This page: http://monitor.mikes.fi/ionclock/
News and notable results
2021-04-30: trapping again after IR-camera BBR-characterization. New machined alu C-beam mount
2021-01-25: first ions trapped after x-mas break. Intensifier+PCO needed for beam-alignment
2021-01-22: new mu-shields delivered. inner and mid-shields installed and fit ok
2020-12-03: Initial measurement of repumper-shift: -2600 (C1) to -2900 Hz (C4)
2020-11-18: New Ion page for clock-run data
2020-11-04: clock run with new HelRes (now 14.424 MHz) and new B-beam direction, but only 800us probe
2020-09-24: Move 2019 results to index_2019
2020-09-24: Secular frequencies with tickler + photon-correlation
2020-09-23: new 14.404 MHz Magic-RF Helical-Resonator
2020-07-02: Fourier limited 4 Hz linewidth with 180ms probe-pulse. Night, no metro.
2020-06-11: Second clock run, OD2 with 4ms probe, only C1, C2 peaks.
2020-05-17: D5/2 lifetime, random-telegraph method with 405nm and fit
2020-04-24: First clock run! 2.5ms probe, stability 2e-13/sqrt(tau_s)
2020-04-01: trapping single ion again, after 405 laser broke
2020-01-31: Move comb-plots to comb
2020-01-15: For temperatures, see DAQ
2019-12-19: Micromotion map, 3D pictures
2019-11-29: Cavity mode number, Comb mode number. Cavity absolute frequency.
2019-11-08: -C2 Zeeman component linewidth estimate of 800 Hz.
2019-11-05: Zeeman peaks with B-shield (1.6 uT estimate).
2019-11-04: Now with first B-shield. Laminated plywood + metglas.
2019-07-12: Zeeman peaks in ambient (66uT estimate) B-field. First clock-spectrum!
2019-07-10: Initial clock-transition search...
2019-07-10: Page created.
2019-06-18: ion photo and secular frequencies
422nm OD0 max power
check if ion still stored? 422nm laser on, all beams, repumper+clearout, 405nm on
10:45local turn down AC to 30Hz - to stop oscillation?
RF-rectifier vs T108 temperature -0.000548 V/C, average ca 0.0586 V, so -0.94 %/C
Copper resistance temperature coefficient around room temperature ca +0.4 %/C
Shut down lasers - leave RF on and ion trapped
Rb-lock too unstable vs. temperature changes? Finfoam on floor limiting air currents maybe helps a bit?
422-laser unlocked again!
ID16991 peak-search OD3+OD2 with 90ms probe
+C1 75 872 649
16:18 ion went dark...
422 laser unlocked from Rb-line
relock to Rb-line restores fluorescence
ID16986 is peak-search with 35 ms probe
turn up FALC-gain, was on "4-5", now "6-7"
+C1 fit fails 75 872 644
-C1 75 845 308
ID16984 is rabiflop on +C1 - OD1+OD3 = OD4 seems to give less than 100-fold tau_pi compared to OD0
fit gives 35.7 ms for +C1
ID16986 is peak search with OD1+OD3 / 40 ms probe
large rabi-sideband, probe too long?
+C1 fit fails, visual 75 872 640
-C1 visual 75 845 305
+C2 75 886 225 (middle of scan, so linecenter and B ok)
-C2 75 831 725
+C4 75 927 124 (left of scan center, B-estimate too high? or two equal peaks?)
-C4 75 790 808 (two equal height peaks?)
ID16980 is peak search with OD2 / 4 ms probe
+C1 75 872 635
-C1 75 845 299
C1_center 75 858 967
comb_center 75 858 971
ID16977 is peak search with OD0 / 400us probe
Measure C1-peaks to find B, then add OD-filter?
comb_center 75 858 975
+C1 75 872 621
-C1 75 845 310
C1_center 75 858 965
B1 4.86644 uT
Counts now lower, because of higher temperature in lab? , ca 30 / 1.5 in 10ms
14:50local turn up A/C from 25 Hz to 35 Hz - (LeCroy 4GHz scope was heating lab?)
Fits to peak-search without (ID16729) and with ASE repumper (ID16736)
400 us probe, 2 kHz Fourier limit
linecenter-comb/ASE off linecenter-comb/ASE on 2020-12-03 result
C1 +42.4 Hz -5741.3 Hz -2683 Hz
C2 -34.1 Hz -5167.4 Hz -2694 Hz
C4 -25.6 Hz -3363.8 Hz -2933 Hz
Counts with newB and OD1.3 (T=5%) filter (slot 10, OD1 is T=10%):
47 / 1.1 in 10ms
505 / 12.4 in 100ms
4704 / 115.6 in 1000ms
Try filter-wheel slot 11, OD1.3 for 422nm beam
10:09 fluorescence restored after maybe 1-2 min of 405nm
422 power to max, check rb-lock, 405nm on
10:00local ion dark?
check uMotion 300kphotons: A('newB') 0.2% in 37s, B('oldA') 1.0%, C 1.2%
(old beam names!)
last tickler run image time-stamp is 06:53local
First voltage is V0:
each run duration ca 4 min per scan, 12 min per voltage. 52 voltages so 12*52 = 10.5 hours
4 voltages * 13 changes = 52 runs in total
Each voltage change set starts and ends with V0: [V0, V0+0.1*dV, V0+0.2*dV, V0+0.3*dV, V0+0.4*dV, V0+0.5*dV, V0, V0-0.1*dV, V0-0.2*dV, V0-0.3*dV, V0-0.4*dV, V0-0.5*dV, V0], length=13
Order is now: V12cm, V12diff, V34cm, V34diff
18:45local start long tickler run
nominal freqs in prediction: 1.931939 0.960010 0.913348
18:20 local z= 1.928443 y 0.958707 x 0.911807
17:50 local z= 1.928327
ID16747 x-scan 18:27
ID16745 y-scan 18:25
ID16744 z-scan 18:20local
ID16738 z-scan better centered
ID16737 search for z 1.928 to 1.932 MHz
Rerun the partially failed secular tickler measurement from last night - now with voltages as planned.
Predicted secular freqs from previous data. Only one voltage is changed in these graphs!
Use prediction to center tickler-scan - in particular for V12cm shift is large. Overall all frequencies may need scaling depending on temperature and helres-tuning
connect RF-rectifier and collect data with DMM
17:40local turn off clock-laser and comb
Similar peak-search to last one, but with repumper override on should give roughly repumper-shift
slightly wider span, expected shift 2-3 kHz optical?
2020-12-02 measured rempumper shift (AOM units)
C1 -1302 -1381 Hz
C2 -1246 -1448
C4 -1349 -1584
Automatic gaussian fit doesn't work very well, need to do manual fits later. Compare against ID16729 normal peak-search.
comb center 75 858 918 (predicted line-center from comb data and AHM3-model)
One more current trial I(-35deg) = [-2.1843 -1.3241 -16.8557]
This looks like the best B-direction found so far.
+A -A +R -R Rmean
C1 0.49 0.45 1256 1265 1230
C2 0.45 0.50 1131 1127 1129
C4 0.47 0.48 1225 1298 1262
ID16730 is flop on +C1
ID16731 is flop on -C1
ID16732 is flop on +C2
ID16733 is flop on -C2
ID16734 is flop on +C4
ID16735 is flop on -C4
ID16729 is peak search, 400us probe (Fourier limit 1000 Hz AOM)
+C1 75.872600, height 50.6, FWHM 1004 Hz AOM
-C1 75.845284, height 48.7, FWHM 955 Hz
C1_center 75.858942 , C1_B 4.867467 uT (comb center 75 858 921)
+C2 75.886185, height 45.5, FWHM 952 Hz (right of scan-center?)
-C2 75.831632, height 50.9, FWHM 875
+C4 75.927077, height 50.4, FWHM 878 (far right of scan-center) - looks OK no big rabi-sidebands now!?
-C4 75.790725, height 45.6, FWHM 991
+A -A +R -R
C1 0.51 0.44 1151 1236
C2 0.40 0.46 1255 1188
C4 0.40 0.50 1332 1396
R_artiq( I(30deg) ) R_artiq( I(28 deg) )
+C -C mean ratio tau_pi(us) +C -C mean ratio tau_pi(us)
C1 1127 1125 1126 1.0 444 1165 1098 1132 1.0 442
C2 1153 1067 1110 0.986 450 1040 1076 (repeat +C2 is 1034)
C4 1297 1383 1340 1.19 373 1267 1444 1356 1.20
Try I(28 deg) currents, I1 = -4.0287, I2 = -1.8824, I3 = -16.854
ID16708 is flop on +C1
ID16710 is flop on -C1
ID16712 is flop on +C2
ID16714 is flop on -C2 - NOTE: clock-laser unlocked at some point..
re-lock laser and redo C2 and C4 - number of points reduced to make it faster
ID16720 is flop on -C4
ID16722 is flop on +C4
ID16724 is flop on -C2
ID16726 is flop on +C2
ID16707 is peak search with 400us probe
+C1 75.872515 (-30 Hz from last peak-searc)
-C1 75.845306 (+86 Hz)
C1_center 75.858911, C1_B 4.84825 uT
+C2 75.886099 (-245 Hz)
-C2 75.831737 (+192 Hz)
+C4 75.926920 (-276 Hz)
-C4 75.790861 (+47 Hz)
Try I(30 deg) currents, I1 = -3.4962, I2 = -1.7305, I3 = -16.876
R_artiq( I(30deg) )
+C -C mean ratio tau_pi(us)
C1 1127 1125 1126 1.0 444
C2 1153 1067 1110 0.986 450
C4 1297 1383 1340 1.19 373
ID16694 is flop on +C1
ID16696 is flop on -C1
ID16698 is flop on +C2
ID16700 is flop on -C2
ID16702 is flop on +C4
ID16704 is flop on -C4
ID16693 is peak-search with 200us probe
-C1 75.845223 , C1_center 75.858884, C1_B 4.86851 uT (comb center 75 858 911 Hz)
+C4 75.927196 (right of scan range, true B higher than estimate?)
-C4 75.790814 (roughly centered in scan range?)
Computed currents for equal strength components:
I(30 deg) = [-3.4962 -1.7305 -16.876] mA
I(28 deg) = [-4.0287 -1.8824 -16.854] mA
Polarization once more.
In A-beam (newB) on breadboard after cube, vertical polarization and polarizer reads 48 at minimum (previous notes had 230 = 50 + 180)
Measurement of 674nm beam 'newB' on ASE side of trap, polarizer reads minimum at 18
so beam polarization is 48-18 = 30 degrees from vertical. direction is CCW when moving along beam, so call this -30 degrees
( 2021-12-17 AW result was -26 )
Matlab-fitting gives R = [896 1038 1823] Hz with amplitudes A = [0.42 0.52 0.49]
Today R = [798 1011 1806] Hz = (798 Hz)*[1 1.27 2.26]
ID16686 is flop on +C1
ID16688 is flop on +C2
ID16690 is flop on +C4
2021-12-17 results with new B-beam and these coil currents have rabi-freq ratios [1 1.20 2.14] = [RC1/RC1 RC2/RC1 RC4/RC1] = [730 877 1564] Hz / 730 Hz
ID16685 is peak-search, OD0/200us, new-B beam. beam is now more focused, so tau_pi shorter? this should find something anyway?
Coil currents from last change I = [-0.8532 2.7827 17.4259] mA
comb_linecenter 75 858 926
C1_Linecenter = 75.858785
C1_B = 4.81956
beatnote ca 15.547 MHz
ECDL piezo now ca 57V for correct TEM00 mode. Cavity camera VGA-box needed reboot to show image of TEM00 mode
10:35local, comb on, clock-laser on.
PMT count monitor, new B beam, with 1000ms gate - until more interesting experiments
uMotion check, 300k photons, after long Vbias run (old beam names in figures)
All secular tickler runs
NOTE bugs in voltage-setting program caused voltages not to be set as planned!!
x outside range
x back in range
Nothing?? first change of V12cm
First few result images:
(note we come back to the null-EMM voltages many times later)
Trap parameter scan range:
Predicted voltage changes (change one voltage from null-EMM values) that move ion +/- 3 um
V12cm +/- 3 V
V12diff +/- 1.2 V
V34cm +/- 4 V
V34diff +/- 7 V
Scheduled measurements in ARTIQ (hope that moving 50% of above voltages in a single step doesn't kick out ion from trap!?)
x 910 - 916 kHz, 80 pts
y 955 - 965 kHz, 80 pts
z 1926 - 1936 kHz, 160 pts
Long list of experiments started ca 19:25 on 2021-12-20
ca 4+4+8 min = 16 min/voltage, 49 rows to measure, duration ca 784 minutes = 13 hours
V0 = [ -6.2635, 0.238, 25.0238, 54.9503] is the EMM-null found by uMotion-optimization
V12cm V12diff V34cm V34diff change from V0 xID yID zID
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16477 16478 16479
-6.2635 0.2380 25.0238 55.6503 0.0000 0.0000 0.0000 0.7000 16481 16482 16483
-6.2635 0.2380 25.0238 56.3503 0.0000 0.0000 0.0000 1.4000 16485 16486 16487
-6.2635 0.2380 25.0238 57.0503 0.0000 0.0000 0.0000 2.1000 16489 16490 16491
-6.2635 0.2380 25.0238 57.7503 0.0000 0.0000 0.0000 2.8000 16493 16494 16495
-6.2635 0.2380 25.0238 58.4503 0.0000 0.0000 0.0000 3.5000 16497 16498 16499
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16501 16502 16503
-6.2635 0.2380 25.0238 54.2503 0.0000 0.0000 0.0000 -0.7000 16505 16506 16507
-6.2635 0.2380 25.0238 53.5503 0.0000 0.0000 0.0000 -1.4000 16509 16510 16511
-6.2635 0.2380 25.0238 52.8503 0.0000 0.0000 0.0000 -2.1000 16513 16514 16515
-6.2635 0.2380 25.0238 52.1503 0.0000 0.0000 0.0000 -2.8000 16517 16518 16519
-6.2635 0.2380 25.0238 51.4503 0.0000 0.0000 0.0000 -3.5000 16521 16522 16523
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16525 16526 16527
-6.2635 0.2380 25.4238 54.9503 0.0000 0.0000 0.4000 0.0000 16529 16530 16531
-6.2635 0.2380 25.8238 54.9503 0.0000 0.0000 0.8000 0.0000 16533 16534 16535
-6.2635 0.2380 26.2238 54.9503 0.0000 0.0000 1.2000 0.0000 16537 16538 16539
-6.2635 0.2380 26.6238 54.9503 0.0000 0.0000 1.6000 0.0000 16541 16542 16543
-6.2635 0.2380 27.0238 54.9503 0.0000 0.0000 2.0000 0.0000 16545 16546 16547
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16549 16550 16551
-6.2635 0.2380 24.6238 54.9503 0.0000 0.0000 -0.4000 0.0000 16553 16554 16555
-6.2635 0.2380 24.2238 54.9503 0.0000 0.0000 -0.8000 0.0000 16557 16558 16559
-6.2635 0.2380 23.8238 54.9503 0.0000 0.0000 -1.2000 0.0000 16561 16562 16563
-6.2635 0.2380 23.4238 54.9503 0.0000 0.0000 -1.6000 0.0000 16565 16566 16567
-6.2635 0.2380 23.0238 54.9503 0.0000 0.0000 -2.0000 0.0000 16569 16570 16571
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16573 16574 16575
-6.2635 0.3580 25.0238 54.9503 0.0000 0.1200 0.0000 0.0000 16577 16578 16579
-6.2635 0.4780 25.0238 54.9503 0.0000 0.2400 0.0000 0.0000 16581 16582 16583
-6.2635 0.5980 25.0238 54.9503 0.0000 0.3600 0.0000 0.0000 16585 16586 16587
-6.2635 0.7180 25.0238 54.9503 0.0000 0.4800 0.0000 0.0000 16589 16590 16591
-6.2635 0.8380 25.0238 54.9503 0.0000 0.6000 0.0000 0.0000 16593 16594 16595
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16597 16598 16599
-6.2635 0.1180 25.0238 54.9503 0.0000 -0.1200 0.0000 0.0000 16601 16602 16603
-6.2635 -0.0020 25.0238 54.9503 0.0000 -0.2400 0.0000 0.0000 16605 16606 16607
-6.2635 -0.1220 25.0238 54.9503 0.0000 -0.3600 0.0000 0.0000 16609 16610 16611
-6.2635 -0.2420 25.0238 54.9503 0.0000 -0.4800 0.0000 0.0000 16613 16614 16615
-6.2635 -0.3620 25.0238 54.9503 0.0000 -0.6000 0.0000 0.0000 16617 16618 16619
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16621 16622 16623
-5.9635 0.2380 25.0238 54.9503 0.3000 0.0000 0.0000 0.0000 16625 16626 16627
-5.6635 0.2380 25.0238 54.9503 0.6000 0.0000 0.0000 0.0000 16629 16630 16631
-5.3635 0.2380 25.0238 54.9503 0.9000 0.0000 0.0000 0.0000 16633 16634 16635
-5.0635 0.2380 25.0238 54.9503 1.2000 0.0000 0.0000 0.0000 16637 16638 16639
-4.7635 0.2380 25.0238 54.9503 1.5000 0.0000 0.0000 0.0000 16641 16642 16643
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16645 16646 16647
-6.5635 0.2380 25.0238 54.9503 -0.3000 0.0000 0.0000 0.0000 16649 16650 16651
-6.8635 0.2380 25.0238 54.9503 -0.6000 0.0000 0.0000 0.0000 16653 16654 16655
-7.1635 0.2380 25.0238 54.9503 -0.9000 0.0000 0.0000 0.0000 16657 16658 16659
-7.4635 0.2380 25.0238 54.9503 -1.2000 0.0000 0.0000 0.0000 16661 16662 16663
-7.7635 0.2380 25.0238 54.9503 -1.5000 0.0000 0.0000 0.0000 16665 16666 16667
-6.2635 0.2380 25.0238 54.9503 0.0000 0.0000 0.0000 0.0000 16669 16670 16671
Once more with 'new B' - the one we want to use for clock-runs
x ID16325 0.912335 MHz, FWHM 696
y ID16326 0.959175 MHz, FWHM 747
z ID16327 1.929434 MHz, FWHM 721 (this is the wide resonance we wanted to see. Note peak-height now low with 26 dB att)
z ID16328 24 dB att, not good fit.. (slightly higher peak)
z ID16329 22 dB att - 1.929455 MHz - FWHM 581 Hz
UGH changing beam names is fun... (these measured with the horizontal A-beam, which in artiq for now we call the B-beam...
Secular resonances should be (much) wider now with more focused B-beam?
z ID16318 26 dB - missed the resonance completely... higher temp has lowered secular freq.
z ID16319 26 dB - slightly above 1.929 MHz - fit 1.929216 MHz, FWHM 257 Hz (not much wider than before? BECAUSE BEAM WAS ALMOST HORIZONTAL...)
z ID16320 26 dB - narrower span - 1.929142, FWHM 282
y ID16321 6 dB - 0.958957 MHz, FWHM 1491 (wider than before)
x ID16322 6 dB - 0.912292 MHz, FWHM 629 Hz
uMotion (300k photons): A 0.6% in 42s, B 0.1% (!) in 38s, C 1.2% in 53s
(images use 'wrong' shutter naming!)
Detuning-scans, for beam-waist check:
ID16304 detuning scan with B-beam - counts up to 22k - 208 MHz counts 9306 / 116
ID16306 detuning scan with A-beam - counts up to 21k - 208 MHz counts 8881 / 111
ID16308 detuning scan with C-beam - counts up to 19k - 208 MHz detuning counts in 1000ms: 7226 / 125
Beam power once more:
A 0.62 uW 912 / 10
B 0.5 uW 957 / 12
C 0.55 uW 683 / 14
try focusing 'in' CW -> better! 994 / 12 counts - good enough(?)
slightly lower counts 557
try focusing B-lens slightly outward (CCW) from last position
Counts and 422nm power before any beam adjustments:
use the old beam-names
A 0.6uW 910 / 12
B 0.66uW 580 / 13
C 0.65 uW 799 / 14
Beam naming-convention in ARTIQ. Names A/B/C correspond to TTL channels connected to shutter drivers
change committed to master, but not in use yet? (requires artiq-restart?)
Device-db file should be moved one folder up - and then artiq-dashboard and artiq-master re-started
"a_beam": "ttl14", # horizontal beam, camera side NOT for ionization & loading anymore
"b_beam": "ttl11", # 'new B-beam' - use for ionization & loading from now on!
"a_beam": "ttl11", # old A used for ionization and loading
"b_beam": "ttl14", # old B, had polarization problems.
10:40local back to null-EMM voltages V12cm , V12diff, V34cm, V34diff = [ -6.2635, 0.238, 25.0238, 54.9503]
V34cm change x y z
+2 914983 961497 1930748
+1 914371 961031 1931408
0 913724 960500 1932393
-1 912665 959473 1932167 (temperature changes in lab..)
(Temp. coefficient of Copper resistance +0.393% / C around room-temp, so temperature rise causes lower secular freqs)
(work in lab sometime during this scan - temperature changes!?)
V34cm to -1 V
V12cm , V12diff, V34cm, V34diff = [ -6.2635, 0.238, 24.0238, 54.9503]
x ID16264 10 dB att, resonance far left in scan
x ID16265 10 dB att, 0.912665 MHz, FWHM 531 Hz
y ID16266, 10 dB
z ID16267, 26 dB, wide span
z ID16268, 26 dB, narrower span
z ID16269, 26 dB, narrower span
x 10 dB
y 10 dB
z 26 dB
z 26 dB
V34cm to +2 V
V12cm , V12diff, V34cm, V34diff = [ -6.2635, 0.238, 27.0238, 54.9503]
z ID16256 - 26 dB att, 1.930748 MHz, FWHM 278 Hz
y ID16257 - 6 dB att, 0.961497 MHz, FWHM 591 Hz
x ID16260 - 10 dB att, 0.914983 MHz, FWHM 476 Hz
z 26 dB
y 6 dB
x 6 dB dip at resonance, too much tickler?
x 8 dB dip at resonance
x 10 dB
Try voltages where EMM not minimized, +1 V on V34cm:
V12cm , V12diff, V34cm, V34diff = [ -6.2635, 0.238, 26.0238, 54.9503]
x ID16252 - 6 dB att, 0.914371 MHz, FWHM 578 Hz
y ID16253 - 6 dB att, 0.961031 MHz, FWHM 796 Hz
z ID16254 - 26 dB att, 1.931408 Hz, FWHM 215 Hz
x 6 dB, ca +700 Hz from null-EMM voltage
y 6 dB, ca +530 Hz from null-EMM voltage
z 26 dB, ca -1000 Hz from null-EMM voltage
Secular frequencies at V12cm = -6.2635 V
z ID16246 - 22 dB tickler too strong (beam not focused very well)
z ID16247 - 24 dB att.
z ID16248 - 26 dB att. 1.932393 MHz, FWHM 230 Hz
y ID16249 - 6 dB att. 0.960500 MHz, FWHM 677 Hz
x ID16250 - 6 dB att. 0.913724 MHz, FWHM 419 Hz
z 24 dB, asymmetric?
z 26 dB
y 6 dB, motion along camera-axis - not much movement seen in image
x 6 dB, motion in camra plane clearly side-to-side
stop fluorescence monitor
ion still stored. counts show same lower stdev now - before dark-period stdev was higher (after a few tickler-measurements)
One beam low-power 422nm with repumper+clearout is not sufficient to restore fluorescence
405nm light seems to be essential - not clear if all beams and/or high 422-power required also?
17:41 back to OD1-filter and only 'new A' beam - counts ca 6500 / 1s
up to 35000-40000 counts with all beams at max power
ca 1 min of 405nm laser miraculously restores fluorescence!?
ca 17:30 422-laser seems locked to Rb-line - remove OD-filter for max power in all beams
ca 16 ion lost? all beams on, clearout on (ion loaded 2021-12-14 Tuesday)
10:37 continue with fluorescence monitor.
Back to original voltages V12cm , V12diff, V34cm, V34diff = [ -6.2635, 0.238, 25.0238, 54.9503]
strange results - these trap-parameters somehow bad/unstable?
take 10% step toward V12cm=0
V12cm , V12diff, V34cm, V34diff = [ -5.6372, 0.238, 25.8404, 54.9503]
prediction ca -19 kHz for z, +17..18 kHz for x and y
f_secular = [ 1.913]
fwhm = 
not good fit
21 dB too much excitation?
24 dB still too much?
26 dB still too much?
Vbias = [-6.2635, 0.2380 , 25.0238 , 54.9503 ]
f_secular = [0.914182 0.960982 1.933258] MHz
fwhm = [421 707 220] Hz
09:22 stop flurescence monitor.
17:25local leave lab
comb off, clock-laser off
New A-beam not so well focused, but polarization should be more stable?
Leave PMT count monitor with 1000ms gate
absolute angle for new measurement: vertical is 140 (from cube reference measurement) - beam max 114 - so absolute angle is -26 degrees (moving along direction of beam CW is + and CCW is -)
new polarization rotated by 76 degrees from old?
previous KH results for same minimum 100 / 280 !?
'new A' beam minimum when polarizer at 24 degrees, or 24+180 = 204 degrees. max at 90 degrees to this, so (114 and 294) ?
Polarizer in A-beam on breadboard (should be vertically polarized after cube): min power at angle 230 (max ca 138?) (delta 92 deg - ok)
A- A+ R- R+
C1 0.51 0.42 702 758
C2 0.46 0.46 889 864
C4 0.46 0.51 1542 1586
ID16210 flop on +C1 - visually tau_pi 700us
ID16211 is -C1 700-800us maybe
ID16212 is +C2 shorter ca 600us
ID16213 is -C2
ID16214 is +C4 shortest, ca 350us
ID16215 is -C4
New A-beam might have less 674nm power, so try peak-search with 400us probe?
ID16209 new A-beam peak-search, 400us probe
Initial alignment, PMT counts:
A 479 / 12 in 100ms - needs focusing for optimum waist?
B 891 / 11 (old 'A' - which was best focused?)
C 529 / 14 (not changed)
NOTE: From this note forward A means the tilted beam at the back of the breadboard, camera side, and B means the nearly horisontal beam at the front, also camera side.
Exchange A and B fibers. Focusing of objectives might have changed sliglty - polarization direction also... Objectives were not exchanged.
Matlab Fits for Amplitude and Rabi freqs, B-beam, I = [-0.8532 2.7827 17.4259]
A- A+ R- R+
C1 0.43 0.54 1954 1981
C2 0.57 0.41 1940 1976
C4 0.57 0.43 1963 2052
ID16173 is flop on +C1, duration ca 230s
ID16174 is -C1
ID16175 is +C2 (first max noisy?)
ID16176 is -C2
ID16177 is +C4
ID16178 is -C4
ID16172 is peak search, B-beam, OD0/200us with latest coil-currents
Centers and FWHMs: [75.87159056 75.84463746 75.88520807 75.83124737 75.92563863 75.79072803] [0.00197299 0.00187072 0.00185356 0.0020001 0.00182339 0.00205466]
Next coil-currents to try I = [-0.8532 2.7827 17.4259]
Matlab Fits for Amplitude and Rabi freqs
A- A+ R- R+
C1 0.41 0.56 2222 2237
C2 0.57 0.41 1708 1691
C4 0.57 0.40 2108 2082
Rabi-freqs R = [ 2196 , 2166 , 1570 , 1702 , 2088 , 2062] Hz (+C1, -C1, +C2, -C2, +C4, -C4). tau_pi = 0.5*(1/R)
ID16151 is rabiflop on +C1
ID16153 is -C1
ID16155 is +C2
ID16157 is -C2
ID16159 is +C4
ID16161 is -C4
ID16150 is peak search with OD0/200us - new B-direction
+C1_Center = 75 871 696
-C1_Center = 75 844 637 (B1 = 4.821791349788093 uT)
+C2_Center = 75 885 243
-C2_Center = 75 831 149
+C4_Center = 75 925 859
-C4_Center = 75 790 594
set new coil-currents I = [-1.0079 1.7551 17.8636] mA
09:50 re-lock clock-laser. also enable temperature-control "toptica_dlc_tloop.py"-script to keep FALC-integrator within +/-2V - clock-laser should stay locked forever? (without mechanical disturbances in lab)
Try new B-direction, for equal C1/2/4 strength, possibly less sensitive to polarization-changes also? (see PMT count timeseries for clock-run vs. temperature..)
Old currents I = [1.3378 2.1205 17.0758] mA, clockrun shows abs(B) = 4.8246 uT
new suggestion I = [-1.0079 1.7551 17.8636] mA,
ID16144 is uMotion optimization, 4x4x4 grid centered on previous optimum, now 300k photons/histogram for lower unc.?
run duration ca 2 hours
Optimal voltages (Common mode, diff): [ 0.23866464 25.03642967 55.00744471]
at optimal point 300kphoton uMotion: A 1.2%, B 0.7%, C 0.8%
EQS at final voltages from yesterday (note bias-voltage changed for better uMotion since last 1h "100%" dataset of V12cm-scan)
EQS_zeroed = [ -0.298 (0.269) , -0.040 (0.242) , +0.337 (0.262) ]
compare to last 1h dataset in V12cm scan:
EQS('100%') = [-0.278 (0.143), -0.259 (0.131) , +0.537 (0.175) ] n=7 servo-cycles
Uncertainties not scaled by 1/sqrt(n_servocycles)
eqs-slope figure shows stepped V12cm voltage yesterday, and clock-run after micromotion-optimization.
Note temperature changes in lab causes clock-laser disturbance that peak-servos are unable to track -> also eqs-slope disturbance
Both ion and comb see these disturbances - so look for temperature sensitive components common to both paths?
0800local stop clock-run
clock-laser unlocked, FALC integrator at 4V limit, need to use slow temperature control to keep integrator below e.g. +/-2V
ID16142 try if clock-run will start, using latest stored line-centers. B-beam, OD6/120ms as before
Analysis of secular results:
[ax ay az] = [-0.001027260759455 0.000624538003714 0.000402718960605], qz = 0.365023572362448
old dV = [1, 0, 1.30, 0.02]
new dV = [1.0000 -0.0022 1.3037 -0.0329]
values close to zero can be rounded to zero?
RID16132 is uMotion minimization in 4x4x4 = 64 pts grid, 100 kphotons/histogram, V12cm constant at -6.2635 V
Optimal voltages (Common mode, diff): [ 0.23801151 25.02381342 54.95032671]
uMotion at found optimum: A 0.64%, B 1.25%, C 1.39%
predicted secular freqs @ Vbias_EQS-zero = [0.9153 0.9640 1.9215] MHz
experimental seculars = [0.907454 0.954396 1.919689]
RF-monitor-scope: 3.34 Vpp
uMotion A 8.2%, B 9.4%, C 6.1%
17:02local stop the clock-run for now.
EQS looks linear as function of V12cm. Experimental EQS changes sign at predicted optimum, so use slightly lower V12cm?
Quantum-jump counts are not so good, but 1h of data for each parameter-set looks sufficient with OD6/120 ms probe
16:13 to 100% Vbias (predicted EQS-zero - but polyfit to 10%..80% points show zero-crossing at V12cm = -6.0..-6.05 V)
Vbias = [-6.2635 0.2243 25.047 54.619 ]
V1 = -6.4878, V2 = -6.039199999999999, V3 = -29.572, V4 = 79.666
15:32 to 90% Vbias
Vbias = [-5.6372 0.2243 25.8613 54.6315]
V1 = -5.8615, V2 = -5.4129, V3 = -28.770200000000003, V4 = 80.4928
14:54 to 80% Vbias
Vbias = [-5.0108 0.2243 26.6755 54.6441]
V1 = -5.2351, V2 = -4.786499999999999, V3 = -27.968600000000002, V4 = 81.31960000000001
14:09 to 70% Vbias
Vbias = [-4.3845 0.2243 27.4898 54.6566 ]
V1 = -4.6088000000000005, V2 = -4.1602, V3 = -27.1668, V4 = 82.1464
12:56 to 60% Vbias
Vbias = [-3.7581 0.2243 28.304 54.6691 ]
V1 = -3.9824, V2 = -3.5338000000000003, V3 = -26.3651, V4 = 82.9731
11:57 to 50% Vbias
Vbias = [ -3.1318 0.2243 29.1183 54.6817 ]
V1 = -3.3561, V2 = -2.9075, V3 = -25.563399999999998, V4 = 83.8
11:05 to 40% Vbias
Vbias = [-2.5054 0.2243 29.9325 54.6942]
V1 = -2.7297, V2 = -2.2811, V3 = -24.7617, V4 = 84.6267
EQS_30percent = [10.19 2.68 -12.88] = [69% 72% 69%] * EQS_120ms_2021-12-15 (EQS looks linear vs. dV vector changes)
09:32 30% Vbias
Vbias = [ -1.8791 , 0.2243 , 30.7468 , 54.7067 ]
V1 = -2.1034, V2 = -1.6548, V3 = -23.959899999999998, V4 = 85.45349999999999
08:33 take step to Vold + 20%*6.263535*dV
Vbias = [-1.2527, +0.2243, +31.5611, +54.7192]
V1 = -1.4769999999999999, V2 = -1.0284, V3 = -23.1581, V4 = 86.2803
07:34local take a step of 10%*6.263535*dV towards the zero-EQS bias voltages
Vbias now = [-0.6264, 0.2243, 32.3753, 54.7318]
(V1 = -0.8507, V2 = -0.40209999999999996, V3 = -22.356499999999997, V4 = 87.1071)
With old bias voltages, 8h clockrun 139 servo-cycles:
EQS_120ms_2021-12-15 = [+14.85 (0.23) , +3.72 (0.27) , -18.57 (0.28)] Hz - unc. NOT scaled by 1/sqrt(n_cycles) !!
EQS_40ms_2021-12-14 = [+14.86 (0.50) , +3.69 (0.51) , -18.55 (0.56)] Hz
Clockrun with 120ms probe?
6-peak cycle ca 206 s
suggested bias-voltages for clock-run with low EQS:
Vold = [0, 0.22433, 33.1896, 54.7443]
dV = [1, 0, 1.30, 0.02]
Vnew = Vold - 6.263535*dV
= [-6.2635, 0.22433, 25.0469924, 54.6190]
EQS_theory = [-0.0664 -0.0166 0.00830] Hz for C1, C2, C4
ID16017 is peak search with OD6, 100 ms probe, 0.8 Hz step, 100Hz span, 125 pts/peak
search initialized with comb_center, B-field estimate, and experimental EQS from y-day
Switch back to the normal fiber-noise servo
C1 and C4 can be attenuated more, to get same tau_pi as C2?
ID16086 is flop on +C1 with 19 dB DDS attenuation -> 233 us
ID16088 is flop on +C1 with 20 dB attenuation -> 245.8 us
ID16090 is flop on +C1 with 21 dB attenuation -> 297.4 us
ID16092 is flop on +C1 with 22 dB attenuation -> 327.3 us
tau( 19 dB ) / tau( 18 dB ) = 233 / 215.4 = 1.08
tau( 20 dB ) / tau( 18 dB ) = 245.8 / 215.4 = 1.14
tau( 21 dB ) / tau( 18 dB ) = 297.4 / 215.4 = 1.38
tau( 22 dB ) / tau( 18 dB ) = 327.3 / 215.4 = 1.52
ID16080 is rabiflop on +C1, 18 dB att, 20us to 2000us. estimate of tau_pi = 200us is roughly right
ID16081 is flop on +C2
ID16084 is flop on +C4
tau_pi = [215.4 292.2 276.7] us % C1, C2, C4
ID16079 is peak-search with B-beam, no OD filter, 200us probe, ca 1.1uW 674nm
200Hz step, 10 kHz span, 51 pts/peak, Fourier limit 2000 Hz AOM
+C1 75 871 835, FWHM 1919 Hz AOM
-C1 75 844 761, FWHM 1905 Hz
+C2 75 885 306, 2189 Hz
-C2 75 831 281, 2050 Hz
+C4 75 925 734, 1841 Hz
-C4 75 790 772, 2007 Hz
previously this gave tau_pi ca 200 us - try that?
P_674_B = 1.35 - 0.25 = 1.1 uW (no OD-filter, usual 18 dB DDS attenuation)
remove OD-filters from 674nm beam - try fast rabiflops instead? Zeeman-AOM attenuation could be tuned to give same tau_pi for all components?
ID16076 is rabiflop on -C4 75 790 771 Hz
too much drift to measure flops wit OD5 and ca 40 ms probe...
ID16072 is B-beam peak search with 40ms probe, 200 Hz AOM span, 2 Hz AOM step, "2021-12-15-search3"
NO Fiber-noise cancelling servo - AOM driven with fixed frequency
+C1 75 871 802 (visual approx center)
(earlier OD5 spectra with 60ms probe, which is possibly too long?)
-C1 75 844 736
+C2 75 885 257
-C2 75 831 291 (why is this peak right of scan-center?)
+C4 75 925 788 (? right of scan center)
-C4 75 790 771 (right of scan center)
Try clock-spectrum withoput fiber-noise cancellation?
Instaead of servo+VCO, drive the fiber-noise AOM with fixed 178.5 MHz from maser-referenced R&S signal generator (+1dBm level seems ok)
phase seems to wander 2pi every few-seconds to 10s on scope - so this should not cause major shift in light to ion? (scope sees fiber 2-way, ion only 1-way)
re-lock to right TEM00 mode
13:10local: check uMotion: A 3.2% in 30.3s, B 4.0% in 25s, C 3.0% in 36s, (RIDs 16069 16070 16071)
Clock-laser jumped to TEM01 mode at some point during temperature-swings..
Lab temperature-swing of -1 C and then recovery +1.75C causes beat-note change ca 100 Hz IR, so 200 Hz red!?
Cannot reasonably be real cavity-length changes, since cavity-spacer should have >5 day time-constant
Fiber from ECDL to comb is unstabilized, also unstabilized is ECDL-to-cavity fiber.
09:25local abort clock run, probably not locked to peaks anymore
ca09:00 large ambient changes, MIKES-building service
10 first servo-cycles discarded, also discarded all data after servo-cycle nr 450 - since ambient temperature swings started then
suggested Z-shift sensitivities:
revised theory estimate EQS = [+14.334 +3.583 -18.917]
based on 8h of clock-run data (257 servo-cycles) experimental EQS = [+14.86(0.50) +3.69(0.51) -18.55(0.56)] Hz
ID16067 is clock-run again, 40ms probe, with initialized peak-positions including EQS = [+11.67 +2.92 -14.59] Hz optical
line-center from comb measurement, B-estimate from C1 measurement
Improve peak-position program by adding variable EQS, now:
print:[75.87180381 75.84473186 75.8852534 75.83127352 75.92577057 75.79073884]
ID16063 is new clockrun, 40ms probe, initialized-values from C1-search ID16061
Predicted peak-centers from C1_center and C1_B (no EQS?)
predicted (simple model) This run
+C1 75 871 804.52
-C1 75 844 732.48
+C2 75 885 258.53 (last clock-run -9 Hz ??)
-C2 75 831 278.47
+C4 75 925 784.59
-C4 75 790 752.41 (last clock-run -8 Hz ??)
ID16061 is peak-search with 40ms probe. 2Hz step, 50 Hz span, 25 pts/peak
+C1_Center = 75 871 804.5 (visually 75 871 804 previous clock-run was locked ok to this peak)
-C1_Center = 75 844 732.5 (visually 75 844 732)
C1_Center = 75 858 268.5
C1_B = 4.823956292687948
comb_linecenter 75 858 260
ID16059 is rabiflop on +C1 = 75 871 800, tau_pi ca 40 ms?
results look strange - abort..
ID16057 try to start a clock-run. B-beam, 60ms probe, db-tag "2021-12-14-clockrun1"
Program is "Lock to transitions C1,C2,C4" (hopefully this is the latest one..)
Peak-positions initialized with comb_linecenter and C1_B estimate below (initialization assumes no EQS?)
With large EQS and large 50 Hz side-bands there might be danger of locking to the wrong peak?
Results on Ion page
After 10 servocycles peak-positions:
All frequencies AOM! (25 Hz or 50 Hz AOM offset indicates lock to 50 or 100 Hz side-band?)
clock-run spectrum-scan(ID15980) Delta
C1+ freq: 75 871 804.8 75 871 796 +8
C1- freq: 75 844 732.7 75 844 737 -5
C2+ freq: 75 885 249.2 75 885 249 0
C2- freq: 75 831 277.2 75 831 288 -10 ?
C4+ freq: 75 925 784.0 75 925 754 +30 !!?
C4- freq: 75 790 744.3 75 790 759 -15 ?
Update AHM3 model from Atomic Bulletin 2021-12
MJD0 = 59544
y0 = 4.47416e-13
y1 = -2.64742e-15
y_AHM3 = y0 + y1*(mjd - MJD0)
ID16055 peak-search, B-beam, 60 ms
+C1 visually 75 871 806 (Gaussian fit fails badly because of side-bands..)
-C1 75 844 735 (gaussian fit fails)
C1_center 75 858 271
C1_B 4.82377 uT
comb_linecenter 75 858 264 Hz
After beam-changes uMotions again: A 3.0%, B 3.8%, C 2.7%
Beam power (OD1 for 422nm):
P_repumper = 7.9 mW
P_422_A = 0.75 - 0.17 = 0.58 uW
P_422_B = 0.82-0.22 = 0.6 uW
P_422_C = 0.7 uW
Measure detuning-scans to verify if C-waist is now improved?
C was clearly worst previously, now between A and C
ID16033 is A-beam detuning scan, 1000ms gate, counts 11069 / 148 - max counts approach 25 000
ID16036 is B-beam detuning scan, 1000ms gate, counts 7189 / 124 - max counts ca 17 500
ID16039 is C-beam detuning scam, 1000ms gate, counts 8535 / 142 - max counts ca 20 000
Now counts in 100ms:
A 1093 / 15.6
B 624.5 / 12.4
C 794 / 14.5
422nm laser filter-wheel for attenuating beam
Slot OD Transmission % Linear att. % Filter
1 0 100.0 0.0 no filter
2 0.1 79.4 20.6 NE501B-A
3 0.2 63.1 36.9 NE502B-A
4 0.4 39.8 60.2 NE504B-A
5 0.5 31.6 68.4 NE505B-A
6 0.6 25.1 74.9 NE506B-A
7 0.7 20.0 80.0 NE507B-A
8 0.8 15.8 84.2 NE508B-A
9 0.9 12.6 87.4 NE509B-A
10 1.0 10.0 90.0 NE510B-A
11 1.3 5.0 95.0 NE513B-A
12 1.5 3.2 96.8 NE515B-A
back to OD1-filter (slot 10) and lower 422nm power.
finally got C-beam through both shields and output-beam visible + ion fluorescence.
Tried C-beam focusing and alignment - but probably need to remove mid-shield and/or inner shield to get C-beam working again?
A counts ca 1000 / 100 ms
ID15982 is B-couns in 100ms when heating ion-breadboard B-fiber by hand. Counts from 600 to 1100 in 100ms
heat/stress applied from ca "50" to "150" on time-axis, then removed
B-counts 72.9 / 2.8 in 10ms
ID15980 is same as 60ms peak-search that failed mid-way yesterday
smaller 200 Hz AOM span - this finds carrier and sidebands if we are well centered on the peak..
+C1 done 10:45local 2021-12-14-search1C1+.txt
-C1 done 11:23local 2021-12-14-search1C1-.txt
+C2 done 2021-12-14-search1C2+.txt (50% done 11:42)
-C2 done 2021-12-14-search1C2-.txt
+C4 done ca13:20 2021-12-14-search1C4+.txt
-C4 done 13:55 2021-12-14-search1C4-.txt
comb_center 75 858 261 Hz (-6 Hz from yesterday)
RID15969 is detuning-scan with A-beam, 1000ms gate. Counts at 208 MHz detuning 11952 / 350 in 1000ms
RID15974 is detuning-scan, B-beam, 1000ms, counts 9444 / 325
RID15977 is detuning-scan, C-bean, 1000ms, counts 5211 / 352
(lab-lights off, but lights on in small lab, no holy blanket)
uMotion: A 2.1%, B 1.9%, C 2.0%
07:42Z load with 1.70A. Ion at 4min30s. Note: loaded with 208 MHz detuning and OD1 filter (not full 422 power). Fluorescence counts during detection: ID15962.
PMT-counts rise slightly before ion visible on camera - optmized automatic loading could shut dispenser some seconds sooner than manual operator
07:30Z 422-laser seems locked OK to Rb-line - not the cause of ion-loss..
ion lost during -C2 scan? why is storage time now only 1-2 days?
Plots from clock-laser data over weekend 2021-12-10 (Friday) to 2021-12-13 (Monday) - ca 65 hours.
Data folder "2021-12-13_clocklaser"
18:30local leave lab
ID15958 is peak-search with OD5, 60ms probe. 1Hz step, 300 Hz span, 301 pts/peak
Fourier limit 6.7 Hz AOM
Rabi side-band maybe 10-11 Hz AOM from carrier, 50 Hz AC-sidebands should be 25 Hz AOM away!?
+C1 is slightly right of center in scan (B-estimate is slightly too low?) - but this scan is wide so should capture all peaks anyway
Add OD3+OD2 filter - expect 60 ms probe?
ID15956 is peak-search, B-beam, OD2 filter, 2ms probe, 40Hz step, 1 kHz span, 25pts/peak
+C1_Center = 75 871 802 (80 Hz shift from OD0 result), FWHM 180 Hz vs. Fourier 200 Hz
-C1_Center = 75 844 754 (maybe 70 Hz shift), FWHM 198 Hz
C1_linecenter = 75 858 278 ( comb prediction is 267 Hz for last three digits)
C1_B = 4.819684 uT
+C2_Center = 75 885 234
-C2_Center = 75 831 298
abort search and add more OD
initialize peak-search with comb_linecenter and found abs(B) = 4.8226 uT
Add OD2 filter for 674nm - expect 2 ms tau_pi?
tau_pi = [0.22 0.31 0.29] ms = 0.22 ms * [1 1.41 1.32] # [tau_C1 tau_C2 tau_C4]
ID15953 is rabiflop on +C4, 20us to 2ms, 50pts, 100shots
ID15951 is rabiflop on +C2, 20us to 2ms, 50pts, 100shots
ID15949 is rabiflop on +C1, 20us to 2ms, 50 pts, 100shots
ID15948 is peak-search with B-beam, 0.2 ms probe, 10 kHz span, 400 Hz step, 25 pts/peak
+C1_Center = 75 871 882
-C1_Center = 75 844 818
C1_Linecenter = 75 858 350 (comb preduction still 75 858 270 - but this probe is so short maybe Hz-level agreement not expected)
C1_B = -4.82258108765977 uT
+C2_Center = 75 885 387
-C2_Center = 75 831 361
+C4_Center = 75 925 782
-C4_Center = 75 790 772
ID15947 is peak-search with B-beam, 0.5 ms probe, 4 kHz span, 160 Hz step, 25 pts/peak
still get a double-peak? reduce probe further
ID15946 is peak-search with B-beam, 1ms probe, 2 kHz span, 80 Hz step, 25 pts/peak, "2021-12-13-search2"
+C1 looks like probe is too long - reduce.
Initialize peak-search scan with comb_linecenter and 4.8 uT estimate
Set coil currents for 4.8 uT field, I1 = 1.3378, I2 = 2.1205, I3 = 17.0758
based on polarization measurement suggested B-field for equal C1,2,4 sensitivity:
B03 = [1.18465 0.424435 -0.27965] % residual field at zero currents
I_3.2 = [-0.5988 1.0765 11.7275] mA % abs(B) = 3.2 uT
I_4.8 = [ 1.3378 2.1205 17.0758] mA % abs(B) = 4.8 uT
resulting B-direction is close to vertical [0.3206 0.2738 0.9068]
predicted EQS dnuEQSfh is -14.5875 with V12cm=0, and -0.0640 with V12cm = -6.2 V
B01 = [1.00994 -0.25657 -0.28325] # 2020-06-12
B02 = [1.08592 1.29155 -0.10010] # 2020-08-19
B03 = [1.18465 0.42444 -0.27965] # 2021-01-26, new muShields
comb_linecenter 75 858 269 - similar to last week
B counts 97 / 3 - threshold 17
P_674_B = 1.14 uW
uMotion A 1.8%, B 1.8%, C 4%
OD1 filter for 422nm
focus-adjustments seem to shift the beam position enough so that sometimes we hit the mid-shield on the C-objective side!?
C after adjustments 2400 counts - beam size on through-transmission side is indicator of where we are focused
A B beams with full power ca 3600 / 100ms
old lock-ring ca 2.78mm thick and rotating objective was very close to this distance from mount
replace 0.5" lock ring with thinner one found in lab
right focus is final lense very close to mount - thinner lock-ring would be good?
a bit more outward from mount - counts get worse, ca 600 / 100ms
try to focus C-objective, final lens slightly outward from mount - similar counts 700/100ms
A counts 1009 / 27 bright/dark in 100ms
B counts 1109 / 26
C counts 718 / 29
add OD1-filter for 422nm - 10-fold less power still gives 1/3rd of counts
No OD-filter for 422nm - "full power"
P_422_A = 5.8 uW
P_422_B = 6.8 uW
P_422_C = 7.6 uW
B-alignment -> ca 3000 / 100ms counts
C-beam alignment -> also ca 3000 / 100ms counts
A-counts ca 3000 / 100ms with full 422nm power
A-beam has stayed aligned, but B and C show very low counts - alignment moved while measuring polarization?
Detuning set to 208 MHz
13:39:10Z ion loaded.
13:30:30Z Try to load ion, 1.7A
461nm laser against dispenser: 196.26 mA
13:15Z Try to load ion, 1.6A. No ion after 11 min
422nm detuning to 185 MHz
Polarization for 422/674nm ABC beams measured.
Cavity inner shield temperature has stabilized. Doors were closed 2021-12-08 Wednesday ca 17:15local - ca 5 days ago
Temperature over weekend was stable. Clock-laser box is closed, A/C on 'low' 25.00 Hz (since 2021-12-07 12local) - no people in the lab
Note short visit by AW on Sunday (2021-12-12) lowers temperature permanently (why?)
good clock-laser dataset - little or no work in the lab + stable temperature
start 2012-12-10 17:00 local
end 2021-12-13 10:39 local
Computing optical frequencies from beat-note data
ncomb = 889558
f0 = -35e6 # NOTE negative f0!!
frep = 249 999 896 Hz # comb DDS 19 999 584 Hz ( 1/4 th of difference from 20 MHz goes to rep-rate )
f_IR = f0 + ncomb*frep - fbeat_ch1
f_red = 2*f_IR # usually -330 MHz SRS
f_ion_breadboard = f_red + 178.5 MHz # fixed frequency AOM for fiber-noise cancelling, 2*Zeeman_AOM shift (ca 160 MHz) remains to ion line-center
f_cavity = f_red - 2*drift_DDS # usually -490 MHz SRS
# ch1 + ch2 sanity check
fbeat1 + fbeat2 = frep
# optical frequency from ch2
f_IR_ch2 = f0 + (ncomb - 1)*frep + fbeat_ch2
stop clock-laser measurement while working in lab
10:38local stop fluorescence monitor
Try 405nm laser to revive ion? also 461nm on - no ion?
Add more points to the comb ADEV plot - is there some oscillation with 1h period?
Plot also power-spectrum? Good data since Friday - so we have much more than longest plot of 16h
dark-counts slightly higher with all 422nm beams ABC
picoscope on comb-computer had crashed (out of memory) - can't run picoscope continuously
try 405nm laser for some minutes, all beams ABC - but no ion..
14:45 local, ionization shutter was on manual, so no 405nm went to ion earlier.
09:20 no ion, but continue with fluorescence monitor anyway
405nm on - 262-267 counts with TTL9 on - 252-263 counts with TTL9 off - not sure if SRS shutter in TTL or Manual mode..
clearout on - no ion
detuning-scan - no ion
all beams on - no ion
lab temperature looks stable, clock-laser and comb OK
09:03 stop monitor
ion went dark?
20:05 start fluorescence monitor with B-beam, runs for 24 hours.
ID15860 peak-search with 7ms probe
RID15857 rabiflop on -C2(?) 75.86526
Try 50R temrinators on I1,2,3 cables next?
ID15856 is peak-search with better B-estimate 1.2489 uT
15ms probe, B-beam, OD3 filter, 5 Hz step, 500 Hz span
+C1_Center = 75 861 770 (-1 Hz offset from last scan)
-C1_Center = 75 854 756 (-6 Hz from last - Gaussian not good fit to Rabi-shape)
C1_Linecenter = 75 858 263 (-4 Hz)
C1_B = 1.24990 uT
ID1584 is peak-search with zero currents, assuming B=1.29 uT. 2021-12-11-search3
15ms probe, 5Hz step, 500Hz span, 101pts/peak
+C1 left of assumed position, so true B0 is lower?
+C1_Center = 75 861 771
-C1_Center= 75 854 762 (right of scan center, as expected)
C1_Center = 75 858 267
C1_B = 1.2489 uT
comb_center 75 858 272
C1 peaks: (switch positions so -C1 is left and +C1 is right)
abort search, re-start with better B-estimate
2021-01-27 notes have abs(B0) = 1.29 uT
set currents to zero.
Coil currents were (-8.6056, -1, +7.0059) mA, and peak-search positions B = 2.0227 uT
ID15851 is peak-search with 15ms probe, 5 Hz step, "2021-12-11-search2"
59pts/peak - duration maybe 1/3 of previous search, so 40 min?
start ca 11:25 local
tau_pi at least 4x shorter for C2, C4, compared to C1?
09:20 local, probe is much too long for C2 and C4? Try peak-search with 30ms probe
ID15850 is peak-search with B-beam, OD3-filter, 30ms probe
Span 300Hz, step 2Hz, 151 pts/peak, Fourier limit 13 Hz AOM
+C1_Center = 75.86394799482541
-C1_Center = 75.85260861772092
C1_Linecenter = 75 858 278
C1_B = 2.0205589993747646
Comb_linecenter 75 858 273 Hz
peak-search end ca 11:25, so duration 2 hours
17:05local leave lab
RID15849 is peak-search with OD3-filter and 60ms probe, B-beam
Span is 300 Hz AOM, step 1 Hz, 301 pts/peak. Fourier limit 6.7 Hz AOM
RID15848 is rabiflop on -C1, B-beam, OD3 filter - counts are low but maybe 60ms probe OK?
This is close to 1/50 Hz - so will not resolve side-bands very well? More OD and longer probe?
RID15845 is rabiflop on -C1, B-beam, OD2 filter
first maximum is low, so we might not hit the center of the peak? tau_pi ca 27 ms
Switch to OD2 filter for 674nm
RID15841 is rabiflop on -C1, B-beam, OD1 filter for 674nm
RID15840 is peak-search, B-beam, 11ms probe, "2021-12-10-search6"
+C1_Center = 75.86393639139541
-C1_Center = 75.85259678647058
C1_Linecenter = 75.85826658893299, C1_B = 2.020599594588025
P_422_C = 0.7 uW
P_422_B = 0.65 uW
P_422_A = 0.5 uW
P_ASE = 7.9 mW
RID15830 is A-beam detuning scan, improved ASE-focusing/alignment
RID15831 A-beam counts 12220 / 525 in 1000 ms (lab lights on, no holy blanket)
RID15833 is B-beam detuning scan
RID15833 B-beam counts 11167 / 587 in 1000 ms
RID15836 is C-beam detuning scan
RID15837 C-beam counts 6099 / 565 in 1000 ms
Counts reduce to ca 1800/100ms with OD1-filter for 422nm
These counts with all beams ABC and max power from 422nm laser.
Initially ca 1000 counts/100ms - slight improvement to 1100-1200, then further focusing and >4000 counts!
Try focusing ASE beam better - screw final lens 'in', closer to fiber-launch
Seculars [1.088909, 1.128308, 1.720621] (compare to 2021-12-02 results (1.089, 1.129, 1.722) )
radial modes with 20+4 dB, axial mode now with 20+18 dB (previously 20+15 dB)
with 20+18 dB attenuation axial mode tickler still too much - heats ion so that contrast is reduced at resonance - this maybe indicates cooling works less well now?
RID15823 is axial mode with 20+21 dB attenuation - now ion stays cold enough for photon-correlation also at resonance
P_422_A = 0.82 - 0.21 uW = 0.61 uW
P_422_B = 1.18 - 0.61 uW = 0.57 uW
P_422_C = 0.78 - 0.15 uW = 0.63 uW
P_ASE = 8.8 mW
RID15802 is A-beam detuning scan, 1000ms gate.
RID15803 is A-beam bright/dark counts: 5729 / 169 in 1000ms
RID15805 is B-beam detunings can, 1000ms gate
RID15806 B-beam counts 5084 / 217 in 1000ms
RID15808 is C-beam detuning scan, 1000ms gate
RID15809 is C-beam counts 3717 / 160 in 1000ms
Try to optimize ASE beam (already at optimum).
Adjust 422 AOM. Cooling laser stays now locked during detuning scan.
ID15784 is peak-search with B-beam, 11ms probe, "2021-12-10-search5"
1 kHz span, 7 Hz step, 143 pts/peak, Fourier limit 36 Hz AOM
probe-times close to 1/50Hz = 20ms might be problematic - since we have strong 50Hz and 100Hz sidebands?
Try clock-spectrum with B-beam?
RID15782 is rabiflop on +C1 with B-beam
tau_pi 11.8 ms - so much less 674nm power in this beam?
uMotion(200k photons): A 0.7%, B 0.8%, C 0.5%
Optimal voltages (Common mode, diff): [ 0.22433074 33.18958791 54.7442894 ]
RID15773 is micromotion optimization in 4x4x4 grid with 100k photons/histogram.
12:49 progress 45/64
inv( Cfit ) = % ID15773 on 2021-12-10
-0.0022 2.3757 1.8666
0.2333 -1.4629 0.9545
0.2048 0.9681 -2.5425
inv( Cfit ) = % 2021-11-30 result
-0.0046 2.4796 1.9667
0.2430 -1.5787 0.7812
0.2254 1.1646 -3.1975
% compare to 2021-02-17 notes
inv(-1*Cfit) = % sign is arbitrary..
-0.0004 2.9735 2.3561
0.4009 -1.5572 -0.1910
0.3815 1.1746 -4.2519
ID15771 is rabiflop on +C1 center
ID15770 "2021-12-10-search2" is peak-search, A-beam, 3000us probe/OD1-filter, with updated line-center prediction and old abs(B) estimate
1kHz span, 25 Hz step, 41pts/peak
+C1_Center = 75.8639314334291, FWHM 286 Hz (Fourier limit is 133 Hz - but Gaussian fit is no good for Rabi-shape with sidebands)
-C1_Center = 75.8525797859889
C1_Linecenter = 75.858255609709, C1_B = 2.022745445510969
Initial line-center prediction now 75 858 270 Hz - similar to yesterday
next 8.4 MHz, next 9.0 MHz, finally find right beat-note at Piezo-voltage of 42.8 V (old value was around 57V). Roughly 1V change for every TEM00 mode
Clock-laser on. First TEM00 mode found at beat-note 7.7 MHz !? many modes off from the right one.
C might need alignment? not much gained by trying to align C.. (needs focusing?)
P422_C ca 0.85-0.17 = 0.68 uW
P422_B ca 0.6 uW
P422_A ca 0.6 uW
P_ASE ca 8.2 mW
uMotion A 1.1% in 36s, B 1.3% in 36s, 5.3% in 51s
(plotting now fixed so beam is identifid correctly in figure)
A counts 64.1/1.4
B counts 63.2 /2.6
C counts 38.3/1.9
11:18local stop collision monitor
09:15local ion still stored, collision monitor running
15:00Z start collision monitor, A-beam, no ASE-filter, no clearout, 55/1 counts in 10ms. Db-tag "2021-12-09-run1"
A counts 55.5 / 1.1 in 10ms
looks like reasonable results - PMT collection efficiency seems a bit low? (or uMotion causes lower counts?)
uMotion A 2.1%, B 2.7%, C 5.4%
(check plotting code, should be C-beam for last measurement??)
P422_A ca 0.4 uW (OD1 filter on laser-breadboard filter-wheel)
P422_A ca 4 uW
then one disappears?
after various adjustments, Rb-relock and tests, see two ions
14:04 stop dispenser - see some raised PMT counts but no image on camera. counts disappear with ASE-beam, so fluorescence from Sr+
13:56Z load again 1.7A
re check 461-dispenser, now set to 195.78mA
13:40Z Try to load with 1.7A. No ion after 10 min, switch off dispenser current.
lowering trap RF voltage seems to move ions further apart - throw them away..
two ions appeared 1-2 minutes after shutting down dispenser??
no ion in 10min..
12:30Z Try to load with 1.7A
no ion in 10 min. ASE beam direction could be optimized for filter now..
12:14Z load with 1.7A
422-detuning to 185 MHz
empty trap by switching RF off/on
P461 now ca 190 uW in A-beam
tune IR fiber-alignment, 461nm fiber-alignment, and doubler temperature set-point
check 461nm against dispenser: 194.9mA - 4.642 kOhm (new cables, supplies in coolinglaser-rack). Connector for DC to dispenser/vacuum-cross may need re-soldering
Beat-note now 15 546 560 Hz, and Zeeman AOM line-center prediction 75 858 246 Hz - down 2 kHz from ca 75 860 100 with cavity doors open yesterday. (cavity frequency has shifted up)
Total heater power now 28W after closing cavity doors
Inner shield T-sensor shows ca +5 mK warmer temperature ca 20 hours after closing cavity doors. Will probably settle for 1-2 weeks...
Try also 422-shutter on laser-breadboard - to move it far away from ion and comb, since it seems to cause noise in the beat-note measurement?
No ion seen. Install ND-filters for 422nm - but this requires 461nm laser for fiber-alignment - and 461-laser cables are still ongoing work..
ca09:50 local increase P422 to max, all beams on, remove ASE-filter
ion seems lost/dark?
09:18: 422 lock seems to be ok. All beams on.
09:16 local : Ion seems to have gone dark. Stop collision monitor.
17:20 leave lab
17:15local Close ""dog-house""
start of run might have bad data - throw away first 30 minutes?
15:09:50 local start collision monitor, A-beam, with ASE-filter, no clearout, ASE-filter optimized to give ~same amount of counts as without filter.
counts in 10ms 62.6 / 0.5
Now A-beam P422 ca 0.3 uW
reduce P422 to get ca 500 counts per 100ms
422: A: 1.3 uW
1092: 8.6 mW, without filter
Approx 1840 counts per 100 ms (no filter)
With filter up to 1000-1200 counts, dropping to 800 with filter. 1092 power: 6.3 mW.
with filter, ca 1500 counts in 100ms after adjustments - 6.4 mW power after filter
up to 1800 counts / 100ms after adjusting ASE-beam direction
New graphics card in use, bios updated and bios PCIe settings tweaked. Seems to have solved the problem??
Need to shut down computer and replace graphics card...
Computer graphics card red-screen-of-death. 2021-12-08-dmesg.txt
RID15663 is peak search with 3ms probe and OD1 filter. span reduced to 1 kHz, step is 25 Hz, 41 pts/peak. db-tag "2021-12-08-search4"
+C1_Center = 75.86577708041591, FWHM 152 Hz (Fourier limit 133 Hz AOM)
-C1_Center = 75.85440804573633, FWHM 145 Hz
C1_Linecenter = 75.86009256307612, C1_B = 2.0258436706298917 uT
Add OD1 filter for probe
RID15661 is same rabiflop on +C1, but NO 20ms synchronization
duration 251s, half of 20ms-synced program
RID15659 is rabiflop on +C1, synchronized to 20ms. probe times out to 10ms
measurement takes 500s - longer than non-sycnhronized
RID15657 is peak search without ASE-filter "2021-12-08-search3". A-beam 1200us probe
peak-shape looks better now!? ASE-filter might reduce cooling too much??
comb-data predicts line-center 75 860 100 Hz
+C1_Center = 75.8657749546092, FWHM 300 Hz (Fourier limit 333 Hz)
-C1_Center = 75.85440074345941, FWHM 337 Hz
C1_Linecenter = 75.8600878490343, C1_B = 2.02676606375368 uT
+C2_Center = 75.87142020597837, FWHM 297 Hz
-C2_Center = 75.84873606226346, FWHM 323 Hz
C2_Linecenter = 75.86007813
+C4_Center = 75.8888664606585, FWHM 304 Hz (is Gaussian fit to carrier-peak or one side-lobe??)
-C4 fit fails (too long probe-time gives double-peak?)
remove filter, now 192.5 / 1.2 in 10 ms - threshold 15
A fluorescence is 70.7 / 1.4 in 10ms - with ASE filter
(abort scan after C1 peaks)
RID15654 peak search with 1200us probe
scan is wide, so C2 also visible in scan centered around +C1 and -C1
ID15652: Rabi flop with 20ms synchronization.
new programs for spectrum-scan and rabiflop with 20ms synchronization
7:22Z: Stop collision monitor.
17:25 leave lab
17:22local start collision monitor, A-beam, with ASE-filter, no clearout
tune P422 slightly up, to get counts 47/1 in 10ms. threshold is 6
add ASE filter for collision monitor.
TF made improvements to comb beat-note setup: telescope to focus beam correctly, periscope to flip polarization for grating.
Now up to 53 dB SNR for beat-note (initially on 2021-11-01 got ca 47 dB)
Try to improve uMotion uncertainty estimates, and plot a figure:
(see 2019-12-10 for simulation on how many photons needed to reach given unc.)
Incerease number of photons, to get lower uncerainty for contrast
50k photons in 5s
400k photons, duration 35s
800k photons, duration 74s
more photons causes the last bin to have anomalous count?? RID15586 and 15585
Compare to fits in matlab:
ID num_photons Py_contrast(unc) Matlab_contrast(unc)
15587 50k 2.43 (0.65) 2.43 (0.67)
15581 100k 2.13 (0.46) 2.13 (0.47)
15582 200k 2.60 (0.43) 2.60 (0.44)
15583 400k 2.20 (0.22) 2.20 (0.23)
15584 800k 2.40 (0.18) 2.40 (0.18)
Comparison between simulation from 2019-12 and experiments 2021-12 show rough agreement
Variables form ARTIQ program:
clock_probe = 1000 [us]
dt_cool = 30 [ms]
dt_repumper = 35 [ms]
dt_clearout = 5 [ms]
dt_detect = 10 [ms]
dt_detect2 = 10 [ms]
start_delay = 1 [ms]
clock_decay = 1 [ms]
cool_shutter = 5 [ms]
end_delay = 7 [ms]
RID15565 is peak-search with 1200us probe, A-beam, ASE-filter removed
step now 60Hz, span 3 kHz, 51 pts/peak (all numbers in AOM units)
peak-shape looks better now
+C1_Center = 75.86598858026497, FWHM 346 Hz (Fourier limit 333 Hz AOM)
-C1_Center = 75.85463431058417, FWHM 353
C1_Linecenter = 75.86031144542457, C1_B = 2.023212701495724
+C2_Center = 75.8716461460222, FWHM 365 Hz
-C2_Center = 75.84900696569804, fwhm 308 Hz
+C4_Center = 75.88817853075848, FWHM 281 Hz
-C4_Center = 75.83154739674906, FWHM 288 Hz
Centers, FWHMs = [75.86598858 75.85463431 75.87164615 75.84900697 75.88817853 75.8315474 ] [0.00034556 0.00035261 0.00036455 0.00030762 0.00028072 0.00028792]
RID15563 is rabiflop to 9 ms probe time - to see more oscillations?
RID15561 is repeated rabiflop on -C1, no ASE filter
now maybe nicer oscillation with max at 1200us?
95.4 / 1.07 in 10 ms -> threshold 10
remove the ASE-filter, this increases A-counts from 550-600/100ms up to 900-1000/100ms
RID15555 is rabiflop on -C1
rises to 40% at 600us - then noisy counts around 30-40?
RID15554 is peak-search again, 800us probe
no real improvement..
Turn off Keysight DMM (on optical table close to 461nm supplies)
RID15553 is peak-search again
+C1_Center = 75.8660506996127, FWHM 683
-C1_Center = 75.85453957585881. FWHM 1268
C1_Linecenter = 75.86029513773576, C1_B = 2.05116246505545 uT
doesn't look any better.
re-lock fiber noise cancelling servo, maybe it was not locked?
RID15552 is peak-search with reduced A/C, A-beam 800us probe, ASE-filter
3kHz span per peak, 100Hz step, 31 pts/peak
No major improvement..
12:00 local A/C now at 25Hz
Try with lower A/C setting of 25Hz?
RID1550 is peak-search with 800us probe
+C1 75.866030, FWHM 1172 (double Fourier limit!?)
-C1 75.854535, FWHM 896
C1_center 75 860 282, C1_B 2.048 uT
not so great spectrum..(line-centers etc. all wrong in this figure - script could be improved..)
-C1 75.85415 (single high count? false?)
From these line-center 75.860 (but ion dark-moment creates symmetric high-counts, since we scan left/right outwards from scan-center!)
abs(B) = 2.085 uT (similar to Friday)
RID15547 is 600us clock-spectrum with A-beam, no OD-filter. 100 kHz AOM span, 130 Hz step, db-tag "2021-12-07-run1"
Zeeman AOM line-center prediction is 75.8603 MHz - similar to previous values
BOA current 400 mA
P674 power-servo on (stabilizes power on ion-breadboard, not intra-cavity power)
Typical drift 10 mHz/s for 674nm is 0.8 kHz/day - measured at comb IR 0.4 kHz/day
ECDL piezo voltage now 56.6V - this is the correct mode at 15.548 MHz !? (needed to move 5 pcs TEM00-modes from bootup of laser)
next 16.201 MHz
next 16.854 MHz
next mode 17.507 MHz
increase piezo, next mode at 18.159 MHz
Initial TEM00 mode at beat-note 18.790 MHz - increase piezo-voltage?
10:20 comb on
uMotion A 0.9%, B 2.6%, C 3.8%
A-counts 46.3 / 0.8, B-counts 57.3 / 0.7, C-counts 43.7 / 0.8
re-appears in 1-2 minutes with all beams ABC on
ion dark again!?
10:13 local stop collision monitor
For future remote operations: ionization-laser shutter is TTL9
coolinglaser shutter is TTL8
405nm laser powersupply is TTL4
Note uptime on kavi-machine >7 years ;)
09:33:28 up 2594 days, 22:12, 7 users, load average: 0.07, 0.09, 0.13
New IP for ionclock-computer, ends in 45.15 - replace IP with name 'strontium.ad.vtt.fi' on kavi-machine - this should fix things as long as DNS works.
09:26local ion still stored. camera-image maybe a bit weak?
18:16local start colliosion monitor. A-beam 50/1 counts (threshold still 6?), ASE-filter, no clearout. db-tag "2021-12-06-run1"
uMotion A 0.5%, B 1.0%, C 3.2%
A-beam fluorsecence 49.0/0.8 in 10ms
18:12local ion has re-appeared !? all three beams ABC were on since ca 11local. no clearout
ion loaded monday 2021-11-29, so ca 1 week stored
run detuning-scan program (this helped to restore fluoresescence previously?!) - no ion
422nm laser Rb-lock monitor not running - would help to diagnose problem
try 405nm (but no sure about shutter state when operating remotely..) - no ion
clearout on - no ion
1051local all beams ABC on - no ion on camera image
RID15523 is 10ms counts over ca 600s(?) - highest count is 5, so not detected as a bright state.
1046local stop collision monitor program
camera image shows no ion
counts in 10ms: 2, 7, 1, 9, 2, 7, 3, 8, 0, (artiq program outputs count only when state changes)
threshold set from bright=60, dark=1. (dataset variable thr=6 is in use?)
0845local ion stuck in perpetua blink-mode?
19:45local leave lab.
19:38local start collision monitor, A-beam, ASE-filter, no clearout
uMotion A 0.7%, B 1.5%, C 2.3%
47.9/0.8 in 10ms
Incerease P422 to get same counts as before, ca 450-500/100ms
adding filter in front of ASE-beam reduces counts to ca 340/100ms.
A-beam fluorescence level 44.2/0.4 in 10ms
19:30local stop collision monitor
ion still stored.
15:04:27 UTC: Start collision monitor without clearout
mis-align fiber on 422-breadboard to get only ca 60 counts. measurement gives 62.1/0.7 counts in 10ms
A-beam counts now ca 100/10ms
ASE-filter for collision-monitor tests Edmunds 87813 Center wavelength 1100.00 +/- 3 nm, FWHM 25 +/- 3 nm.
Activate temperature-loop for toptica - this should keep Fine-1 integrator voltage limited by slowly adjusting laser temperature
RID15505 spectrum-scan, A-beam, OD1/3000us probe
clock-laser unlocked during scan...
RID15501 rabiflop on +C1
Try OD1-filter, this should triple tau_pi.
With no OD-filter, A beam P674 = 0.76-0.18 = 0.58 uW (power-servo on, PD-voltage 3.54 V)
uMotion was not checked after working with the shields.
uMotion now: A 0.6%, B 0.3%, C 1.5% - no need to optimize(?)
RID15496 is peak search with 1200 us probe
+C1 Center = 75.86629295865397 (FWHM 334 Hz AOM - at Fourier limit)
-C1 Center = 75.85498667431237
C1_Linecenter = 75.86063981648317
C1 B estimate = -2.0146622133999115
+C2 Center = 75.87196171348045 (FWHM 343 Hz)
-C2 double-peak? not good fit..
Rabiflop suggests tau_pi 1200 us for +C1?
+C1 h = 24.5, Center = 75.86631950607462
C1_Linecenter = 75.86064077073416
C1 B estimate = -2.0237830864052193
RID15488 is 600us clock-spectrum with A-beam, 2021-12-03-run3
(overrides in wrong positions...)
Coil currents now set from artiq [-8.6056, -1.0, 7.0059] mA - this should give ca 2.26 uT?
Test that I2-driver now drives the correct current through 2-coil - OK!
this removes the short to chamber-ground from 2-coil camera side
camera-side window coil was maybe shorted to inner shield. add kapton-tape to wire, where it goes through U-shaped hole in camera-side inner shield
ID15470 is a wider scan, to capture C4, C5 also. "2021-12-03-run2"
600us clock-spectrum, now hopefully without B-field issues?
ID15469 "2021-12-03-run1", A-beam
Connect 1-coil and 3-coil to driver as usual, terminate 2-coil with 50R terminator. Now expect abs(B) closer to 2 uT as before?
coil-driver box seems OK when driving external loads
however any driver-output when driving the 2-coil shows strange currents?? (correct results when driving 1-coil or 3-coil)
2-coil somehow broken, or external current/voltage leaks into it??
driving only 2-back-coil (CaF side?) seems OK
driving only 2-front-coil (FS side?) - very strange results!! this coil seems broken. Must be short to ground or similar problem??
1-coil and 2-coil are floating wrt. chamber ground.
broken 2-coil seems shorted to chamber ground
Measure coil currents with DMM in line with the cabling. Units in mA.
Artiq set-points: I1: -8.6056 I2: -1.0049 I3: 7.0059
I1: -8.601 mA DC, 66 nA std, 330 nA peak-to-peak, 60 samples.
I2: +20.1 mA DC (!!!), 45 uA std, 216 uA peak-to-peak, 60 samples. Broken???
I3: 7.013 mA DC, 1 uA std, 6 uA peak-to-peak, 60 samples. OKish, but with outliers??
Straight from source to DMM:
I2: now gives programmed value -1.003 mA (1.4uA std, span 5.8uA)
test coil resistances (Keysight DMM 2-wire)
coil1: 3.250(30) Ohm
coil2: 1.310(20) Ohm - some connection broken/shorted?
coil3: 3.170(15) Ohm
coil2: front 0.19 Ohm, back 0.19 Ohm (maybe contact problem with coated wire)
09:05local stop DMM B-field logging
08:55local Ion still stored, but clock-laser out-of-lock at some point??
laser-unlock is jump in Fine-in-1 voltage (FALC slowest integrator output) - may try adjusting laser temperature to never reach -2V with integrator?
Leave clock-spectrum peak-search, many queued measurements - to maybe see better spectrum at night?
RID15442 to RID15460
RID15442 start_time 2021-12-02 17:53 local
RID15443 run_time 2021-12-02 19:13 local
RID15444 run_time 20:32 local
RID15445 run_time 21:51 local
RID15446 run_time 23:12 local
RID15447 run_time 00:31 local
RID15448 run_time 01:50
RID15449 run_time 03:09local (clock-laser looses lock during +C4 scan)
RID15450 no quantum jumps
Install the B-field meter vertically, outside breadboard on ASE side. Collect ca 1 point per second using DMM.
Compare to old data (summer 2020?)
+C1: 31.5 75.882759 913 (height, center_MHz, FWHM_Hz)
-C1: 30.7 75.836036 956
C1_center 75 859 397, C1_B = 8.3256 uT
+C2: 20.5 75.905963 1873 (double FWHM?? B-noise??)
-C2: 21.8 75.812753 1784
+C4: 17.8 75.976141 4348 (4.8-fold FWHM compared to C1!!)
-C4: 20.3 75.742525 3855
Fourier-limit with 600us probe is 667 Hz AOM
RID15441 peak-search program (program could be improved by automatic scan-range, frequency-step parameters, based on probe length)
5 kHz AOM span with 100 Hz AOM step, 51 points/peak
peak-position prediction is [75.88278013 75.83589187 75.90608223 75.81258977 75.97627258 75.74239942] = [+C1 -C1 +C2 -C2 +C4 -C4]
This gives line-center 75 859 336 Hz, and abs(B) = 8.355 uT (compare to prediction from comb: 75 859 312)
Fit to +C1 peak [34.9 75.882809 959]
Line-center prediction is 75.859 MHz - so expect to find +C1 at 75.859 + (75.859-75.836) = 75.882
Fit to -C1 peak [height, center MHz, FWHM Hz] [28.7 75.835922 1005]
11:45 UTC: AC to 44.40 Hz. Back to old graphics card.
ID15430 -C1 spectrum, 600us probe "2021-12-02-run8"
Gaussian fit FWHM 930 Hz (AOM?). 0.8/600us = 1333 Hz optical
ID15428 rabiflop on -C1: 75.83547
RID15427 spectrum with 600us probe - looks better now!? (database tag "2021-12-02-run7")
Rabi flop ID 15425 on -C1, tau_pi ~600 us, max probability ~30%.
Computer crashed again, with pci express errors!
Graphics card crashed the computer, switched to a ""new"" card
-C1 with AC at 20 and rectifier monitor box disconnected! :
-C1 with AC at 20 Hz, 300us probe.
reduced AC from ca 10:38local
try reducing AC VFD from high 44-45Hz setting down to 20 Hz?
ID15416 rabiflop, but counts very low and no oscillation..
ID15415 measure -C1 with 300us probe - peak center ca 75.83568 MHz
P674_A = 0.67-0.15 = 0.52 uW in A-beam before chamber, power-servo ON
P674 = 52uW before ZeemanAOM-cube, power-servo ON
P674_A ca 1.1 uW in A-beam before chamber
P674 = 108 uW before ZeemanAOM-cube (no power-servo)
seculars slightly higher than before (1.089, 1.129, 1.722)
previous (x, y, z) measured (1.088 ,1.127, 1.7175)
comb computer was running, but comb-software had crashed!?
Comb beat-note measurement had failed, because reading rep-rate DDS value from comb failed. Menlo computer somehow unstable??
uMotion A 0.4%, B 1.5%, C 2.2%
09:23local ion still stored
Leave wide spectrum scan. A-beam, 300us probe. no 674-power servo (to get max power).
500us probe spectrum:
Can't find a reasonable clock-spectrum, go back to the old coil currents [-8.6056, -1.0049, 7.0059]
change 422-shutter to the one used for 461/405 beams - this one is much more silent
422-shutter seems to buzz with a strange sound (not present before??) during clock-cycle run
ID15392: Rabi flop at f = 75.873408 MHz. Detection probability is ~40% at long probe times, but without oscillation:
power after AOM almost doubled after removing 2-channel DDS connection and splitter (used for bichromatic probe tests)
ca 0.26 uW from A-objective towards ion
ca 3.1 uW before A-fiber
ca 10uW after double-pass AOM
before Zeeman-AOM cube-splitter ca 54 uW (note beam-splitter BSN04 10:90 R:T for power-monitor has quite high losses(
Maybe some problem with Zeeman-AOM causes strange clock spectra. check also 674nm power.
run2 try to increase 674nm power slightly, and longer probe 1000us
1ms should have 800 Hz FWHM, 400 Hz AOM. Step is 80 Hz AOM, so 5 pts in FWHM.
run1 spectrum with 500us doesn't show much..
Zeeman AOM line-center prediction now 75.8616 (yesterday rough measurement was 75.8617 MHz - so OK!?)
Set all currents to zero.
Last run of coil-current program has settings (-8.6056, -1.0049, 7.0059) mA.
use AB2021-11 numbers for AHM3 shift. Maser frequency is high by +5e-13, so we observe optical frequencies too low by ca 5e-13*SRS = 200 Hz
(previously AHM3 was too low, and we observed optical frequencies shifted up)
increase piezo-voltage, now beat-note 15.550 MHz - right mode!?
increase piezo-voltage, now beat-note 16.202
turn on clock-laser and comb. initial beat-note 16.855 MHz (2 modes too high beat-note?)
z-secular slightly higher than before
repeatability 1 719 596 - 1 719 594 = 2 Hz !? (experiments run consecutively without delay, duration maybe 2min/experiment)
uMotion A 0.6%, B 0.2%, C 1.4%
09:12local ion is still stored. stop fluorescence monitor
17:35local leave lab
Leave fluorescence monitor, A-beam
no issues with artiq-boxes found
rabi-flop doesn't give sensible results, spectrum with shorter probe doesn't look better?? check thath ARTIQ-boxes are properly locked to 100 MHz!?
from peaks abs(B) = 8.23 uT
measured line-center 75.8617 MHz. Prediction from comb-data is 75.8626 MHz (should include latest AHM3 model?)
+C1 center 75.8848 MHz
-C1 center 75.8386 MHz
double-peak indicates too long probe time?
A-beam 600us spectrum
674nm in A-beam ca 0.58-0.32 = 0.26 uW
this would give abs(B) = 8.4 uT ?
4 jumps at 75.8842 MHz (+C1 ?)
5 jumps at 75.837 MHz (-C1 ?)
ID15348 try to find clock-spectrum with A-beam. 200us probe
A-beam fluorescence quite low: 55/2 counts in 10ms
Prediction for Zeeman AOM line-center is 75.8626 MHz
Repeat Comb-number measurement: (see 2021-10-28)
reprate_DDS reprate_comb beatnote1 comb_number
19 999 584 249 999 896 -15 549 941 n1
20 000 713 250 000 178.25 -16 627 505 n1-1 (m=-1)
n1 = (m*frep2 + f02 + fbeat2 - f01 - fbeat1) / (frep1-frep2)
n1 = round( 889557.988485 ) = 889558 (same result as 2021-10-28)
f674 - SRS = 2* ( f01+n1*frep1 + fbeat1 ) - SRS = -330223432.5
reduce piezo-voltage, next TEM00 at at 15.550 MHz - this is the correct one (?), measure comb-number here
reduce piezo-voltage, next TEM00 at 14.897 MHz
reduce piezo-voltage, next TEM00 at 14.245
with rep-rate DDS at 19 999 584 Hz previously correct TEM00 mode was at 15.476 MHz. We are now at 14.231, so 2 modes off?
beatnote moves up when rep-rate adjusted up, so sign is negative
increased piezo-voltage gives TEM00 mode at beat-note of 13 592 465 Hz (spacing -652 489 Hz)
previously correct mode at 15.9947 MHz, and TEM00-spacing at comb 0.6525 MHz - so probably need to move two TEM00 modes higher?
Clock laser on, comb on.
Initial TEM00-mode beat-note at 14 244 954 Hz (not the correct TEM00 mode, most probably)
Maybe good to re-do the comb mode-number measurement
Tuned parameters for fitting of cooling-lineshape
# [start,lower,upper] values for fit
Rc0 = [0.9,0.5,1.65];
Rr2pb0 = [0.2,0.1,1.0]; # old [0.66,0.5,1.2]
scale0 = [450,300,600];
z-secular similar to y-day
images rotated to show zero-plane. green/red show sign of contrast.
For notes on the linear uMotion model see for example 2021-02-17
Recently changed to alu-mount and Polaris-mount for B beam, so it may have moved slightly (within mu-metal shield holes)
Changing voltages by column of inv(Cfit) should change only one contrast by 100%
% Micromotion contrast model
% S = C (V - V0)
% S has [A, B, C] contrasts
% V and V0 have voltages [V12dif, V34cm, V34dif]
% V12cm is constant 0 V in recent experiments
inv(Cfit) = % 2021-11-30 result
-0.0046 2.4796 1.9667
0.2430 -1.5787 0.7812
0.2254 1.1646 -3.1975
% compare to 2021-02-17 notes
inv(-1*Cfit) = % sign is arbitrary..
-0.0004 2.9735 2.3561
0.4009 -1.5572 -0.1910
0.3815 1.1746 -4.2519
Now get uMotion A 0.7%, B 1%, C 2% ( 100k photons per histogram )
Optimal voltages (Common mode, diff): [ 0.22497105 33.21311168 54.61181913]
ID15337 uMotion optimization 5x5x5 grid, 125pts, 50k photons per histogram
uMotion: A 4.4%, B 5.4%, C 3.9%
ion is still stored
09:30local stop fluorescence monitor
Strong correlation between temperature and B-beam counts. Polarization change from linear to partially circular (in which component?)
NOTE: different sign of correlation between 2021-11-24 and 2021-11-29 measurements.
First measurement had waveplate before B-fiber input (not sure about waveplate orientation)
In the later measurement waveplate was after 422nm-launch on ion breadboard, optimized to zero the lost power in the A-beam polarizing cube
Leave 1000ms fluorescence monitor, B-beam
Try to improve lorentzian fit by setting limits for parameters to be optimized:
lorentzian(f, C, A, f0, g) = C + ( A / ( ( f-f0 )**2 + g ) )
# C = constant background
# A amplitude
# f0 center freuency
# FWHM = 2*sqrt(g)
# C A f0 g
fwhm0 = 500e-6 # guess for FWHM in MHz
p0 = [ 0, np.max(fit_contrast), freq[np.argmax(fit_contrast)], pow(fwhm0/2.0, 2) ]
p0_lower = [0, 0, freq, pow(1.0e-6/2.0, 2)] # lower bound for fit
p0_upper = [0.2, 1.0, freq[-1], pow(0.01/2.0, 2)]
Example run (for z mode):
p0 = [0, 0.49300631348921775, 1.7173684210526317, 6.25e-08] # starting point for fit-parameters
p0_lower = [0, 0, 1.715, 2.5e-13] # lower bound
p0_upper = [0.2, 1.0, 1.72, 2.5e-05] # upper bound
Detuning scans (0.1 s gate): A: ID15320, B: ID15318, C: ID15316
Background counts (0.1 s gate): A: 26.73, B: 24.66, C: 25.0
uMotion: A: 4.7%, B: 6.4%, C: 2.8%
Reduce beam power: A: 0.7 uW, B: 0.9 uW, C: 0.9 uW
lambda/2 -waveplate moved to in front of 422 delivery collimator, A-beam PBS waste beam power minimized. B-beam extinction ratio now slightly worse, ~18 dB
uMotion A 3.7%, B 2.3%, C 2.6%
C 250couts/10ms - power 7.2uW
B also 300 counts/10ms - power 7.2uW
A-beam ca 300counts/10ms - Power 5.6uW
camera image of ion looks blurry. micromotion?
13:19:10 ion! (4 min 10s)
405nm ca 4 mW
13:15z load with 1.7A
check 461nm dispenser: T = 4.642 kOhm, I = 196.60 mA (no major change from before)
13:11 no ion loaded
12:56z load with 1.6A.
empty trap by RF off/on
Measured at B-beam output (through trap chamber) 20.6 dB
Polarization extinction ratio on ion-breadboard 25.6 dB (after fiber from laser). Only small changes when heating fiber by hand.
try 405nm laser - no ion in 2-3 minutes
try runnign detuning scan? no ion
09:35local coolinglaser on. initially no ion seen..
coolinglaser oscilloscope had freezed - need to improve logging program
18:10local leave lab.
leave ion trapped without cooling
shut down coolinglaser, clocklaser, comb, PMT
wider spectrum with 200us probe. A-beam
maybe A-beam gives clearer results?
peak-position prediction is [75.88579188 75.84100812 75.9080481 75.8187519 75.97508808 75.75171192]
-C1 at 75.841 and +C1 at 75.8858 would indicate line-center 75.863 and abs(B)=7.98 uT (?)
reduce cooling-laser power, so B PMT counts ca 120/10ms
ID15270 clock spectrum with 200us probe
NOTE: beat-note on spectrum analyzer shows visible jumps when we run the clock-cycle!? From coolinglaser mechanical shutter?? Shakes comb, or maybe 1348nm fiber?
rabiflop suggests tau_pi = 200us
initial clock spectrum, predicted line-center 75.865 MHz. 1000 us probe pulse
-C1 (maybe?) at ca 75.841 MHz / +C1 maybe 75.8858
Coolinglaser power was increased again - maybe 5-7uW in beams now
coil currents (I1, I2, I3) = (-8.6054, -1.0047, +7.005) mA (settings from 2021-02-25?) - abs(B) estimate was 2.26 uT (?)
alu-mount installed, some adjustment of B-objective focus restores roughly same level of fluorescence as before
B-beam objective: 51 mm from mid-shield to lower end of the objective case. Height from breadboard to objective lower edge 158mm
install alu-mount for B-beam
C fluorescence 8520 / 118
B fluorescence 5666 / 76 (heating fiber gives ca 1100/100ms = 11000/1000ms)
A fluorescence 13522 / 87
ID15224 wider span scan, with larger frequency-step. B-beam(?)
we might have the wrong comb-number in the calculations? check what rep-rate was used previously?
automated line-center prediction is 75.864 MHz for Zeeman AOM - try to search there?
fbeat1 = 15 994 723 Hz, with rep-rate of 249 999 896.5 Hz
11:25local start clock-laser. found TEM00 mode 1348.052-053 nm is the correct one?
uMotion A 1.7%, B 2.7%, C 1.9%
09:47local stop A-beam fluorescence monitor
17:36local leave lab.
Leave fluoresecence monitor with A-beam. A polarization should be OK?
back to V34cm = 33.188V
ASE power ca 8.3 mW
(x, y, z) = (1.089855, 1.129059, ) MHz
ID15205 z-secular, same range as before? (now using C-beam)
now uMotion A 21.9%, B 9.5%, C 12.2%
Add +1V to V34cm: 34.188 V
uMotion A 1.8%, B 2.4%, C 0.6%
These at uMotion-optimum set earlier, V12cm = 0.0, [ 0.22942917 33.18804714 54.8568037 ]
results (x, y, z) = (1.089309, 1.128581, 1.721865) MHz. RF rectifier DMM reads 58.6 mV
(fit fails.. python numpy lorentz fit worse than matlab??)
(narrower range, peak centered)
ID15195 1.71-1.73 MHz, 20+15 dB
ID15194 1.12-1.14 MHz, 20+4 dB
ID15193 1.08-1.10 MHz, 20+4 dB
ID15192 20+4 dB lower mode with B-beam. 1.07-1.09 MHz 50pts, 20kphotons/pt
(peak not very well centered in scan)
higher one is ca 1.12775
0+12 dB coarse tickler reveals one mode at ca 1.0883 MHz
notes from 2021-05-19 show coarse tickler with 0+11 dB attenuation
Secular prediction is (x, y, z) = (1.0864, 1.1260, 1.7193 ) MHz
China camera images for A, B, C beams (stack of 0.5s exposures, 30-60s in total)
New Standa table installed in the 'mini-lab'.
Check time-stamps of these measurements and correlate with RF-rectifier measurements?
Secular frequency slightly lower than yesterday:
uMotion (100k photons): A 0.7%, B 0.5%, C 0.6%
Optimal voltages (Common mode, diff): [ 0.22942917 33.18804714 54.8568037 ] - note when these are auto-applied by the optimizer, manual GUI-program retains old (wrong) values!
11:09 progress 50/125
10:55 local ID15170 run uMotion minimization, 5x5x5 grid, 125 pts. 50 k photons per histogram (maybe need more to accurately find zero?)
uMotion now A 1%, B 2.3%, C 2.8%
10:41 local: stop PMT counter program that was running over night.
fits to data from yesterday:
before installing shields, a photo with Sigma 105mm macro lens (no extension tubes)
Compare to 2019-06-18 photo:
Leave fluorescence monitoring over night
plot updates once per minute
data is in textfile 2021-11-24-counts.txt
C-beam power ca 1.7 uW
ID15164 C-beam detuning scan
C-beam Bright / Dark: 14505 / 150
B-beam power ca 2.4-0.9 = 1.5 uW
ID15161 B-beam detuning scan
B-beam Bright / Dark: 131748 / 83.3
A-beam power ca 1.5 uW
ID15157 is detuning scan, A-beam, reduced 422nm power
ID15158 Bright/Dark A-beam 19558 / 117.7
re-locking to Rb-line recovers fluorescence
lost ion while trying to reduce coolinglaser power??
Now about 1uW in B-beam
Now get ca 259 uW onto ion breadboard from coolinglaser - this is maybe 10x too much - misalign laser-ion fiber to get 30uW
ID15153 B-beam, 15 dB DDS attenuation - now fit works
ID15152 repeat B-beam, less DDS attenuation 18dB to get higher correlation peak - fit still fails
ID15151 repeat B-beam, narrower range
ID15149 B-beam shows much wider resonance (but fit fails..)
ID15148 A-beam has worst cooling for axial mode, so narrowest resonance
coarse sweep shows resonance at 1.727 MHz
Try old secular sweep settings, they might still work? 1.786 - 1.79 MHz, DDS att 22dB - no peak
P_repumper = 8.3 mW
C-beam detuning scan is ID15139, C beam P422 = 7.5 uW
Bright/Dark (ID15140) = 22234.7 / 512 (1000 ms gate)
B-beam detuning scan is ID15137, B beam P422 = 7.1 uW
Bright/Dark (ID15142) = 23283.5 / 61.6 (1000 ms gate)
A-beam detuning scan is ID15135. A beam P422 = 4.4 uW
Bright/Dark (ID15144) = 26363 / 128.1 (1000 ms gate)
B-beam fluorescence weakness may be becaus of polarization-effect?
at [0.0, 0.2273, 33.1706, 54.9760] get A 0.2%, B 1.7%, C 0.4%
Optimal voltages (Common mode, diff): [ 0.22733392 33.17064191 54.97603239]
B-beam now gives weakest fluorescence - realign?
ID15122 uMotion-optimizer, 4x4x4 = 64 pts range [[0.22, 0.24], [32.5, 34.5], [54.1, 56.1]].
at [0, 0.23, 33.5, 55.1] get A 5%, B 0.8%, C 1.5%
ID15115 uMotion-optimizer Optimal voltages (Common mode, diff): [ 0.22903766 33.52124375 55.10656201]
Try [0, 0.2451, 31, 55] -> A 28%, B 23%, C 14%
ID15103 NPL uMotion scan 3x3x3 grid (27pts) Optimal voltages (Common mode, diff): [ 0.25156391 31.7416589 57.15082204]
bias voltages from past experiments were 0.0, 0.2451, 34.2701, 50.0363
repeat C 42%
uMotion with inner+mid shields: A 22%, B 52%, C (very weak now? mid-shield C-hole was blocking beam?)
install middle shield
ion is still stored
Install inner shield
Switch to ASI camera. focus-micrometer at 7.0mm. slight focus difference between intensifier/PCO and ASI cameras
uMotion A 20%, B 47%, C 37%
ASE-beam alignment doesn't give more fluorescence - original alignment was probably OK
counts rise to 300/10ms when detuning set to 208 MHz - so we probably have a lot of micromotion (no shields, old vbias voltages)
rough PMT alignment gives 200 counts/10 ms gate. Note PCO image inverted wrt. china-cam (so PMT pinhole is top left on china-cam, bottom right on PCO)
now all beams give fluorescence!
adjust C-beam to get fluroescence.
C-beam gives no signal (but 0.5" post was loose)
no signal at all from A-beam? adjust A-beam and recover fluorescence
fluorescence mostly from B-beam
14:34:30 found ion! turn off dispenser
small A-beam adjustments while dispenser on
14:30:30 try to load again, 1.7A
adjust A-beam alignment using diffraction-pattern of electrodes
14:12:30 load with 1.6A
PCO camera, 400ms exposure, 4V intensifier gain, lab lights off
Try loading ion with intensifier.
When intentionally unlocking the laser from Rb-line (three points towards end of graph), this shows as reduced VPP/AC noise on CH1, CH2 (laser current) goes out of range of the scope (collected numbers 1e38 = NaN?), and higher DC (lower AC) level on photodiode CH3
TPA current 1.8A, seed laser 68.2 mA
New coolinglaser monitoring plots at bottom of comb page
monitor Rb locking-signal via oscilloscope, now at 192.168.10.56
CH1 (error signal) average should stay close to zero, noise (Vpp) around zero shows that lock is working
CH2 (laser current) should stay bounded - will loose lock if this goes too high or too low (what are the limits?)
CH3 (photodiode) should stay at constant level, proportional to total 422nm power(?)
align camera to electrodes
try intensifier + PCO camera?
08:25 no ion, stop loading
08:15Z try to load with 1.6A
Rb-signal after optimizations
Optical feedback made lines move around a lot, and spacings between the three lines to vary - i.e. tuning laser frequency via laser-current was very non-linear
CH1 (yellow): sat.absorption differential locking-signal
CH2 (blue): laser current drive
CH3 (magenta): photodiode signal
Shiner Table1, Line ID "i" is 710 962 401 328 kHz
Sr+ line 710 962 837.3 MHz - so AOM-shift to line-center +435.97 MHz
Further beam-alignment improved Rb-signal a lot!
There seems to be optical feedback from the first surface of the DLI-1 isolator, after 844nm seed laser
mis-aligning isolator removes optical feedback and makes Rb-signal more stable
Rb error-signal still quite unstable and variable
Try lower A/C setting of 20Hz - less air-currents over laser/Rb-cell ?
Seed laser temp 31.5C set-point - change slightly?
check 844nm Linos DLI-1 isolator tuning - in reverse configuration transmitted intensity minimized with tuning-screw
changes in AOM-frequency give changes in beam-direction -> DC-level changes in locking signal
422nm laser tuning..
375um in from electrode-focus, z-translator at ca 7.0 mm
up/down center between lectrodes.
Inner electrodes in focus when Z-translator at 7.5mm
Thread-lock on SM1 tube and camera T-ring interface
check camera alignment? need to remove mid-shield for this
11:36 stop dispenser - no ion loaded?
11:27 cooling-laser out of lock? relock
11:24:15: start to load with 1.6A, power 1.6W initially
Bias voltages: V1 -0.2451, V2 +0.2451, V3 -15.7661, V4 +84.3065
1033nm: 70 uW
1092nm: 10.4 mW
trap RF on at 10Vpp from keysight gen ch2. RF-rectifier shows 58 mV on DMM
405nm adjusted to ca 3 mW in A beam
AOM detuning 185 MHz
422nm power in beams now more than before, 4uW A, 5.5 uW B, 6.5 uW C
power 165 uW in A-beam
461nm check against dispenser 196.68mA and 4.642 kOhm
remove outer B-shield, to install camera objective
C-beam post was loose? why? tighten M6 screw
see also Thorlabs Fabry-Perot Interferometer Tutorial
With R1=inf. and R2= 1.0m get mode-spacing of 92.32 MHz
Tuning R2=1.0247 m gives observed spacing 91.06 MHz
Note FSR/mode-spacing = 5.5, so in principle modes close to TEM00 are:
TEM(N, 0, 0) at 0 MHz
TEM(N, 0, 5) +455.3 MHz
TEM(N+1,0,0) +501.3 MHz
TEM(N, 0, 6) +546.4 MHz
(in practice TEM05 and TEM06 are quite weak or not excited at all, if mode-matching and beam-alignment is OK)
TEM00 - TEM01 - TEM02 - TEM03 mode spacing determined as:
Cavity modes given by
f / f_FSR = [q + (1/pi)(m+n+1)*acos(g1*g2) ]
gi = 1 - L / Ri (R is radius of curvature of mirror i)
L = 0.299012 m (from FSR-measurement 501 304 785 Hz)
For fun, try to measure frequency of non-TEM00 modes of cavity?
comb rep-rate 249 999 896.5 Hz (DDS 19 999 586 Hz)
TEM01 mode (vertical, silghtly tilted to right) beat1= 28 878 149 Hz
lower piezo-voltage, beatnote goes through zero (changes sign) and for TEM00 get beat1= 16 654 261 (- sign in previous measurements!?)
back to same comb-line as TEM01, now even more piezo voltage TEM02-mode (vertical, tilted to right)
beat1 ca 74 MHz, but this isn't measurable with SDR.
adjust rep-rate to get beat1 within range. rep-rate: 249 999 962.5
TEM02 beat1= 15 696 724 Hz
increase piezo to get TEM03 mode, adjust rep-rate to 250 000 012.5
beat1= 16 757 321
TEM03 intensite seems a bit weak, and variable intensity on camera image
adjust rep-rate down, beat-note moves up (so sign is +)
decrease piezo voltage to go back to TEM02
set DDS 19 999 840, so reprate is 249 999 960 Hz
beat1= 17 920 607 Hz
decrease piezo voltage to find TEM01
set DDS 19 999 630, so rep-rate is now 249 999 907.5 Hz
beat1= 19 093 037 Hz
decrease piezo voltage to find TEM01
DDS now 19 999 440 Hz, so rep-rate 249 999 860 Hz
beat1= 15 814 581
(the later measurements should all be against the same comb-line, with variable rep-rate)
sanity-check, back to old rep-rate of DDS = 19 999 586 Hz, or rep rate of 249 999 896.5 Hz
beat-note changes sign as expected (now -sign)
beat1= -16 654 245 (compare to first TEM00 measurement above)
work on cooling laser...
unlcok comb rep-rate and f0 locks
M-nir with flip-mirror to RF-lab
15:17 turn off clock-laser
14:06 lock laser again
up to ca 0.7 V after adjusting two last mirrors
up to 0.56-0.57V by adjusting final mirror before cavity
cavity transmission monitor 0.46-0.47V, with BOA-current 401 mA
open cavity doors for adjustments
12:55local stop comb-measurements for visitors etc
re-start CombControl - seems to work
at 10:41 local, computer reports uptime of 6:27 - so rebooted at 04:xx local?? (last datapoint in DB 04:08 local)
looks like menlo-computer rebooted during the night??
ping 192.168.10.18 - does not work
related to reading rep-rate from new menlo comb?
re-start of gnuradio and ZMQ-logger doesn't help?
gnuradio ZMQ logging had failed for some reason?
12:20local laser jumped to next TEM00 while moving cart in lab - still quite sensitive to mvoement in lab?...
Upgrade Picoscope to latest version from website - to see if that helps
PicoScope program on comb computer had crashed. Out of memory error, PicoScope seems to consume 3.3 GB of memory. Comb computer has 8 GB of memroy installed.
When PicoScope re-started it consumes only 50 Mb of memory.
Laser stayed locked to cavity over night.
Improved input-offset trim on FALC keeps error-signal bounded?
Also computer-control of laser temperature, to keep Fine-1 limited to +/-0.5 V - seems to work?
Cavity frequency during last 16 hours. Note step due to temperature spike in whole MIKES building!?
Cavity frequency (674nm) is computed from measured IR 1348nm at comb and the applied frequency to the double-pass drift-AOM
With A/C setting high enoguh, oscillations go away, at least with no changing heat-load in lab
Some disturbance in whole MIKES building morning of 2021-11-11 - this has large effect on measured laser frequency also (through cavity? or through fiber from laser to comb?)
Collect 100 samples of Fine-2, and average them together - since signal is noisy. averaged data in DB from about 16:01 local
Connect FALC-monitor, via low-pass filter, to DLC Fine-2, to allow monitoring of this voltage
15:35 local, Fine-in has not drifted much, ca -0.01 V, FALC-monitor average at ca -0.1 V (scope with low-pass filter)
14:20 local re-lock laser with new improved(?) input offset trim
With both FALC-inputs grounded, monitor-output showed almost +3V !? adjust offset so that now monitor-output shows ca -4 mV (trimpot quite sensitive, this is as good as it gets?)
FALC integrator always seems to drift up? Try adjusting input offset pot?
Cavity drift was wildly changing yesterday when doors were closed, now after a short measurement seems to have settled at 10 mHz/s (similar to old values)
11:35 local, increased A/C VFD 40 -> 45 Hz
Try increasing further, to eliminate oscillations?
A/C VFD was increased from 35Hz -> 40 Hz on 2021-11-09 at 14:35 local
40Hz setting gives triangular-shaped T110 oscillation with peak-peak of ca 0.3 C
11:20 local re-lock laser to TEM00
Clock-laser stayed locked until ca 02:15 local
Reason for unlcok is not that fine-1 goes out of range - but something else?
Maybe check alignment into cavity, coupled power in/out, and optimize PDH error signal?
Temperature adjusted by python script in 0.0002 C increments to keep abs(Fine-in) < 0.5 V. Seems to work
With doors closed innermost T-sensor settles around +3...+4 mK higher than with doors open
Slow adjustment of Toptica temperature seems to work at least initially. Fine-in is a bit noisy, could be better to look at average over past 10min?
Leave laser locked and comb running for the night. abs(Fine-in) should stay < 0.5 V if everything works!?
Try adjusting laser-temperature by 0.0001 C once per minute whenever abs(Fine-in) is too large.
Limit for abs(Fine-in) set to 0.5V while testing
Limits on possible temperature range set to 18...22 C.
max range of Fine-in seems to be +/- 4V ?
sensitivity roughly -0.1 V fine-in / 0.001 C laser temperature
setting temperature to 19.999 C makes fine-in rise to ca 1.52V - timeconstant maybe 10-20s
16:10 local Fine-in ca 1.44V - maybe need to very slowly adjust temperature in order to keep Fine-in voltage limited?
PDH lock FALC output DA2 (Fine-in-1 on DLC) voltage now up to 0.92 V (what is the maximum?)
ca 14:35 local: increase A/C VFD from old settings 35 Hz to new setting 40 Hz - more air flow to eliminate oscillations?
ca 14:30 local: close 4th and final door
As usual total heating power falls from 90 W with doors open to maybe 30 W with doors closed. Takes around 4 hours.
Old results on thermal expansion from SIB04 IonClock project:
y = - alpha * ( T - T0 )**2
Coarse values: alpha = 7(1)e-10, T0=32(1) C
Measured with HeNe laser in steel vacuum tube.
674nm repeated results. Thermalization for 10 days after set-point change.
alpha = 6.2(2)e-10 and T0 = 31.4(2) C
Observed shift at 1348nm is ca 20 Hz - when cavity inner temp changes ca 2 mK.
20 Hz / (0.5 * f_SRS ) = 9e-14
dy = -2*alpha*(T-T0)*dT
with alpha=7e-10, T-T0 = 40 mK and dT = 2 mK get -1.1e-13
Inner shield temperature sensor over one week:
Laser frequency and lab-temperature seem to be correlated.
Try closing all doors and/or stabilizing lab temperature
Toptica PDF mpdulation amplitude -0.2 dBm, phase-shift 57.8571 degrees (from Windows monitor program)
ARC-factor is -0.5 V/V, so PC voltage changes -0.5 V/V based on Fine-in-1 voltage.
Toptica IP 192.168.10.52
Logging from toptica required new install: sudo pip3 install toptica.lasersdk
09:45 local Lock laser again, found TEM00 mode gives beat-note 16.657 MHz
Laser unlocked at some point, found in TEM01 mode.
Comb PicoScope program had crashed during the night
Leave clock-laser and comb locked, measurement running
Drift-changes during the day are significant? But one door of cavity-box is still open..
ARTIQ drift-dds adjusts to the measured value of 2h drift - but this means the applied drift-correction always lags behind a bit
Make a list of TEM00 modes in the cavity and their beat-notes?
Comb frep_dds = 19 999 896 Hz -> frep = 249 999 896.5 Hz
TEM00 modes of the cavity appear at similar beat-note frequencies because FSR of cavity is ca 500 MHz, and comb-spacing is 250 MHz in IR, ca 500 MHz in red
Old measurement of TEM00 FSR was 501.305 MHz -> should give 652.5 kHz shift of beat-note for each TEM00-mode with 250 MHz rep-rate
17 962 088 Hz Delta-from-last (start, 5min measurement, only valid datapoints)
17 309 598 Hz -652 490 Hz (increase ECDL piezo-voltage)
16 657 106 Hz -652 492 Hz (ECDL piezo 51.856V)
16 004 610 Hz -652 496 Hz (piezo 53.033V)
15 352 114 Hz -652 496 Hz (piezo 54.202V)
Cavity FSR for 674nm seems to be 2*(249 999 896.5 + 652 496 ) = 501 304 785 Hz (in agreement with old estimate 501.305 MHz)
Note uncompensated laser will drift during a 300s measurement by maybe 20 mHz/s*300s = 6 Hz!
Currently not sure how well AOM driftcomp matches true drift of cavity
close 3 out of 4 doors of the clock-laser, maybe it stays locked better like this?
ARTIQ drift-DDS was re-started (old-setting of -6.050 mHz/s (single-pass 674nm))
if measured 2h drift is within reasonable limits
drift_low = 2e-3
drift_high = 30e-3
Then ARTIQ applies the new drift value.
Testing SDR-tracked with lower loop bandwidth of pi/10 - 100 kHz with 2MS/s sampling. (default value was pi/5, 200 kHz using 2 MS/s)
No major change to beat-note sum histogram
Note applied drift of ARTIQ-DDS adjusts to the 2h-drift measurement value (measurement needs to run 2+ hours without disturbances)
Note predicted Zeeman AOM frequency flattens out when drift-correction matches true drift of cavity
(AHM3 model was not updated, need to adjust these values before ion-experiments!)
ADEVs as expected. This could be improved by adding the cavity-ADEV also, but need to merge 1S/s frequency with drift-DDS for this
Add also monitoring of measured rep-rate and f0 from comb?
Improved beatnote-tracker program was running for 1-2 hours with clock-laser locked
fbeat1 + fbeat2 = freprate seems to hold with low uncertainty
Old 100MHz comb had 5 mHz tolerance for this sum - but now all valid points seem to be within 1 mHz of the programmed rep-rate
histogram of 1 hour (3600 points). Only 4 points are outside the histogram (no people or movement in the lab during measurement)
SDR-tracker now auto-tunes to center the IF frequency near zero.
InfluxDB results now also take into account menlo rep-rate tuning in the valid-flag.
Menlo computer connected to instrument-network: "http://192.168.10.18:8123"
XML-RPC reading of data seems to work from this IP also.
Work on gnu-radio SDR frequency-counter for beat-notes
pass-tags on ZMQ block gives len(data_str)=27. without tags len(data_str)=8
unpacked with struct.unpack('ff',data_str)
Comb PicoScope problem (not found). Re-plug USB and power (no effect), reboot computer -> issue fixed!?
SDR RF level.
GUI-sink just after USRP-source shows real signal level (PLL block outputs normalized data?)
peak at ca -77 dB
With comb blocked noise floor: ca -112...-113 dB - this is the noise-level of the photodetector
what are the noise peaks??
noise-floor of SDR-input only is much lower, -140 dB or so
Constellation-display shows signal range ca -0.015 to +0.015
With sine-wave input to SDR the same IQ level achieved at input of 50 mVpp
500mVpp (siglent siggen, 50ohm load) gives ca -0.15 to 0.15 IQ
1 Vpp gives ca -0.31...+0.31 IQ SDR spectrum-peak -19 dB
NOTE: photodetecto output at 250 MHz rep-rate is -4.5 dBm (100kHz RBW) - so we can't amplify a lot here?
Beat-note at 15 MHz has amplitude of -18.3 ... -12 dBm (depends on 1348nm fiber polarization changes? use PM fiber from laser to comb!?)
ZHL-500LN has max
output of +5 dBm at 1dB compression - we are now within ~9 dB of this
when detuning far away from the peak, SDR floor ca -111 dB, at peak ca -62 dB, so 'wide-band-SNR' ca 49 dB - ok!?
Menlo comb remote data interface?
local CombWatch interface is 127.0.0.1:8123 with "comb" / "system" as username/password (?)
VTT IP is 220.127.116.11
example field to read e.g. "counter0.freqref" should show the DDS value for reprate-lock (around 20 MHz)
password/username set to None in the fibercomb.setup file.
Now we can ask the server for a list of (non-system) methods like this:
comb_ip = "http://18.104.22.168:8123"
with xmlrpc.client.ServerProxy(comb_ip) as proxy:
methods = proxy.system.listMethods()
for m in methods:
and we get an answer:
$ python3 menlo_test.py
Try reading reprate: (This is the DDS setting, rep-rate is 245MHz + DDS/4)
h = proxy.hello()
Fiber Comb Control v0.6.16
rep_set = proxy.comb.reprate.synth.getFreq()
Try listing parameters:
plist = proxy.param.list()
for p in plist:
Try reading a parameter:
x = proxy.param.get("Actuators.SYNCRO_XPS::DDS0/FREQ")
x = proxy.param.get("Counter.Offset")
What is the data-function?
data.getInfo() returns a large dict describing data:
data history size: 3502
data history max: 3600
data last_timestamp: 1635855968.875568
data.query() seems to return the entire history-buffer as a nested dict
Beatnote received with SDR, 2 MS/s
FFT 32 kpts, average N=10
first minimum ca -72 dB, peak ca -38 dB, so peak-SNR ca 34 dB
(what is dynamic range of N210 SDR? should we use more or less gain?)
With these settings (P -6dB, I 20 Hz) SNR with 100 kHz RBW seems to be >47 dB
tace is N=20 average
central peak is ca +35 dB from first minimum
Now keeping Pgain constant at -6 dB, try different I gain settings
I 315Hz (trace D)
I 150Hz (trace C)
I 75 Hz (trace B)
I 20 Hz (trace A)
Try measuring beat-note with different RRE gain settings.
P 18 dB (trace A), 10 dB (B), 2 dB (C), -6 dB (D):
I gain settings constant 315 Hz in these measurements
SA trace average of N=20
C or D seems best?
SYNCRO-CEO settings from test-report: P 2d B, I 16000 Hz
Laser temperature setting was ca 23.5C (set too low accidentally?) - set to test-report value 25.5C
Back to RF lock
- Pico D: Monitor RRE
- C1 signal (lower)
- settings from test-report
Menlo setup/training was on Friday 2021-10-29
Locking to optical reference tested, settings:
- Beat-note RF to back of rack OFD input I
- PicoScope D: Monitor CW
- Input C2 (top) used
- P gain 12.6 dB
- I gain 20 kHz
- these P/I gains not optimized, but seems to work.
Try to measure absolute frequency with new comb
TEM00-mode with wavemeter reading 1348.052 nm
SDR-counter with one channel directly measured, the other through mixer (LO set to 220 MHz)
beat1 moves up when frep moves up
frep_DDS beat1 beat2
19999584 -15 475 999 +234 523 897 (beat1 moves up when frep moves up)
20000569 +15 470 491 -234 529 651 (beat1 moves down when frep moves up)
frep = 245 MHz + frep_DDS/4
(4th harmonic of rep-rate at 1 GHz is mixed with 980 MHz PLO and then compared to frep_DDS)
beat1 results give n1 = 889557.979 = 889558.
this gives 2*f_IR - SRS = -330.075547 MHz
m = n2-n1 = -1
sign of f0 seems to be negative.
f0 = -35 MHz
beat2 results give k1 = 889557 (one comb-tooth lower than beat1)
m = k2-k1 = +1
2*f_IR-SRS = -330 075 547 Hz (same result as for beat1)
Eqn for mode-number from Menlo manual
m = n2-n1
Delta1 and Delta2 are uncertainties in measurement of fbeat1 and fbeat2.
Here frep1-frep2 = 250 000 142.25 - 249 999 896 = 246.25 Hz
To compute n1 with uncertainty <<1 we need to measure fbeat1 and fbeat2 with e.g. uncertainty <<25 Hz
This is roughly the maser-limit at 1s of averaging, so averaging for 100s or so gives n1 with low uncertainty.
Note (uncompensated) cavity drift when settled 10 mHz/s, so laser drifts 1 Hz during a 100s measurement.
Found more 1348nm power from clock-laser -> much better SNR for beat-note
power-meter: -2 dBm clock-laser (was -13 dBm!), comb only -21.5 dBm (M-NIR diode currents reduced, 121mA, N/A, 1.2 A, 308 mA).
>40 dB SNR quite easily achieved
Zoom around one beat-note. Bumps around main peak are from PDH-lock
beat-note tests with new comb. M-NIR free-space output
test-report spec is >35 dB SNR with 100 kHz RBW. Initial try gives ca 20 dB SNR, sensitive to 1348nm-fiber (from clock-laser) bends - replace with PM fiber?
black fiber-power-meter shows -13.2 dBm from clock-laser, and -11.8 dBm with clock-laser + comb-light.
(center peak are beat-notes against N and N+1 comb-mode, right peak is 250 MHz rep-rate)
Detector is Fermionics FD80FC InGaAs photodiode, 5.1 kOhm transimpedance amp (OPA818),
with Mini-Circuits ZFL-500LN providing 28 dB postgain
Install new Menlo comb
Move old Menlo comb
Newport table pressure setting is 2 bar
VFD back to old setting 35Hz - seems to give damped(?) oscillations.
With VFD at 25Hz system seems stable again - but temperature rises when working in lab..
At 11:30 local, set KSK105 to 25 Hz
KSK105 -3 floor was on 35 Hz.
Switch on lights ca 10:30 local time.
Note temperature oscillations during the weekend. Monday morning all lights were off (too little heat-load from lamps causes oscillations?)
Old 100 MHz comb also off.
Ion still in the trap!
Both RF-monitor and secular frequency show relative change of ca 0.00005/0.06025 = 8e-4, for the 1.8MHz axial frequency this is 1500 Hz.
Compare to oscilloscope resolution at 8-bit which is 1/256 = 4e-3.
Similar settings in late May gave 1.7898 MHz, now ca 1.7885 MHz. NOTE axial mode has lower frequency than normal because V12common = 0V.
Secular axial mode measurements from yesterday, RIDs 15041 - 15072, together with new RF-monitor measured with DMM
ca 14:00Z turn off lasers. turn off PMT. leave ion trapped.
RID 15042 - 15072: secular tickler, axial mode, C-beam
RID 15041: secular tickler, axial mode, C-beam.
RID15036: secular tickler, axial mode 1.781-1.79 MHz with A-beam
u-motions again A 15%, B 25%, C 28%
Fluorescence level changes very much when heating the B-fiber!
Optimize beam alignment A, B, C, ASE - but all were quite ok to begin with..
422 lock to Rb-line is quite unstable. unlocked twice while measuring u-motions etc.
umotions: A: 13% (7 s), B: 13% (15 s), C: 22% (8 s)
set 422-detuning to 208 MHz
11:44Z: Try to load: 1.6A, 1.59 W. Ion! at 11:52:40Z!
Compensation voltages: [-0.24511, +0.24511, -15.7661, +84.3065]
405: 3.3 mW.
ASE powers: 1092: 11.5 mW, 1033: 68 uW
Using ASI camera, no intensifier
Only inner magnetic shield is installed
422nm detuning: 185 MHz.
422nm on ion breadboard in A free-space beam ca 20uW. A-objective 1.5uW, B 2.5 uW, C 1.8 uW
461 power: 440 uW on ion breadboard. Check with Sr dispenser: 196.48 mA, 4.642 kOhm.
11:50Z Switch lasers off.
11:00Z Switch lasers on. No ion in the trap?
14:28Z turn off lasers. leave ion trapped
Detuning-scans at V12cm=0V:
NOTE 0.5*axial plotted, to get all datapoints close to eachother
Summary of results:
[-5.8 x 26.6721 x] V -> ( 958830, 1003244, 1975363) Hz
[-5.0 x 27.7201 x] V -> ( 982384, 1025684, 1950869) Hz
[-4.0 x 29.0301 x] V -> (1010912, 1053126, 1919859) Hz
[-3.0 x 30.3401 x] V -> (1038747, 1079756, 1888188) Hz
[-2.0 x 31.6501 x] V -> (1065803, 1105771, 1855910) Hz
[-1.0 x 32.9601 x] V -> (1092152, 1131241, 1823320) Hz
[ 0.0 x 34.2701 x] V -> (1117918, 1156046, 1789792) Hz
V12diff = 0.2451 V constant
V34diff = 50.0363 V constant
shift per volt = (+25800, +24800, -33500)
axial scan 1788-1792 kHz
rhi scan 1154-1158 kHz
rlo scan 1116-1120 kHz
uMotion A 2.6%, B 5.5%, C 7.2%
Set Vbias = [0 x 34.2701 x]
shift per volt = (+25500, -32600)
axial scan 1821-1825 kHz
rhi scan 1130-1134 kHz
uMotion A 3.4%, B 4.1%, C 7.1%
Set Vbias = [-1 x 32.9601 x]
shift per volt = (+2700, +26000, -32300)
axial scan 1854-1858 kHz
rhi scan 1104-1108 kHz
rlo scan 1064-1068 kHz
uMotion A 3.9%, B 2.4%, C 6.9%
Set Vbias = [-2 x 31.6591 x] V
shift per volt = (+27800, +26600, -31700)
axial scan 1887-1891 kHz
rhi scan 1078-1082 kHz
rlo scan 1037-1041 kHz
uMotion A 3.4%, B 3.2%, C 7.3%
Set Vbias = [-3.0 x 30.3401 x]
Observed shift per volt = (+28500, +27400, -31000) Hz
axial scan 1918-1922 kHz
rhi scan 1051-1055 kHz
rlo scan 1009-1013 kHz
uMotion A 4.4%, B 4.6%, C 7.6%
set Vbias = [-4 x 29.0301 x] (x not changed)
Observed shift per volt V12cm (x, y, z) = (+29200, +27700, -31300) Hz
axial scan 1947-1954 kHz
rhi scan 1023-1027 kHz - could use even more tickler than 7dB att
rlo scan 980-984 kHz
uMotion A 4.4%, B 5.1%, C 7%
rhi ca +2.7 kHz/0.1V, expect +22 kHz shift
axial ca -3 kHz/0.1V, so expect -24 kHz shift
Set new Vbias = [-5.0 0.2451, 27.7201, 50.0363]
(x, y, z) = (958830, 1003244, 1975363 Hz ) Yesterday: (959049, 1003521, 1975940)
fit fails - too wide scan?
Vbias = [-5.8 0.2451 26.6721 50.0363]
NOTE: figure P422 and P1092 not measured, just guesses. Program doesn't include window-transmission difference for A/B and C beams
uMotion 100kphot: A 4.7% in 13.9s, B 3.7% in 18.9s, C 7.5% in 13.8s
Fit uses -1 MHz correction, so 2*208 - 436 - 1 = -21 MHz (this agrees with the scanned points in the figure)
Is 208 MHz a good set-point for cooling?
ion appears maybe 20s after completed detuning-scan! (only points -23 MHz from 208MHz were scanned, no points in blue direction)
try detuning-scan to get cold ion?
no ion during first 60s with only A-beam, or during about 60s with ABC beams
detuning is 208 MHz
11:06Z turn on lasers.
synced with gitlab, new secular_prediction.py program
14:25Z lasers off. leave ion trapped
New version 'v2' of detuning-scan program with approximate lineshape fit.
TODO: automatically measure background-counts before or after the scan (or take this from blue-detuned data points?).
Add parameters for P422_A,B,C and P1092
uMotion A 6.1% in 10s, B 3.3% in 15.8s, C 4.3% in 10.5s
lasers on, ion still trapped
14:31Z turn off lasers. leave ion trapped.
Test how repeatable axial secular results are. RID14875 - RID14884 (10 repeats). each measurement ca 3 minutes.
(measurements via remote desktop, no activity in the lab)
f0 = [1975659, 1975684, 1975645, 1975599, 1975633, 1975598, 1975663, 1975607, 1975562, 1975502]
FWHM = [671, 559, 580, 610, 695, 628, 508, 630, 795, 657]
Standard deviation for resonance frequency ca 50Hz, for FWHM ca 75 Hz.
Systematic effects due to lab temperature change or helres-drift, cooling-laser intensity/polarization are probably (much) larger?
Trap parameter results:
Vbias -> [omega_secular x, y, z] (uMotion A, B, C)
[-5.8 0.2451 26.6721 50.0363] -> [ 959049 1003521 1975940] (7.7%, 1.2%, 2.8%)
[-5.9 0.2451 26.5411 50.0363] -> [ 956205 1000753 1978978] (7.1%, 1.6%, 2.5%)
[-6.0 0.2451 26.4101 50.0363] -> [ 953220 997869 1982072] (not measured)
[-6.0 0.2451 26.4101 50.0363] -> [ 952917 997615 1981450] (7.9%, 2.1%, 3.5%)
seculars at V12cm = -5.8 V
get uMotion A 7.7%, B 1.2%, C 2.8%
Go to Vbias = [-5.8 0.2451 26.6721 50.0363]
Seculars at V12cm = -5.9 V
get uMotion A 7.1%, B 1.6%, C 2.5%
Go to Vbias = [-5.9 0.2451 26.5411 50.0363] (change is +0.1 * dVbias)
Trap parameter analysis suggests voltage-vector that doesn't change field at ion:
dVbias = [1, 0, 1.31, -0.03] (V34diff change is small - ignore?)
Repeat secular measurement:
953.2 / 997.9 / 1982.1 kHz @ Vbias = [-6 0.2451 26.4101 50.0363], uMotion
(these are +315 PPM / +301 PPM / +339 PPM compared to values measured before lunch )
Ion went dark during axial scan!? (too much tickler with 20+10 dB setting!) Fluorescence recovered with detuning-scan
RID14835 C-beam, axial secular mode with C-beam, 20+20 dB
RID14836 C-beam, rhi 20+10 dB scan 995-1000 kHz (peak height 15%, could use less attenuation?)
RID14838 C-beam rlo 20+10 dB scan 950-955 kHz, 50 points (side-to-side blur at resonance in camera image, with good imagination..)
952.9 / 997.6 / 1981.4 kHz
PMT counts with 1000s gate, time-series ca 60s
feels like there is quite a lot of variation in P422
P1092 8.90 mW
422 beam power (signal-background): A 0.55-0.08 uW, B 0.90-0.15 uW, C 0.79-0.07 uW
RID14821 C-beam scan, max 21.7 kcounts, C fluorescence 8859 (1384) / 167 (16)
RID14818 B-beam detuning scan. max 12.7 kcounts, B fluorescence 5531 (228) / 64 (8) bright/dark
RID14815 A-beam detuning scan. max counts ca 21k/1000ms. A fluorescence 9342 (195) / 65 (9) bright/dark.
ID14808 detuning-scan, A-beam, 1000ms gate, -23...+10 MHz from 208 MHz set-point in 60 steps
A-fluorescence in 1000ms is 6854 (240) / 57.3 (8), repeat 5938 (281) / 67 (6)
uMotion A 7.9% in 15.9s, B 2.1% in 17.5s, C 3.5% in 14.5s
This ion loaded Monday 2021-05-17, and spent several nights + the weekend un-cooled.
07:57 ion appeared! after re-locking to Rb-line!? (ionization lasers were on also)
no ion in the trap? (tried 405nm and 461nm lasers also) tried 422 detuning-scan.
07:41Z turn on lasers
Notes on geometry and voltages
Current voltages are from uMotion optimization run 2021-05-18 ID14607 NPL search, [ 0.24512619 26.41013719 50.03631515 ], V12cm = -6 V
13:38Z: Switch off lasers for the weekend. Leave ion in the trap.
uMotions: A: 7%, B: 3%, C: 4%
12:13Z: Switch lasers on. Ion still in the trap!
15:42Z ion is still stored and cooled. Turn off lasers, leave RF on.
P1092=11mW assumed. Note new C-beam focus seems OK, almost as good as A. Matlab-code fix for C-beam (IR window transmission)
422-beam power before chamber: A 0.65-0.08 = 0.57uW, B 1.15-0.18 uW = 0.97 uW, C 0.9-0.15 = 0.75 uW
RID14790 C-beam detuning scan
RID14787 C-beam fluorescence 10743 / 220 bright/dark (note >2-fold more dark-counts! 422-scatters into PMT!?)
RID14787 B-beam detuning scan, 1000ms gate
RID14785 B-beam 1000ms fluorescence 7785 / 67 bright/dark
RID14783 detuning-scan with A-beam, 1000ms gate
RID14784 A-beam 1000ms fluorescence @ 208 MHz is 9742 / 71 bright/dark (compare to ca 90 counts for the most blue-detuned points in the scan
axial around 1.982 MHz (same as before)
uMotion A 7.5% in 13.8s, B 1.7% in 17.7s, C 4.5% in 13.5 s
last experiments ca 17 local on Wednesday, so ion stored ca 23 hours without cooling
ion appears on camera ca 6-7 seconds after opening the 422-shutter. Detuning 208 MHz as usual
12:55Z turn on lasers
Mid-shield was removed y-day, inner shield is installed
Set RF back to 10 Vpp and switch off lasers
RF 4 Vpp, Monitors: CH3: 0.720000 V, CH4 1.400000 V :
Photon correlation tickler:
Lower radial 20+12 dB:
Upper radial 20+14 dB:
Axial 20+29 dB:
Axial 0.7637 MHz
Radials 0.34 MHz and ~0.448 MHz (??)
RF 7 Vpp, Monitors: CH3: 1.200000 V, CH4 2.419999 V :
uMotions: A: 10%, B: 2%, C: 5%
Photon correlation tickler:
Lower radial, 20+10 dB:
Upper radial, 20+12 dB:
Axial, 20+25 dB:
Coarse tickler, 0+11 dB:
Radials, 0.6517 MHz, 0.7146 MHz:
Axial: 1.363 MHz
RF 10Vpp, Monitors: CH3: 1.719999 V, CH4 3.479998 V :
Photon correlation ticker:
Lower radial, 20+8 dB:
Upper radial, 20+10 dB:
Axial, 20+23 dB:
Coarse tickler, 0+8 dB:
ID14673 Radials, 0.9529 MHz, 0.99753 MHz:
ID14674 Axial 1.982 MHz
uMotions: A: 4%, B: 4%, C: 3%
Remove mid-shield and install 2nd RF-monitor probe
10:18Z Switch lasers on. Ion still in the trap!
14:36Z turn off lasers. leave ion trapped.
Install fixed 20dB attenuator. Axial with 20+23 dB
Try axial mode without fixed attenuator, 0+30 dB dds attenuation.
Too much tickler amplitude.
15:53+03 uMotion 100kphot: A 3.6% in 13.2s, B 2.7% in 20.9s, C 3.9% in 22.2s
14640, 14641, 14642 improve program so beam-letter is included in figure? (include also photons/point?)
Lower radial: (easier to excite, so peak 2-fold higher with same tickler-attenuation)
height also inversely proportional to FWHM (cooling rate) ?
Upper radial just below 998 kHz (harder to exite wrt. lower radial)
A, 0+23 dB. FWHM is large, so mode aligned along beam??
Lower radial around 952.8 kHz
Correlation with 0+23 dB attenuation
C, 0+23 dB. Peak height 0.6, too high?
B, 0+23 dB.
A, 0+23 dB.
RID14625 photon-correlation tickler, 20+20 dB, A-beam. 20kphotons/pt. inner+mid shield installed. ASI camera
A, 20+20 dB
A, 20+23 dB
B, 20+23 dB, FWHM is almost 2-fold larger than A
C, 20+23 dB,
Switch to ZWO ASI camera
uMotions at new optimum: A: 6%, B: 1%, C: 2%
ID14607: NPL search, [ 0.24512619 26.41013719 50.03631515]
Intensifier powered from Siglent lab-supply; axial:
Ion was lost?
ID14591: NPL search for uMotion minimum. 422 lost lock, fail.
uMotions: A: 8%, B: 20%, C: 22% at [-6, 0.239, 25.919, 51]
06:00Z Ion still in the trap!
Tried with another +12V source. Similar result. Investigate intensifier power box?
After removing intensifier power supply:
ID14563: Switch intensifier off:
ID14562: After installing mid-shield, now 0+28 dB:
ID14555: Intensifier on
ID14552: After installing shield, intensifier still off.
Install second half of inner shield.
ID14551: Intensifier off, second peak diminishes slowly?
Switching off intensifier, still two peaks visible? Relaxation time?
ID14543: Removed foil, now two peaks visible for C beam axial:
ID14510-13: Intensifier switched off, lower radial, photon timestamp correlation, 20+11 dB
ID14508: Intensifier: lower radial, photon timestamp correlation, 20+11 dB
ID14507: Intensifier: lower radial, 0+8 dB
Install intensifier and set to max gain.
ID14494: ZWO ASI: lower radial, 0+12 dB
Results below were obtained with ZWO ASI camera
ID14490: Lower radial, C beam, photon correlation, 20+11 dB
ID14489: Upper radial, C beam, photon correlation, 20+11 dB
ID14488: Axial, C beam, photon correlation, 20+25 dB:
ID14487: Axial, C beam, attenuation 0+25 dB. 1.9815 MHz
ID14485: Upper radial, C beam, attenuation 0+8 dB. 0.9972 MHz
ID14484: Lower radial, C beam, attenuation 0+8 dB. 0.95228 MHz
ID14482: Coarse tickler: Radials at approx. 0.95234 MHz and 0.9972 MHz
uMotions at V34diff = 51 V: A: 4%, B: 24%, C: 24%
Sweeping V34diff gives new guess for minimum for A: 51 V.
uMotions: A: 60%, B: 67%, C: 70%
11:20Z Try to load. I = 1.6 A. Ion! at 9min 30s
Check 461 dispenser. Set I from -196.83 mA -197.11 mA
10:49Z Try to load. I = 1.6 A. No ion at 10 min. Flush trap and continue. After 6 min two ions appear. Flush trap and continue. No ion for 10 mins.
Check resonator Z-matching: Minimum 1.01 at 14.4236 MHz. 1.05 at 14.424 MHz. Adjust minimum to 14.424 MHz, now 1.01.
10:29Z No ion in the trap.
14:13Z turn off lasers
uMotion A 45%, B 41%, C 46%
axial with photon-correlation, B-beam, 20+25 dB attenuation, 10 kphotons/pt.
Upper radial with correlation method. 20+11 dB attenuation, B-beam, 10kphotons/pt
coarse tickler 0+3 dB. half of axial (?) and upper radial
coarse tickler 0+14 dB, lower radial, B-beam
photon-correlation method, 20+11 dB attenuation, B-beam, 10 kphotons/pt - looks ok!?
lower radial with 0+14 dB, ca 950.7 kHz
0+14 dB artiq DDS, scan around upper radial
0+4 dB artq dds tickler shows both of these.
Upper radial scan: peak below 0.99 looks like half the axial mode, around 0.996 is the upper radial
Switch to ASI USB3 camera, to get more magnification
siglent siggen 1.8 Vpp hiZ amplitude for upper radial, manual tickler
0.996 and 0.95 MHz radial modes. Upper harder to excite.
Axial mode 0+24/26 dB attenuation
siglent siggen 280mVpp in hiZ mode - for manual tickler
coarse tickler shows resonances at 1.979 and 1.980 MHz
moving camera up/down and l/r increased PMT counts to >200/10ms (from 20/10ms)
12:36Z ion! (tried adjusting 461nm laser-current during loading)
12:14:15Z load with 1.60A
1s camera exposure
detuning to 185 MHz for loading
throw away two ions
align A-beam (not much improvement in fluorescence counts 20-40/10ms)
At 10Vpp see axial mode around 1.980 1.982 MHz
reducing trap drive to 4Vpp shows two ions
detuning to 208 MHz
11:39:15Z (ca 10mins)
11:29Z load with 1.60A (on 6.5 loading took up to 12 mins!?)
Bias voltages from previous settings -6 0.239 25.919 55 V
Detuning 185 MHz for loading
RF on, 10 Vpp drive, Rigol monitor voltage 3.46 Vpp
old camera with intensifier
11:27Z turn on lasers.
Try to record images of secular modes, to see what is causing double-peaked secular resonances?
NOTE: results below are due to filter series-resistance together with 10 MOhm input impedance of DMM!
Sanity-check of bias voltages?
V1 on bias-voltage source dispplay: -6.23901 V
V1 from bias-voltage BNC-output (no cable connected): -6.2379x V (last digit x unstable, Keysight 34461A DMM)
V1 from end of pink SMA-cable -6.2380x V (not connected to filter-board)
V1 from green screw-connector (with blue wire to trap conencted): -6.2116x V (voltage is 0.0264 V more compared to unloaded bias source? has this been measured before?)
V2 on bias-source display: -5.76101 V ( 0.5*(V2-V1)=0.239 V is the V12diff programmed - seems OK with V12cm=-6V programmed)
V2 from bias-voltage BNC-output: -5.7601x V
V2 from end of pink SMA-cable (not connected to filter-board): -5.7602x V
V2 from green screw connector -5.7356x V (0.0245 V more than from unloaded bias-source)
14:40Z Turn off lasers, leave ion trapped
back to -6V V12cm
Test with V1 and V2 bias-box outputs disconnected and V1/V2 cables terminated with 50R. Still double-peak. Now 10Vpp drive
1.783 and 1.7848 MHz
two axial modes ca 2 kHz apart seen also with siglent generator connected to piece of wire, held in front of the IR-window (IR-side inner shield removed)
Axial secular scanning with Siglent generator (no ext ref). No external attenuator. -12 dBm output (into 50 ohm). Weaker excitation ~1.559 MHz, stronger excitation ~1.5609 MHz.
RID14286 Try correaltion method with more photons 40 kphot/pt. C-beam 20+20 dB att.
8 Vpp C-beam axial, 20+9 dB att, 80 ms gate.
8 Vpp C-beam axial, 20+9 dB att, 40 ms gate. longer gate reduces false peak? measurement is long enough to show 422 power variations
8 Vpp C-beam axial, 20+9 dB att, 20 ms gate
8 Vpp C-beam axial, 20+9 dB att, 10 ms gate
ID14277 try C-beam correlation method, with a lot more attenuation 20+20 dB. 10 kphotons/pt
8 Vpp C-beam axial, 20+20 dB att
8 Vpp C-beam axial, 20+15 dB att
8 Vpp C-beam axial, 20+9 dB att
8 Vpp B-beam axial, 20+9 dB att (B should have more projection onto z?)
8 Vpp A-beam axial, 20+9 dB att
8 Vpp A-beam axial, 20+8 dB att
8 Vpp A-beam axial, 20+7 dB att
uMotion A 3.5%, B 5.7%, C 8.9%
Set RF-drive 8Vpp
9Vpp A-beam axial, 20+7 dB att. NOTE sudden shift down -2kHz(?) when working close to the trap
9Vpp C-beam axial, 20+7 dB att
9Vpp B-beam axial, 20+7 dB att
9Vpp A-beam axial, 20+7 dB att
9Vpp A-beam axial, 20+5 dB att
uMotion A 4.3%, B 4%, C 8.5%
Set RF-drive 9 Vpp
shift of about -0.5 kHz during a few minutes?
A-beam axial, 20+5 dB att
C-beam axial, 20+5 dB att
B-beam axial, 20+5 dB att
Shape of Rb-lock signal changes a lot when moving close to the laser...
Axial ca 1.9745 MHz, 20+4 dB with all beams
axial, 20+4 dB att
uMotion A 6%, B 2.4%, C 6%
Ion seems to be still trapped, fluorescence re-appears a few s after opening 422-shutter
07:23Z turn on lasers.
Turn off lasers, intensifier. Leave RF on and ion trapped
Double-peak is some nonlinear effect due to too strong tickler??
axial, 20+7 dB att
axial: lower resonance seems weaker, upper is strong with 28 dB att.
axial, 0+28 dB att
axial, 0+25 dB att
upper radial 0+8 dB att
lower radial 0+12 dB att
coarse tickler 0+8 dB shows radials ca 944 and 989 kHz. upper harder to excite.
back to V34dif=55V: get uMotion A 2.6%, B 5.9%, C 7.5%
uMotion A 68%, B 65%, C 60%
All beams illuminate ion ok
detuning to 208 MHz
13:40:30 signal - why 12 min loading-time?
13:28:30Z load with 1.60A
Detuning to 185 MHz
RF on/off to throw away the bright/dark pair
one axial mode maybe 947 kHz
sometimes shifted sideways position when scanning axial tickler
Ion jumping side-to-side seems to indicate one bright ion and one dark ion?
Axial mode ca 1.979 MHz. Coarse tickler with 0+28 dB seems too strong
uMotion A 8%, B 68%, C 44%
A-optimum at V34diff 50V
Trap voltage down to 4Vpp - no clear sign of 2 ions.
uMotion 100kphot: A 66%, B 77%, C 6% (but 11% 2nd harmonic)
A-alignment OK, B some optimization needed, C completely dark - optimization needed
set detuning 208 MHz
12:59 signal! weak cloud (not cooled?) at first - then cold and small image (later this looks like it was one dark and one bright ion)
12:53Z load with 1.60A.
405nm power ca 3.5mW
rf off/on to empty trap
12:47 no signal... check Rb-lock. check 461-dispenser 196.90 mA (close to previous setting)
12:34:30Z load with 1.60A, 1.58W initially. Detuning ca 186 MHz. Yesterday loading time was ca 4 min
Back to voltages -6.0 0.239 25.919 55 V
Install new filter-board - ion was lost during start of removal of old filter-board
Set 0 0 33.719 55 V (uMotion A 23%, B 49%, C 34%
11:03Z turn off monitor voltage logging with rigol
At 0 0.239 33.719 55 V get uMotion: A 0.6%, B 4.8%, C 5.5% - seems to confirm 1:1.3 ratio for V12cm:V34cm
Previously 1.3-fold change in V34cm gave micromotion-null. 25.919+1.3*6 = 33.719 V
Down to V12cm=0V in 0.5V steps. This gives uMotion A 70.4%, B 64%, C 67%
Voltages were -6 0.239 25.919 55 V
Try to lower abs(V12cm) before changing filter-board
Fluorescence appears again some 10s of seconds after re-locking to Rb-line.
Rb-signal changes shape a lot, and is unstable. Now quite a lot of negative DC-offset (previously positive)
10:51 cooling laser Rb-lock unlocked!?
Something in RF-drive/helres not very stable, now axial suddenly higher again, and monitor voltage also higher?
briefly seem to see 1.87 MHz or higher axial, then goes lower to below 1.86 in a few minutes!?
Axial mode with correlation method
B-beam 0+30 dB, 20 kphotons/pt
B-beam 0+30 dB, 20 kphotons/pt, wider range 1858-1863 kHz. cooling doesn't work for upper resonance!
B-beam 20+13 dB, 20 kphotons/pt, (added physical attenuator)
Check spectrum of tickler-DDS with Siglent SA. Looks OK, no visible sidebands or spurs.
With 18 dB DDS attenuation ca -8 dBm peak with spur-free SNR of ca 80 dB
correlation method for upper radial
B-beam 0+20 dB
B-beam 0+18 dB, 935-945 kHz
correlation method for lower radial. no physical attenuator. C-beam, 10kphotons/point
C-beam 20kphotons/pt, 0+16dB attenuation
A-beam 20kphotons/pt, 0+16dB attenuation
B-beam 20kphotons/pt, 0+16dB attenuation. ion not cooled at lower resonance
B-beam 20kphotons/pt, 0+20dB attenuation. ion image/cooling ok through resonance
axial secular now around 1.86 MHz (down from 1.886 MHz at end of Tuesday)
Radials ca 892 khz and 940 kHz
Upper radial is harder to excite.
C-beam 0+27dB, 10ms gate
C-beam 0+8dB, 10ms gate
C-beam 0+10dB, 10ms gate
Detuning is 208 MHz
Fluorescence counts in 10ms: A 158/4, B 111/4, C 129/6
uMotion 100kphotons: A 5% in 9.3s, B 3% in 12.3s, C 4.5% in 10.5s
07:15Z lasers on, ion seems to be still stored. camera image maybe a bit defocused?
This is maybe the 2nd (?) time ion has been stored without cooling over night?
sometime during the night, monitor-voltage dropped from 3.36Vpp to 3.2Vpp (rigol scope)
14:30Z turn off lasers. leave rf on and ion trapped
B-beam counts 66/0.22 in 10ms
B-beam power ca 0.6 uW
C-beam 0+28dB, 100ms gate
C-beam 0+30dB, 100ms gate
B-beam 0+28dB, 100ms gate
C-beam power ca 0.6 uW
C-beam ca 70/1.5 in 10ms
misalign 422 fiber-coupling (laser breadboard)
C-beam power ca 1.3 uW (too much?). Fluorescence counts in 10ms is 137/7
uMotion 100kphot: A 1.9%, B 4.6%, C 6.2%
Back to 208 MHz
0+26dB, 100ms gate
0+26dB, 100ms gate
Try detuning 185 MHz
0+26dB, 50ms gate
0+26dB, 100ms gate (note different rage)
Try changing V12cm +0.2V to -6.0, and V34cm +1.3*0.2V to 25.919V - should change axial/radial ratio?
axial 0+24 dB
axial 0+27 dB
axial 0+26 dB
axial 0+26 dB (high to low)
axial 0+26 dB (lo to high)
B-beam, sweep over both radial modes
0+8 dB now also upper radial?
Try to measure axial mode around 1.7-1.9 MHz? 0+22dB attenuation
(sweep now from high to low)
(sweep now from high to low)
(low to high)
ca 1893 and 1895 kHz.
radials were 898 and 899 kHz
A-beam, 0+10dB attenuation 0.9-1 MHz, 2000 pts, 10ms gate (down from 100ms previously). check how repeatable results are?
Zoom in on 0.88-0.92 MHz and repeat: now 0+8 dB attenuation
peaks seem to move around a lot?? changing Rf-amplitude?
further coarse-tickler scanning 0+8 dB shows peaks: ca 862 kHz, 867 kHz, and 944 kHz !?
ID13938: Maybe two peaks? Too much power?
Attempt to measure with photon correlation.
ID13933: Two-peaked lower radial now with 13 dB instead of 3 dB!
Disconnected new tickler coupling (to lower electrode). Attached old coupling capacitor to upper electrode terminal.
ID13912 photon-correlation tickler 933-935 kHz, 0+29dB att, 100pts, B-beam 10 kphot/point
ID19311: Radial (lower) sweep (coarse): 0.93236 and 0.93372 MHz
ID13903: Axial secular sweep (coarse): 1.9632 and 1.96496 MHz
ID13902: Radial (upper) secular sweep (coarse): 0.978 and 0.980 MHz
setting V34diff to 55V gives A 2%, B 5%, C 2.5%
lowering V34diff to 56V gives A 18%, B 22%, C 21%
uMotion 100kphot A 39.9%, B 42%, C 42% (1st harmonic looks ok, with only little 2nd harmonic)
set detuning to 208 MHz - image now much brighter
10:28Z load with 1.60A. 1.56W initially, ca 1.60W after 1 min, 1.607W at 2 min,
set detuning 185 MHz
no cooling during lunch -> ions seem to have disappeared?
when lowering trap-drive from 10Vpp down the glitch is between 3.6->3.5 Vpp
two ions visible when lowering trap-drive to 4 Vpp. blink-test with 405nm confirms this
uMotion A 15%, B 3%, C 3% (but 11% 2nd harmonic?)
set detuning 208 MHz
08:50 ion signal!
08:43 load again with 1.60A. Now with better 461nm current value
08:41 signal - blinking? looks like many ions, so throw away
08:27 load with 1.60 A dispenser, 461 current driver at too high set-point. Set I=-197.0mA at 10 min.
08:15Z turn on lasers
no time for trapping today, turn off lasers
no sign of ion...
08:15Z turn on lasers, check if ion is still stored?
Turn off lasers. leave RF on and dark ion trapped (until Monday?)
422 power check after detuning scans (compare to before values below): A 1.0 uW, B 1.6 uW, C 1.3 uW
Detuning-scans with improved beam alignment now looks better:
ID13862, A-beam, 1s gate. Fluorescence at 208MHz in 1s gate: 13462 / 64. scan is -23 MHz to +12 MHz from nominal detuning in 60 steps
ID13865 B-beam 1s gate. Fluorescence(208MHz, 1s): 12106 / 55
ID13868 C-beam (new!) 1s gate. Fluorescence(208MHz, 1s): 10208 / 386 (compare the dark backround count to blue-detuned values in the scan?)
uMotion now A 9%, B 3.6%, C 3.9%
aligning ASE-beam improves counts from 50/10ms up to >200counts/10ms - a 4-fold improvement.
422 beam power check: A 0.7 uW, B 1.3 uW, C 1.3 uW - similar as before.
uMotion check: A 4.9%, B 3.2%, C 3.2% - these look OK, but why are detuning-curves so bad??
Typical detuning curve:
wide scan with 0+6 dB:
With 0+6 dB attenuation also a double-peak around 933 kHz!??
Stronger tickler possibly reveals lower radial mode around 933 kHz
radial sweep, 0+17 dB attenuation. 978.3 kHz peak, 980 kHz peak (half of axial?)
tickler-spectrum with 0+28 dB attenuation:
ID13810 detuning-scan C-beam, 100ms gate. Fluorescence at 208 MHz in 100ms is 326.53/14.98 (note almost 2-fold more dark-counts with C-beam, 422-beam scatters more into camera/PMT?)
ID13807 detuning-scan B-beam, 100ms gate. Fluorescence at 208 MHz in 100ms is 285.1/6.46
ID13803 detuning-scan A-beam, 100ms gate, -23 and +12 MHz from 208 MHz. Fluorescence at 208MHz in 100ms is 355.6/7.9
Fluorescence levels (10ms) are A 35.6/0.86, B 27.8/0.72, C 35/1.24
Algorithm suggests 0.239 25.659 57.290 where we get: (100kphot) A 6.4% in 31s, B 0.6% in 36s, C 1.7% in 32s
ID13788 uMotion minimization 2x2x2 grid.
(Note NPL minimization program previously had wrong V12cm=-5V setting, now set to -6.2V as used while manually scanning)
(Note V12cm = -6.2 V is held constant)
Based on this 0.202 25.5 57.5 should be A+B optimal? Get A 11%, B 2%, C 8% ( by luck? we found also a C-optimal point!?)
V34cm 26V: A-opt (26, 57), B-opt (26, 58)
V34cm 27V: A-opt (27,56) and B-opt (27, 59)
Try same scans at V34cm=28V: B-opt (28, 60), A-opt (28, 54)
further scanning shows B-optimal point ca (V34cm, V34diff) = (29.301, 61) V . should now change something so A-optimum at (29.301, 52.5) coincides?
now A 1.7%, B 82%, C 76% at voltages 0.202 29.301 52.5
scanning V34 diff shows A-minimum at 52.5V
at this point uMotion A 48%, B 74%, C 68%
optimum found 0.202 29.301 55.057
ID13755 uMotion scan 4x4x4 grid, 50 kphot/point, estimate duration 50min (lunch)
uMotion A 8%, B 70%, C 72%
scanning V34diff shows A-min at 61V
uMotion A 64%, B 81%, C 77% (2nd harmonic content lower)
08:37:20 ion signal!
08:28 try to load with 1.60A dispenser current. Rb-lock of cooling laser seems a bit jumpy, very sensitive to air-currents above laser (shield this with alu-plate or plexi?)
trap RF OFF and the ON, to empty trap
Ions seem to be lost
08:10Z turn on lasers.
two ions were left in the trap yesterday, lasers turned off.
Blink test doesn't look great. two or more ions
Suggested optimum 0.199 23.38 65.08 - here get A 23% B 22% C 27%
ID13727 try uMotion scan 3x3x3 grid with 10kphot/point
Try detuning 208 MHz as usual?
suggested optimum 0.253 26.89 59.51. At this point A 33%, B 23%, C 21% (all with a lot of 2nd harmonic) - make sure we have only one ion!?
ID13719 try uMotion minimization with 2x2x2 grid, 50 kphot, ranges [0.2, 0.25], [26, 27], [58, 59]
Now uMotion 100kphot A 5.5%, B 36%, C 29%
Try scanning V34diff with A-beam: A-contrast minimum at ca 58.4 V
based on bias-voltage program we are now at -6.2 0.239 26.3 62.4 V
uMotion A 55% B 19% C 21% - but large 2nd-harmonic correlation, so bias voltages require adjustment
Note detuning now ca 190 MHz (because cooling didn't work previously with 208 MHz)
Install IR-side inner shield. C-beam passes through the shield hole OKish
C-beam gives ca 50 counts/10ms (maybe focus is a bit off)
Install new C-beam. Requires 0.5" extension lens-tube, so that fiber-connector doesn't interfere with Polaris M4 mounting screws
NOTE: 1.958/2 = 0.979 - so one peak in the radial spectrum might be half the axial mode?
With 0+26 dB attenuation axial mode around 1.958 MHz
Tickler with 0+13 dB attenuation shows two peaks at 0.978 and 0.979 MHz (don't know if this is the higher or lower radial mode)
uMotion A 17%, B 15%
mostly side-to-side adjustment of B-beam gives 70 counts/10ms for B also.
ca 70 counts/10ms from A-beam, very little from B-beam
detuning to 208 MHz, ionization lasers off.
CMI FLIR IR-camera monitors dispenser, and window/chamber black tape reference areas
11:00Z: Try to load, 1.7 A, Ion! after 7 min.
10:48Z: Try to load, 1.6 A. No ion after 10 min.
7:38 UTC: start working in the lab.
Check 461 dispenser: -196.95 mA, 4.462 kOhm.+
turn off lasers - not sure why loading didn't work?? :(
12:31Z no signal... dispenser off
no signal in 8-9 min, try lowering RF-drive to 4.0 Vpp (same as earlier today when loading)
12:07Z load with 1.60A dispenser (1.54W initially... rising to 1.604W)
empty trap by turning RF off, and then back on
set detuning to 185 MHz
seems ion was lost. not sure why - higher RF-voltage might give unstable trapping parameters?
ion went dark!?
Upper radial see two peaks with 0+15 dB: 1.45822 and 1.458975 MHz
Upper radial seems easier to excite?
Matlab secular predicition: 1.436, 1.469 3.06 MHz
Monitors now 5.2 and 2.16 Vpp
Set Keysight to 4.8 Vpp - 20% higher than before
Monitors 4.4 Vpp and 1.76-1.8 Vpp
lower radial 1.179495 MHz with 0+2dB
lower radial ca 1.18 MHz. Found with 0+2dB attenuation.
radial 0+10dB, 1.218889 MHz
axial again 2.490206 MHz, 20+9 dB attenuation
axial at 2.4894 MHz, 0+24 dB attenuation
with 0+6dB get asymmetric peak (too much heating)
Coarse tickler 0+10dB att finds one peak at 1.2184 MHz
Counts at 208 MHz are 43.8/1.3 in 10ms
Also try this with the lights off in the prep-room: (3-4/10ms lower background in 10ms, so 30-40 lower in 100ms?)
Repeated detuning scan with A-beam
optimizing PMT/pinhole-position gives slightly better counts 50/10ms
at -6.2 0.239 26.3 62.4 detuning-scan with A-beam shows a better slope over the detuning-range
Counts still quite low 40/5 in 10ms
(detuning scan with 100ms gate)
V34d 62.4V slightly better giving A 1.7%
at -6.2 0.239 26.3 63V get uMotion A 11% and B 31%
Further scanning V34diff shows A-minimum around V43dif=63V
NOTE: IR-side inner shield not present. Imaging this side with the FLIR camera
A-beam PMT counts 42/6 in 10ms (lights on in the prep-room give higher dark-counts)
B-beam detuning scan.
A-beam detuning scan. For secular correlation we want a large slope at 208 MHz detuning - not so great here..
Scan range (from 208MHz) is -23 MHz to +12 MHz
leave at V34diff=45V. uMotion A 50%, B 49%
Try V34diff 44..48V - no clear minimum.
Try V34diff 48..52V - optimum might be at a lower voltage
Try sweeping V34diff 52...56 V and search for B-minimum? no clear minimum
uMotion 100kphot: A 83% in 28s, B 82% in 36s
PMT counts B-beam 34.5/1.5 in 10ms (similar to A, so maybe this is due to micromotion)
PMT counts with A-beam are 39.9/1.8 in 10ms - seems a little low (micromotion, or alignment?)
Set detuning to 208 MHz - giving much brighter image
Both A and B beams give fluorescence - so their alignment is ok
dispenser off, photoionization off.
07:45 try to load with 1.60A (1.54W initially)
intensifier + PCO camera (200 ms exposure)in use. Objective attached and camera focus to "7" on the micrometer-translator. Field of view should be large, almost see top/bottom inner electrodes in image
Latest bias voltages in ARTIQ are -6.2, 0.239, 26.3, 54 V (these probably in use now). Actual voltages on device screen -6.439 -5.961 -27.7 +80.3 V
RF on at 4.0 Vpp, giving monitors 4.4 Vpp and 1.8 Vpp with Rigol scope
Cooling laser on. Rb-signal seems to wander a lot? Optimize AOM-angle for 185 MHz detuning
Keysight siggen glitches: 4.0->4.1 Vpp when moving up, 3.6-3.5 Vpp when moving down
Setting RF to max (and testing) at 11:03 UTC
Setting RF drive to nominal 2.9 Vpp at 10:36 UTC with amplifier
Setting RF drive to nominal 10 Vpp at 10:30 UTC without amplifier
Adjusted impedance matching of resonator. Now optimal SWR = 1.01 at 14.424 MHz
Setting RF drive to nominal 10 Vpp at 10:14 UTC without amplifier
Setting RF drive to nominal 2.9 Vpp at 09:26 UTC with amplifier
Using trap drive of 1.5 Vpp:
ID13539: axial: B-beam: 28.5 dB, peaks at 0.8874 and 0.89087
ID13553: radial: B-beam: 0+2.5 dB, peaks at 0.39152 and 0.388136
ID13522: radial: B-beam: 0+1 dB, peaks at 0.4431156 and 0.445226
ID13516: radial: B-beam: 0+15 dB, peak at 0.48653
Using trap drive of 2.5 Vpp:
ID13506: axial: B-beam: 0+28.5 dB, peaks at 1.4898 and 1.4922
ID13494: radial: B-beam: 0+13 dB, peaks at 0.764045 and 0.766067
ID13490: radial: B-beam, 0+2.5 dB, peaks at 0.744638 and 0.745797
ID13479: radial: B-beam, 0+3 dB, peaks at: 0.7044925 and 0.70652.
Three peaks ??? :
Leave cooling laser and ASEs on for the night. Switch off intensifier and PMT.
Switched back to 208 MHz detuning
Bias voltages: [-6.2, 0.239, 26.3, 54]
ID13382: B-beam detuning scan
Secular: correlation method with 20 dB attenuator:
ID13446: Axial 20+17 dB attenuation: split into 2 peaks approx. 2 kHz apart!
Seculars: old method: no additional attenuation:
ID13454: axial: B-beam, 0+27 dB, with 185 MHz detuning.
ID13445: axial: B-beam, 0+28 dB, peak at 1.737692
ID13423: lower radial, B-beam 0+3 dB att. peaks at 0.826577 and 0.8282684
ID13360: upper radial, B-beam, 0+13dB att. 0.8785 MHz
ID13380: Radial scan, peaks at 0.877826 and 0.880029
V34diff: 54 V, V34common: 26.3: A: 2%, B: 2%
55.6 V V34diff: A: 70%, B: 1%
ID13332: Trying lower radial with B :
Set voltages to: -6.2, 0.239, 28.0873, 52
49 V V34diff: A: 54%, B: 81%
Search for rough umotion minimum: 52V V34diff: A: 8%, B: 67%
Install inner shield ASE side (both parts of inner B shield now installed).
Ion from Monday still in the trap!
RF drive is now 2.9 Vpp, with RF amp and 35 dB input attenuation.
Switch to Rigol scope for RF monitoring.
ID13299: Trying lower radial with B :
ID13298: Trying axial with B, two peaks? :
A 63%, B 3% at V34diff = 69 V
A 4%, B 62% at V34diff = 65 V
A 32%, B 42 % (no C-beam installed)
maybe we have too much micromotion?
axial secular seems to drift higher as measurement proceeds?
12:58Z ID13266 correlation-measurement of axial, 1.724 - 1.73 MHz, 20 + 17 dB attenuation
maybe around 1.727 MHz - not so great shaped resonance..
coarse tickler 0+20 dB att axial 1.8-2.0 MHz -> somewhere in 1.72-1.74 MHz range
upper radial ca 885.4 kHz, 0+10 dB tickler att.
coarse tickler shows lower secular 833.4 kHz.
A-beam gives ca 100counts/10ms
blink-test with 405nm (low power) seems to show only one ion.
disconnect 422nm fiber to use it for aligning the PMT-splitter. Fiber-end at PMT-position, watch for centered 422-spot with the IR-side camera
Currently no PMT-splitter installed...
coarse tickler-sweep 800 kHz - 1 MHz seems to show two resonances - OK
found A-beam alignment also (at first only B-beam gave fluorescence)
12:03 signal! from B-beam? dispenser off
11:58:30 load with Dispenser 1.7A
RF signal-gen 2.9Vpp, monitors 3.12 Vpp and 1.33 Vpp
422nm A 0.7 uW, B 1.0 uW
405nm 3.5 mW
repumper 11.4 mW
clearout 62 uW
Check 461 dispenser: -197.13 mA, R 4.462
align intensifier + pco camera to inner electrodes
connect 2nd monitor to scope. monitors 4.11 and 1.68 Vpp
remove IR-side inner shield
remove outer shield, remove mid-shield
08:13Z lights on in the lab, remove finnfoam-boards
(note summer time now, UTC+3)
13:36Z dispenser off - could not get any cloud or signal visible??
12:59Z dispenser 1.66A
lab dark, intensifier to 5V, ca 400-800 ms exposure on PCO camera
422nm detuning 185 MHz
RF generator 2.8 Vpp (+attenuation + RF PA), monitor 3.1Vpp and 1.25 Vpp
re-align camera+intensifier to electrodes
Try to trap!?
17:04Z: Install all shields, and leave RF on (will switch off after midnight). RF drive voltage: 4.0 Vpp.
14:16Z Try to load, 1.65 A
13:41Z Adjust ASE,try to load, 1.65 A (1.66W).
13:14Z Adjust A-beam, try to load, 1.65 A (1.66 W). No ion.
12:49Z Adjust A-beam, try to load, 1.65 A (1.66 W). No ion.
12:10Z Adjust A-beam, try to load, 1.65 A (1.66 W). No ion.
12:01Z: Try to load, 1.65 A (1.66 W). No ion.
11:48Z: Try to load, 1.65 A (1.66 W). No ion after 10 min.
Install intensifier and ethernet camera.
10:41:30Z: Try to load, 1.65 A (1.66 W). No ion after 10 min.
10:20Z: Try to load, 1.6 A (1.56 W), RF drive 3.9 Vpp, CH4 monitor 3.02 Vpp. No ion after 10 min.
Check power: 422 A: 1.1 uW B: 1.5 uW , 461: 144 uW , 1033: 55 uW, 1092: 10 mW
Check 461 nm dispenser: T = 4.642 kOhm, I = -197.27 mA
Added 3 dB attenuation. Now 'normal' drive voltage is 3.9 Vpp (CH4 monitor 3.02 Vpp), and high is 8.9 Vpp (CH4 monitor 6.77 Vpp). Voltage can be adjusted now between normal and high without glitching.
RF generator glitches when moving from high voltage (6.4 Vpp) to 'normal' voltage (3.2 Vpp)
Logging monitor-voltages from scope failed during the night
Resonator connection, filter board and Cu tube shield had to be removed to put thermal paste
between alumina insulator and shapal heat sink.
Significant drop in monitor voltage after re-assembly was due to resonator helix being
off center, towards the trap. After centering (tuner end opened), resonance frequency
was <14.414 MHz.
After tuning, min SWR 1.01 @ 14.424 MHz, R 50.6 Ohm, X -0.24 Ohm. Q visually as before.
Monitor voltages even higher than before. Note: old monitor very sensitive to helix position,
new monitor reflects real trap voltage better?
12:42Z set to 6.4 Vpp, which gives roughly doubled monitor voltages compared to previous night
12:37Z With 6Vpp drive get monitors of 5.71 Vpp and 2.6 Vpp
12:33Z RF PA on and connected to helres with 3Vpp drive from siggen CH1
Test the RF PA. Attenuation from siggen to PA-input: 35 dB
Previously 70 mVpp (-19 dBm) from siggen gave ca 1 and 2 MHz seculars.
With 10Vpp (+24dBm) an attenuation of -35 dB gives -11 dBm (ca 178 mVpp) - OK max voltage at PA input?
siggen 3.1Vpp, 35 dB attenuation, and RF PA seems to give similar monitor-voltages as before 2.93 Vpp and 1.33 Vpp
Temperature rise with shapal heatsink and new Cu-tube.
RF 10Vpp from signal-generator, giving ca 3 Vpp on helres monitor. Seculars not measured since last helres/cu-tube changes, probably ca 0.9/1.8 MHz as usual
Trap efficiency was now estimated from a 3D FEM simulation (KH 2021-02/03), giving:
eta_in,z = 0.7264
eta_in_r = 0.7263
eta_out,z = 0.6978
eta_out_r = 0.6976
This can be compared to the old 2D cylindrically symmetric FEM result (AW 23.6.2019 and 9.11.2019, TLi notes 13.11.2019):
eta_in = 0.75
eta_out = 0.67
Installed shapal-rod heatsink, new D42mm Cu-tube, new U-shaped filter-board (with MMCX-connector for tickler)
Black tape on RF-parts for IR imaging
measured with DMM:
V1 = bottom outer electrode = -5.215 V
V2 = top outer electrode = -4.739 V
ARTIQ settings are V12cm = -5V, V12diff = 0.239 V (?check if these are the ones last used?).
Restarting clock laser to try new comb rep-rate piezo.
PDH settings were lost. These seem to work: Freq 25 MHz, amplitude 1.8 dBm, phase 51.4deg, lock level 0.0. (ARC gain notes 2020-11-21)
Composite-VGA converter for cavity camera is fried. Camera works ok with portable mini-screen.
New 1-channel Thorlabs piezo drive for Menlo comb rep-rate lock. Seems to work (check beat-note noise and compare to old data?)
buy more tape for label-maker: Dymo D1 (12mm, 0.5")
14:45Z Open cavity doors
Arrange cavity-rack in preparation for moving clock-laser seed there
shut down lasers
ion seems lost, storage time ca 11 days
stop 405nm after 10 min - still no ion
Re-lock to Rb-line, seems ok
Try all beams on and 405 nm on, repumper + clearout on
07:58Z stop B-count program.
ion went dark at some point, 04:36 local?
17:07 start B-count logging (with B-beam)
Abort A-count logging at 17:03 local.
Results from A beam fluorescence count monitoring:
Counts and ambient temperature (110 PT100):
Correlation (XY-plot) of counts and temperature:
Leave A-beam fluorescence monitor over night. start ca 15:47Z
gate 1000 ms. Should run for 24 hours.
Datafile: 2021-03-03-counts2. Note 1000ms gate-time is accurate, but there is dead-time between each data point.
09:30Z people in the lab again, may cause temperature-oscillation?
Could be related to 422nm polarization issues also?
C-beam counts, hands-off before t=300
B-beam counts when heating the fiber. Hands-off a bit before t=200. Towards end camera-image is quite weak also
A-beam counts when slightly heating the 422laser -> ion-breadboard fiber. A-beam has polarizer, which converts polarization-rotation to power fluctuations(?)
Seems low, did voltage-changes move ion out of beam-center?
Fluorescence in 10ms: A 53.76/1.82, B 22.64/1.33, C 61.05/1.95
Try again: A 12%, B 2.9%, C 2.9%
At this point uMotion: A 42%, B 49%, C 48% - user ERROR, wrong voltages applied
Optimal voltages (Common mode, diff): [ 0.23901015 28.08731594 48.19857708], V12cm = -5 V
ID13189 uMotion scan 2x2x2
uMotion A 23%, B 29%, C 32%
Voltages now at -5 0.2347 28 50.051
ID13185 z with A-beam
ID13184 z with C-beam
ID13183 rhi with A-beam - broad resonance
ID13182 rhi with C-beam
ID13181 rlo with A-beam
ID13180 rlo with C-beam
z lorentzian fit may not be very good?
ID13179 rlo scan 935-939 kHz
ID13178 rhi scan 980-984 kHz
ID13177 B-beam z-scan with 20dB, fit finally works. peak height ca 40%
B-beam tickler z-secular with 17 dB gives 60% high peak - and fit fails - maybe more attenuation needed
lorentzian fit to z-secular with A-beam seems difficult..
ID13172 try z-secular with A beam again, 27 dB attenuation - but fit fails again.
ion still stored
No experiments during the night
Leave B-beam on.
Summary: Seculars with -5 0.2347 28 50.051. A-beam
axial/radial ratios ??
z-scans where lorentzian fit failed:
Note these are also asymmetric/hysteretic. we sweep the tickler from low to high frequencies
ID13170 try again with 23 dB DDS attenuation (up from 17 dB) - fit fails for the 3rd time
ID13169 narrower range 1843-1844.5 kHz - too strong tickler gives blurry image? may not give lorentzian resonance? - fit fails again!??
ID13168 fit fails :(
ID13168 z scan 1842-1846 kHz - peak up to 50% with blurry camera image. As predicted A cools z-secular badly, so resonance is narrow
z search 1840-1860 kHz - found at 1844!
old z(-6.1)=1876 kHz and z(-6.2)=1879 kHz, so expect z(-5) at -33kHz around 1843?
ID13166 rhi scan 980-984 kHz - as before, rhi resonance is broad with A-beam
rhi search 970-990 kHz - found around 982 kHz
previous rhi was 954 kHz, so expect it around 985 kHz maybe?
ID13164 rlo scan 935-939 kHz
930-940 kHz - peak at ca 936.5 kHz - ca +31.5 kHz from last value with V12cm=-6.1V
search 920-930 kHz - no peak
search 910-920 kHz - no peak
tickler search 900-910 kHz (previous rlo was 905 kHz) - no peak found. 0.1V change in V12cm moved r-secular 3 kHz yesterday, so expect 30-33 kHz now?
Using A-beam for these tickler measurements
Go back to -5 0.2347 28 50.051 - uMotions A 2%, B 9.9%, C 13%
At this point: A 46% !!, B 2.3%, C 16% - repeat A 46%
Optimal voltages (Common mode, diff): [ 0.19201393 27.06730014 49.21492668]
ID13146 uMotion scan 2x2x2 grid
at -5 0.2347 28 50.051 get uMotion A 2.8%, B 11.5%, C 12%
Try scanning V34cm, A-minimum at 28 V
Now uMotion A 38%, B 27%, C 31%
Try reducing V12cm to -5 V. Other voltages as before 0.2347 26.7769 50.051
ID13112 uMotion scan with 5x5x5 grid
Optimal voltages (Common mode, diff): [ 0.23465246 26.7769039 50.05097994], V12cm=-6.1 V
At this point A 2.9% in 16.3s, B 2.2% in 13.4s, C 2.6% in 21.9s
ID13106 uMotion scan 2x2x2
scan gives Optimal voltages (Common mode, diff): [ 0.18054628 26.89479311 50.30482989]
At this point get A 3.9% in 20.5s, B 12% in 16s, C 4.4% in 19.8s
at -6.1 0.2277 26.3891 51.2 get uMotion: A 6.7%, B 22%, C 25%
Sweep V34 diff: A-min at 51.2 V
uMotion (100kphot): A 70%, B 71%, C 72%
Fluorescence (10ms): A 74.27/1.48, B 114.38/1.34, C 66.76/1.98 (lab lights are off)
after ca. 1 min of 405nm irradiation - ion re-appears!?
re-lock to Rb-line
08:54Z ion seems dark
After a stable period, lab temperature oscillates again?? why?
16:52 leave lab.
Leave A beam on. No experiments during the night
Maybe Rb-lock was OK?
Now A 97.6/2.13, B 38.85/1.62, C 74.11/1.93
Try re-locking to Rb-line
repeat A 93/1.9, B 98/1.47, C 72/1.66
Fluorescence (10ms): A 86/1.6, B 90.67/1.43, C 64.57/1.61
What if we try to minimize micromotion at V12cm = -6.1 V? Previosly V34cm change was -1.3 V34cm/V12cm
ID13058 uMotion scan 2x2x2 grid
B gives really low fluorescence suddenly? 50 seconds to measure 50 kphotons
Optimal voltages (Common mode, diff): [ 0.22771198 26.38907221 56.11298177] (note previous V34cm was 26.18V so this is now higher?)
At this point get 100kphot uMotion A 7%, B 2.1%, C 1% (ca 21-22s measurement time with A and C, but with B 102s)
Tickler measurements now with A-beam
ID13064 z-scan, with -6.2V was at 1876.4 kHz (B-beam in use, tickler program sets beam!?)
A-beam fluorescence also quite weak, laser Rb-lock possibly jumped? (to next Rb-line +60 MHz more detuned?)
Previous peak-height was just over 30%, now get 60% because of less efficient cooling/damping?
ID13065 rhi scan 952-956 kHz - previously at 953.9 kHz. Now with A-beam, giving very broad resonance (A cools rhi well?)
ID13066 rlo scan 903-907 kHz - previously at 905.1 kHz. Note FWHM is almost 3-fold narrower
ID13053 better z-scan
z scan 1870-1880 kHz - found just above 1876 kHz
rhi scan 951-961 kHz - found at 954 kHz
rlo shifted up by +3 kHz, around 905 kHz
ID13047 900-904 kHz rlo scan
Try V12cm = -6.1 V (no change in other voltages)
uMotion A 8.6%, B 19%, C 21%
Tickler measurements with B-beam
axial/radial ratios 1.967 and 2.073
axial/radial ratios 1.976 abd 2.083
with V12cm = -6.2 V , seculars today are:
ID13045 rlo scan
ID13044 search for rlo 897-907 kHz - 1% higher than Friday would be 902.8 kHz - found at just over 902 kHz
ID13043 scan 949-953 kHz
ID13041 rhi search 940-950 kHz. same scaling would give 946.7 kHz - found peak at 951 kHz
ID13040 1876-1882 scan.
ca 0.55% higher than on Friday
ID13039 search 1870-1880 kHz. 10 kHz/30 pts = 333 Hz step - peak around 1879
Search for z-secular 1860-1870 kHz
at this point A 4.0%, B 2.2%, C 1.2%
Optimal voltages (Common mode, diff): [ 0.23807795 26.17776091 57.07942005], V12cm = -6.2
ID13032 uMotion search 2x2x2 grid
ca 57V seems better, uMotion A 13%, B 40%, C 43%
previous V34diff was 49.8V
uMotion A 75%, B 74%, C 72%
Fluorescence A 122.37/1.85, B 119.98/1.83, C 82.64/1.95
install inner shield camera side, and mid shield
With old objective, when focused to Z=7mm, gap to vacuum window is ca 5mm
However image with new objective not so great :( so go back to old objective
Try new objective: focus is closer to Z=6mm, and new image ca half screen to the right of old objective (as viewed in ASI camera software)
For old imaging objective XYZ Z=7mm translator X=55, Y= 36<->0
got C-beam aligned, via electrode diffraction pattern (only inner shield installed)
aligning C-beam with mid-shield installed seems difficult
ca 12-12:30Z (14 local) work in the lab (check how temperature responds?)
Ion stored for ca 1 week now
07:13:00Z Stop collision monitor.
16:03Z start collision monitor. B-beam, detuning 208 MHz, fast 3 ms gate. Threshold set as 40 bright 1 dark (artiq sets geometric mean of these?)
clearout is on
NOTE: scripts assume 10ms gate time, use last column (machine unit) for analysis
B fluorescence in 10ms: 157/1.14, in 5 ms: 71.97/0.44, in 3 ms 44.27/0.42 - lab lights are off
Set RF drive to 10 Vpp
uMotion A 2.8%, B 2.6%
Summary of secular results (Compare to 2020-12-09)
Z-frequency 94.95% of Dec value, with voltage-slope 93.09% of Dec value
rhi 95.55% of Dec value, with voltage-slope 93.37% of Dec value
rlo 95.05% of Dec value, with voltage-slope 93.90% of Dec value
Set RF drive to 9.0 VPP.
z maybe 1681 kHz? found at 1675 kHz
rhi maybe 836 kHz? found at 852 kHz
rlo maybe 800 kHz?
Set RF drive to 9.5 VPP. Expect 95% of 10Vpp seculars?
z ca 1775 kHz? actual resonance ca 1771 (94.8%)
rhi maybe 888 kHz? found at 897 kHz
rlo maybe 843 kHz? found at 847 kHz
lower radial slightly lower than expected, below 894 kHz
Try to find r_hi 934-940 kHz
Behold, artiq autogenerated figure output from the correlation tickler secular sweep program
(these measured with the usual 10Vpp RF drive, B-beam, fixed 20 dB physical atteunator on dds-output, software attenuation adds to this)
ID12911 same scan, modify lorentzian fit to match matlab 'lorentzfit' functional form. resonance at 1868.59
try 1866-1870 kHz - 22% peak at 1868.6 kHz
ID12909 secular z-sweep, B-beam, 1862-1866 kHz
RF monitor 3.06 Vpp
uMotion A 1.8%, B 3.4% (C not aligned to ion...)
Set -6.2 0.2445 26.9255 49.8
scan V34diff, A-optimum at 49.8V
uMotion: A 37%, B 28%
C-beam alginment seems lost...
Fluorescence countsin 10ms are: A 97.8/1, B 106.4/1, C 42.2/1.5 (lab lights are off)
Detuning back to 208 MHz
11:59 stop collision monitor
11:58Z ion still stored. fluorescence quite weak
lights off, leave lab.
16:49Z start collision-monitor ID12879, DB 2021-02-25-run1
B beam, Counts / 10 ms: Bright: 35, dark: 3.
Detuning 190 MHz. Clearout on
Set detuning to 190 MHz, giving ca 30-40 counts in 10ms
Use B-beam (may have polarization issues? correlates with lab-temperature?)
uMotion A 4.8%, B 5%, C 2.8%
Set voltages -6.2 0.2445 26.9255 51.6
Scan V34diff, A-minimum at 51.6V
uMotion A 12%, B 11%, C 11%
Counts A 141/4, B 62/3, C 103/4
Try I1 = -8.6056, I2 = -1.0049, I3 = 7.0059, giving abs(B) =
ID12845 peak search with 300us
C1+ 18%, C1- 32%
C1 B estimate = 2.055935940628926
C1 Linecenter = 75.97028908312774
NOTE: B-beam fluorescence down to 53.57/3.17, indicating maybe Rb-unlock? (or polarization-effect giving unbalanced C+ and C- peaks?)
Try I1 = -8.6056, I2 = -3.0049, I3 = 7.0059, giving abs(B) = 2.2646, 2.2526, and uT
ID12842 peak search, 300us. Should now compute line-center and B after each pair
C1+ 24%, C1- 64% (!)
C1 Linecenter = 75.97024614563541
C1 B estimate = 2.2646399566052304 uT
C2+ 24%, C2- 14%
C2 Linecenter = 75.97025830181504
C2 B estimate = 2.2525933966917435
ID12840 peak-search with 300us, and new B-estimate 2.119 uT
C1+ now lower 29%, and C1- higher 53%
C1 Linecenter = 75.97016610304007 MHz
C1 B estimate = 2.1102318715100816 uT
C2+ 45%, C2- 19%
re-initialize search with C1 line-center and B-estimate
ID12838 peak search, 300us probe, with B=2uT estimate
C1 Linecenter = 75.97022862199493 MHz
C1 B estimate = 2.119082202980859 uT
Set coil currents I1 = -8.6056, I2 = -2.0049, I3 = 7.0059
TLi notes on Magnetic Field (2021-02-15)
For B-beam, using A-matrix and B0 residual field from 2021-01-21.
Should give equal C1/C2/C4
abs(B) I1 I2 I3
2 -8.6056 -2.0049 7.0059
3 -10.672 -2.5016 9.9934
4 -12.739 -2.9984 12.981
ID12835: Peak search without OD filters, B beam. clock_probe = 300 us
C1 Linecenter = 75.97019812711028
C1 B estimate = 1.046407999323715
ID12831: Rabi flopping on B beam, no OD filters. tau_pi ~ 300 us.
Problem with comb rep-rate lock
The closed-loop rep-rate error signal started to show glitches, with about 400-500ms period
This caused all SDR counter-data to be flagged invalid
A Thorlabs MDT693 Piezo-driver was tested - and showed no glitches
Also the cable from the rep-rate lock-box to the piezo driver was replaced!? (suspect looking ends??)
With the new cable, also the old Piezo driver seems to work? (since about 15:25 local, at least for about 10 minutes)
Leave it running, to see if problem re-appears
however the problem re-appears after a 10min warm-up period!!??
14:08Z Go back to the Thorlabs driver. This gives glitch-free lock and valid SDR-counter data
after lunch fluorescence seems weak. Re-locking to Rb-lines improves counts 3-fold
At this point get: A 1.3% in 11s, B 0.9% in 15s, C 0.8% in 15s
Optimal voltages (Common mode, diff): [ 0.24445048 26.92545019 51.96431002]
ID12794 uMotion search 4x4x4 grid
ID12793 repeat - [2.90072747e-02 2.89735223e-04 1.86382763e+00 2.69222756e-04], Lorentzian FWHM = 2 g = 538.4 Hz
ID12792 scan 1862-1866 - 30% just below 1864 kHz - Lorentzian fit: [2.33235839e-02 3.39371929e-04 1.86379851e+00 3.83917947e-04] (constant, amplitude, f_center, gamma)
ID12781 search for z-resonance at higher frequencies 1856-1866 kHz
ID12790 tickler-correlation, z - no resonance
These can be cropped to show only the ion by:
convert 2021-02-25_A_15s.jpg -crop 150x150+1050+120 A4.jpg
Ion images with different beam combinations:
15s stack of 1s exposures. Old imaging objective, ASI camera, high-gain, no binning (1936x1096 pixels). Pixel size is 2.9 um x 2.9 um. Estimated magnification is 30x
Fluorescence levels (10ms): A 132.1/3.8, B 127.6/4.2, C 75.8/4.0
Multiple beams: AB 117.7/4.2, BC 154.4/4.3, AC 143.0/4.6, ABC 181.85/4.6
uMotion A 2%, B 0.9%, C 6.1%
Set -6.2 0.23 26.8378 52.4
sweeping V34diff shows A-minimum around 52.5V, finer resolution sweep gives 52.4V
uMotion A 47%, B 45%, C 49%
Voltages now: -6.2 0.23 26.8378 55
Increased AC-setting at 30 Hz gives slightly higher amplitude oscillations, compared to weekend setting of 25 Hz
Weekend had less heat load in the lab
07:37 stop collision monitor
07:36Z ion still stored
ID12763: Start at 14:48:52 UTZ: Collision monitor with clearout on, with B beam. Tag: 2021-02-24-run1
Counts / 10 ms: Bright: 121, dark: 3.
usual detuning for operation, 208 MHz (check?)
AC setting now increased to 30 Hz
Sweeping V34_diff gives new 'optimum':
uMotion A 4.1%, B 1.2%, C 6.1% at [-6.2, 0.23, 26.8378, 55] (down 1.6V from morning value of 56.6V)
uMotion A 30% in 12s, B 30%, C 36% at [-6.2, 0.23, 26.8378, 56.6]
re-lock to Rb-line -> fluorescence restored!
14:33 ion seems dark/gone?
ID12745 test2 of secular scan Lorentzian fit: [2.00147896e-02 2.50019692e-04 1.85449329e+00 3.99132412e-04]
ID12744 test of secular scan with constant+Lorentzian: [2.17395264e-02 2.64289160e-04 1.85434871e+00 4.59689160e-04] (constant, amplitude, f_center, gamma)
Compare this to Matlab fit to same data points?
Try the third (lowest) piezo-channel of the driver. This again initially gives ok comb-lock and beat-note data - but after 1-2 minutes glitches appear again!
rebooting the rep-rate piezo driver seems to remove the glitch - however the problem appears after a few minutes again
Comb rep-reate lock shows 400ms heart-beat error, can't determine line-center from comb data....
ID12736 carrier C1 peaks with 600us probe
+C1 ? see two peaks?
Now optical carrier and side-bands
ID12739 z 1.848 - 1.856 MHz, Lorentzian fit: [3.72032165e-04 1.85437750e+00 6.67644320e-04]
ID12738 r_hi 937-943 kHz,
ID12737 r_lo 885-892 kHz,
(these over lunch). more points and 100 kphotons per point
ID12734 r_lo 885 to 891 kHz, 7 dB - 22% peak at ca 889.6 kHz - Lorentzian fit: [3.85734824e-04 8.89606990e-01 5.82098050e-04]
ID12733 r_hi 937 to 943 kHz, 7 dB - 16% peak at ca 929.1 kHz - Lorentzian fit: [2.93961269e-04 9.39172793e-01 7.09967596e-04]
ID12731 z 1849 to 1855 kHz, 17dB - 22% peak at ca 1852.1 kHz - Lorentzian fit: [3.88708773e-04 1.85215658e+00 6.40710257e-04] (curret ARTIQ fit doesn't have constant background term)
ID12730 r_lo 891 to 897 kHz - no clear peak
ID12729 secular correlation search 943 to 948 kHz, 7 dB att - no clear peak
B-beam in use, 50 kphotons/point
RF monitor ca 3.040 Vpp (slightly lower than y-day?)
Fluorescence levels (10ms): A 139.2/1.56, B 126.46/1.1, C 110.8 / 1.9
At -6.2 0.23 26.8378 56.6 get: A 1.4%, B 3.1%, C 8%
Scanning V34diff shows A-minimum at 56.6 (scanning only V34diff seems to be a quick way to find the optimum from day to day!?)
Maybe V34diff has settled to a lower value again?
uMotion 100kphot: A 74% in 9s, B 73% in 9s, 75% in 9s
Voltages -6.2 0.23 26.8378 63.3441
08:23Z ion still stored. secular sideband measurements completer over night
15:21Z leave lab
ID12705 z, search21
ID12704 r_hi, search20
ID12702 B-beam r_lo, search19
ID12701 switch to B-beam
ID12700 z, 2 kHz AOM span, search18
ID12699 r_hi, 2 kHz AOM span, search17
ID12698 r_lo sidebands, 2 kHz AOM span, search16
ID12697 z sidebands, search15
ID12696 r_hi sidebands, search14
ID12695 A-beam, r_lo sidebands, wider 1 kHz AOM span, 16 Hz AOM step, C1+ and C1-, 5 ms probe (no OD filter). DB: search13
63 points per peak
C1+ red appears far right in span?
ID12690 r_lo sidebands
uMotion A 66%, B 64%, C (very weak)
ID12676: Photon correlation secular sweep for higher radial: 20 + 7 dB att. Peak at approx. 0.9447 MHz.
ID12675: Photon correlation secular sweep for lower radial: 20 + 7 dB att. Peak at approx. 0.8942586 MHz.
ID12674: Photon correlation secular sweep for axial: 20 + 17 dB att. Peak at approx. 1.8524828 MHz.
ID12673: Secular sweep with 0dB + 22 dB: Lower radial ~ 0.89449 MHz, upper radial ~ 0.94558 MHz
ID12672: Secular sweep with 0dB + 20 dB: Axial: peak at approx. 1.8528 MHz
ID12665: Carrier search for C1, tau_pi = 600 us. 2021-02-23-search11-
ID12659: Carrier search for C1, tau_pi = 600 us. 2021-02-23-search7-
12657: C1+ red low radial?
12658: Rabi on previous. Tau_pi maybe 5000 us
ID12654: Secular search for C1, 2021-02-23-search4-
Tests with objective, umotion again high! V34diff needs 10 more Volts to get uMotion acceptable (sharp camera image)
New imaging objevtive did not show a good image... further testing required
Here uMotion A 0.3%, B 1.1%, C 2.5%
Back to -6.2 0.2358 26.8378 48.3441
ID12569 tickler 20+17 dB, 1864 - 1868 kHz, 30 points - no peak..
At these voltages get 100kphot: A 9.2% in 9s, B 3.4% in 10s, C 5% in 12s
Optimal voltages (Common mode, diff): [ 0.2358202 26.83783696 48.34408991]
Test how trap parameters change at V12cm of -6.0V (previous setting was -6.2V 0.2289 26.8187 48.4554)
ID12562 uMotion scan 4x4x4 (lunch) - USER ERROR: this scan at V12cm=-6.2V as previously....
Note matlab lorentzian fit is variant '3c' of lorentzfit:
L3C(X) = P1./((X - P2).^2 + P3) + C
FWHM is computed as:
FWHM = 2 * sqrt( P3 )
fine-resolution correlation measurement of seculars (now with B-beam)
ID12555 scan 890-893 kHz, 30 points, 20 + 7 dB attenuation - peak height up to 40% contrast
ID12556 scan 937-941 kHz, 30 points, 20 + 7 dB attenuation - peak height 14 % (upper radial is harder to excite with tickler - as before)
ID12557 scan 1.864 - 1.868 MHz, 20 + 17 dB attenuation - peak height 30%
RF monitor during these 3.045 - 3.048 Vpp
search for seculars (A beam)
tickler 0 dB + 15 dB attenuation - no clear fluorescence change. search 0.9 to 1.1 MHz
0 dB + 10 dB attenuation
0 dB + 5 dB attenuation. one mode at 939.6 kHz
0 dB + 2 dB attenuation: lower at 891.4 kHz, upper at 939 kHz
0 dB + 20 dB for axial, 1.8-2.0 scan, at 1.86644 MHz
500kphot A 1.8% in 41s, B 1.3% in 49s, C 3.7% in 61s
100kphot A 1.3% in 9s, B 0.9% in 10s, C 2.9% in 13s
found Optimal voltages (Common mode, diff): [ 0.22887765 26.81872373 48.45535893]
ID12530 2x2x2 uMotion search
Voltages at -6.2 0.2832 27 47
uMotion 100kphot: A 17.6% in 9s, B 17.3% in 10s, C 32% in 12s
Fluorescence levels (10 ms): A 160.7 / 2.14, B 133.59 / 1.94, C 98.25 / 2.04
back to 208 MHz detuning
08:31Z stop collision monitor
ID12517: Start at 15:00 UTZ: Collision monitor with clearout on, with all beams, detuning = 190 MHz. Tag: 2021-02-22-run1
Counts / 10 ms: Bright: 79, dark: 3.
reducing RF-voltage to 4.1Vpp doesn't show two ions - so this is probably a single ion
These are higher than previously during collision-monitoring/clock-runs (ca 60-70 bright counts?)
10ms fluorescence levels A 166.5/2.25, B 145.54/1.94, C 109.88/2.54
all beams ABC show fuzzy image, any single beam sharp image?
uMotion A 14% in 9s, B 9% in 11s, C 22% in 11s
14:39:30 signal (in 3 min)
bias voltages -6.2 0.2832 27 47
RF back to 10Vpp
14:36:30Z try load with 1.60A (1.54W)
changing V12cm from -6.2V to 0V in a single step threw the ions away
reduce RF to 4.0V increases ion-separation clearly on camera
14:30:05Z signal, blurry at first, then two separated ions
14:23Z try to load with 1.65A. detuning 185 MHz, 405nm was up to 7mW - reduce to 3 mW
lowering RF-voltage shows up to three ions! (saved some images)
detuning to 208 MHz
14:16Z signal! fuzzy image
back to voltages -6.2 0.2832 27 47
14:11:15Z load again, 1.7A, detuning to 185 MHz
adjusting RF-voltage down -> maybe two ions in the trap? lost both at RF-glitch
bright ion seems to jump between two positions?? (when switching beams) do we have two ions trapped?
-6.2 0.2831 27 42.4 gives A 15%, B 2.2%
-6.2 0.2832 27 Scan[41 43] B
-6.2 0.2832 27 Scan[38, 43] gives A-optimum at 42V
A 65%, B 68%, C 70%
Optimal voltages (Common mode, diff): [ 0.2778 26.13 47.58]
ID12458 2x2x2 uMotion scan
[0.25, 0.32] [26, 27.5] [46.5 48.5](too large?)
uMotion A 70% in 8s, B 64% in 11s, C 70% in 11s
Voltages -6.2 0.2832 27 47
Lowering trap RF to 5.7Vpp gives stable high fluorescence, 1200 couns / 50ms all beams
try different V34diff voltages, but no clear effect
seem to have something trapped, but fluorescence comes and goes
13:39:35 signal, but fuzzy image of ion? gone quickly?
13:35:30 try to load, 1.7A, detuning to 185 MHz
13:32:15 signal! but lost quickly?
13:26:15 load with dispenser 1.7A. test loading with 208 MHz detuning?!
back to volgaes
ion lost during uMotion search?
ID12410 seems to have finished ok?
fit fails because V34diff range was [46.5, 46.5] (no scan at all)
not so reasonable Optimal voltages (Common mode, diff): [ 1.80241039e-15 9.93688436e-17 -2.15053763e-02]
ID12410: NPL search 2x2x2.
Optimize beams and 422 waveplate.
0.2832 27 47 gives A 14%, B 9%, C 21%
0.2832 28.1299 47 gives A 28%, B 26%, C 40%
0.2832 28.1299 46 47 gives A 28%, B 26%, C 40%
ID12380 scan V34diff vs A-contrast. Minimum at ca 47V
uMotion 100kphot A 76% in 9.9s, B 70% in 16s, 71% in 15s
ASE also required re-aligning, got >2-fold more fluorescence (all beams) by aligning ASE
uMotion 100kphot: A 81% in 22s, B 76% in 27s, C 75% in 29s
B > 200 counts in 50ms, C was already quite well aligned, up to 280 / 50 ms afer adjustment
re-aligning beams is required, since V34dif changed som much? 3-fold more A-fluorescece ca 300counts in 50ms
check Rb-lock, locking to the right line gives better fluorescence
422 power only 5uW on breadboard before A-fiber (this usually 10-15 uW), adjustment of 422 breadboard fiber-coupling gives ca 9uW
lowering V34diff to 53V gives the usual diffraction-pattern image of ion
detuning to 208MHz
12:05:50 some weak signal? (ca 3 min loading time - good!?)
12:02:45 load with dispenser 1.7A
405nm adjusted to 3.7 mW
422nm 185 MHz detuning
461nm dispenser test, Ractual 4.642 kOhm, -197.32 mA, test against 5A dispenser ca 2 min to see fluorescence.
RF monitor 3.045 Vpp
Voltages to last known values: -6.2, 0.2832, 28.1299, 85.6211
camera focus back to last value
11:52Z install mid-shield, and imaging objective
08:20Z turn on stuff to prepare for trapping again
Menlo oscillator was on continuously.
A/C VFD was at 25 Hz
With most stuff turned off on Friday, see strange oscillations in lab temperature
Leave the RF siggen on/off program on
Clamp down helres trap-side
Trim 3mm off the helres-mount clamp
Before final fine-tuning, vswr with newer rigexpert analyzer:
Helical resonator min SWR was 1.12 @ 14.422, which probably explains
lower seculars (qz). HR moved a few mm to the left and towards the trap
and rotated ~0.5…1 deg CW seen from tuner end to center helix and to
align shield adapter with trap Cu pipe.
Also 0.1 mm soft Cu shield replaced by stiffer 0.2mm Cu. This increased
Vmon (helix moved closer to antenna) and required tuning and impedance
After: min SWR = 1.01 @ 14.424 MHz and Vmon = 2.995 Vpp (averaging 32).
R = 50.2 Ohm, X = 0.15 Ohm, Q ~ 800 as before.
NOTE: adjusted A/C to 25 Hz - to possibly reduce temperature swings?
detuning to 185 MHz, 461nm and 405nm lasers on - but no ion?
08:08Z Ion seems lost? Rb-lock was unlocked (large temperature-swings in lab!?), re-lock, but this doesn't restore fluorescence
NPL uMotion program had aborted, due to ARTIQ error? (because of Rb-unlock, or ion loss?)
Leave 8x8x8 uMotion scan over night
ID12326 ranges [0.1 0.4] [26 30] [80 90]
16:09Z uMotion 100kphot: A 34% in 8.7s, B 31% in 14s, C 24% in 9.7s
The required V34diff might be drifting down during the day?
TLi computed from these: qz=0.3643, a=(-0.00038, 0.0016, -0.0012)
These measured with 10Vpp drive, ca 2.70 Vpp helres monitor, and fixed -6.2 V12cm
Reults to be plotted + Lorentz-fit: axial ID12318, r-lo ID12320, r-hi ID12322
upper radial with photon correlation. B-beam
ID12321 978 to 982 kHz, 20+7 dB att, 10 kphotons
ID12322 with 50 kphotons/point
lower radial mode with photon correlation. B-beam
ID12319 923 to 927 kHz, 20+10 dB att, 10 kphotons
ID12320 20+7 dB att, 50 kphotons
fine-resolution secular scan using photon correlation. B-beam. ( 2020-12-09 results use 13 dB for radials and 21-23 dB for axial)
search areas at 925, 980, and 1892.5 kHz.
use the physical 20dB attenuator
ID12312 1.89 to 1.895 MHz in 30 steps, B-beam, 10 kphotons per point, 20+22 dB attenuation
ID12313 20+26 dB attenuation, 20 kphotons, - too much attenuation, peak height low
ID12314 20+23 dB attenuation - 0.2 peak height
ID12316 20+20 dB att, 10 nbins,
ID12317 20+17 dB att, 0.3 peak height ok? but peak looks a bit noisy?
ID12318 20+17 dB att, 50 kphotons/point (ca 6-7 s/point)
Coarse search for secular frequency.
20dB physical attenuator on tickler-DDS. 0 dB software attenuation
resonance at ca 1.892 - 1.894 MHz ca 96% of 2020-12-09 result (lower than before, 2020-12-09 result with 10Vpp drive is 1.968 MHz)
Try 6 dB sw attenuation, resonance at 1.892-1.893 MHz.
(2020-12-09 results have radial modes at 985 and 940 kHz, or 50% and 47.8% of axial)
This would predict current radial modes at 947 and 904 kHz
tickle 0.85 - 1.0 MHz, 6dB, 400 points, but no clear signal (50ms PMT gate).
Try 0dB , also try 100ms gate, remove 20dB physical attenuator and use 15 dB sw attenuation
with 10dB sw-attenuation, one mode at 925 kHz?
try 5dB attenuation, try 0 dB attenuation -> this reveals upper mode at 980 kHz!
B fluorescence 75.7 / 3.96 in 10ms (mid-shield installed, back lab lights on)
B intensity still changes with polarization! Adjust waveplate again.
uMotions: A: 37%, B: 33%, C: 31%
ID12264: Peak search, B-beam, no OD-filters. 2021-02-17-search6-.
C1 linecenter: 75 971 017.0 Hz
C1 B estimate: 1.2443346 uT
ID12258: B-beam Rabi flopping, tau_pi = 0.6 ms:
Adjusting waveplate on 422 breadboard reduces fluorescence change to negligible levels:
Heating B or C fibers by hand seems to give >50% fluorescence change
ID12245 peak search A-beam 350us, no OD filters
C1+ ca 75 974 7, height ca 40
C1- 75 967 5, height 60
Found C1+, rabiflop shows maube 350us tau_pi with A-beam?
seems Zeeman AOM diffraction-order was wrong. Switching to right one gives quantum-jumps as expected.
ID12238 spectrum-scan with 1ms probe, no OD-filter, A-beam
beat-note and excel-sheet suggest line-center around 75.971 150 MHz
Coil currents (-8, 0, 0) mA - same as previously?
A-beam fluroescence 97.4 / 2.4 bright/dark in 10ms
Try to search for the clock-transition?
At this point get: A 3.8% in 13s, B 2.9% in 10s, C 2.4% in 19s
Optimal voltages (Common mode, diff): [ 0.2832 28.1299 85.6211]
ID12223 Try 2x2x2 uMotion scan?
V0 - 0.33VA = 0.2610 27.8855 87.2692, here get A 27.5 %, B 29.8%, C 36.3% - ??
maybe -sign is right?
V0 + 0.33 VA = 0.2612 28.1501 87.5210 here get A 39%, B 29.6%, C 35.5 %
Try moving along voltage-vectors to zero uMotion?
uMotion 100kphot: A 33.8% in 15s, B 29.4% in 12s, C 36.7% in 13s
detuning constant at 208 MHz
Rb-signal has a lot of DC-offset, so this may give some extra detuning?
Re-lock to Rb-line, this gives ca 5-fold more fluorescence and brighter camera image
uMotion 100kphot A 9.7% in 91 seconds!?, B 9.0% in 95s, 11% in 85s
Fluorescence level seems low? camera image weak..
Comb f0 unlocked due to low SNR. re-locked
ID12186: 4x4x4 NPL search around previous value.
Result: [ 0.26114758 28.01782364 87.39507497]
uMotions at optimum: A: 3.6%, B: 5.5% C: 9.2%
Check that the columns of inv(Cfit) are voltage-vectors that change only one contrast!?
% Micromotion contrast model
% S = C (V - V0)
% S has [A, B, C] contrasts
% V and V0 have voltages [V12dif, V34cm, V34dif]
% V12cm is constant -6.2 V in recent experiments
-0.9303 -2.0943 -0.4214
-0.2217 0.1219 -0.1284
-0.1447 -0.1542 0.1619
inv(-1*Cfit) = % sign is arbitrary..
-0.0004 2.9735 2.3561
0.4009 -1.5572 -0.1910
0.3815 1.1746 -4.2519
Cfit * [-0.0004 0.4009 0.3815]' = [-1 0 0]'
moving along first column-vector of inv(Cfit) changes A by 100% and leaves B and C unchanged.
Cfit * [2.9735 -1.5572 1.1746]' = [0 -1 0]'
Cfit * [2.3561 -0.1910 -4.2519]' = [0 0 -1]'
Manually: 29%, 30%, 31% at [0.25, 29.4, 87]
Manually: A: 53%, B: 1%, C: 21% at [0.3306, 27, 86]
ID12094 leave uMotion scan over night
8x8x8 grid. voltage ranges [0.5, 1.5], [20, 30], [95, 105]
Result: (Common mode, diff): [ 3.99139299 12.57404019 85.31937744]
1 scanned 25.8, 98: A-optimal (A 1.9%, B 74%, C 40%)
1 25 99 seems roughly A-optimal (A 7% B 65% C 8.4%)
go back to the AC-optimal 1.0 25 100: A 9.7%, B blur 62%, C 7.1% - so AC-optimal point is roughly repeatable
but scanning V34diff or V34cm around here gives more or less the same flat A-contrast values
at 0.3 27 48 get A 21% B 29% C 26%
at 1.0 27 50 get A 22%, B 32%, C 26% - images look blurry, but this may be due to focusing of camera? or beam-alignment needed?
These voltages are far from the 2021-02-03 values [0.28, 27.0, 48.0]...
B-blur better at 1.0 32 102.5 (A blur 57%, B 77%, C 58%)
C-nubunyn at 1.0 28, scanned 102.5 (A 78%, B blur 64%, C 4.1%)
B-blur is sideways, so change C34cm? at 1.0 28 100.2 get (A 71% ,B 78%, C 46%) all images sharp
C-minimum at 1.0, 25, scanned 100.2 (A 1.9% !, B blur 63%, C 0.3%) first found point that is AC-optimal
B-blur better at 1.0, 25, 98.2 (A 45%, B 77%, C 45%)
C-minimum at 1.0, 22, scanned 98.2 (A 75%, B 63% blurry sideways, C 5.2%)
A-blur reduced at 1.0, 22, 96.2 (A 75%, B 75%, C 36%)
C-contrast minimum at 0.6, 22, scanned 96.2 (A 59% blurry, B 66%, C 1.3 %)
C-contrast minimum at 0.6, 23, scanned 97.4 (A 77%, B 68%, C 3.7%)
A-blur reduced at 0.6, 23, 96.44 (A 75%, B 61%, C 20% )
C-contrast minimum at 0.3543, 23 and scanned 96.44 (A 61,6% blurry vertical image!, B 19.5%, C 4.6%)
C-contrast minimum at 0.3543, 24 and scanned 97.33 (A 75,6%, 19.4 %)
C-contrast minimum at 0.3543, 23.4963 and scanned 96.88
Manually: A: 75%, B: 27%, C: 13% at [0.3306, 23.455, 97.98685]
Further searching gives 0.3, 29, 93 V. This gives reasonably sharp ion image with clear diffraction rings
Here uMotion A 75%, B 77%, C 76% - correlation histograms roughly sine-shaped
Manually found better voltage combination for A and B (harmonic, but high contrast): [0.9, 28, 91]
detuning changes fluorescence very little - indicating lots of uMotion?
detuning back to 208 MHz (from loading value of 185 MHz)
similar blurred image as before - showing blurred diffraction-rings typical of the imaging objective
blocking some beam briefly gives a cooled ion
12:48:05Z signal (axial mode? not cooled?)
12:43:15Z load with 1.7A dispenser
12:40 throw away old ion(s)
lots of uMotion? maybe have 2 ions?
ID11722: Leave NPL umotion search over lunch.
Ion image blurry, lot of micromotion?
Blink test shows only 2 levels, single ion?
Readjust PMT splitter.
try to align all beams, so we get bright fluorescence
422 detuning to 208 MHz
07:44 signal! quite far right from where camera center was aligned?
07:43 load with 1.7A dispenser
align A-beam with help of output diffraction pattern
remove PMT-splitter, align camera to electrodes
07:31Z dispenser off, no clear signal
07:18Z try to load, 1.7A dispenser.
Adjust 461 current.
06:55Z try to load, 1.7A dispenser. No ion after 15 min.
Adjust A-beam position
Cavity drift with minus-K floating now seems better? This day Lunch 12-13 and team-development 13-15 - chronograf figure with local time on X-axis?
need to remove PMT-splitter, to get better image with intensifier?
15:04Z no signal..
intensifier gain 3.77V, PCO camera exposure 222ms
14:53Z try to load, 1.7A dispenser.
install intensifier + PCO-camera
need to install intensifier? to align A-beam?
14:19Z no signal in 14 min.
14:05Z try to load, dispenser 1.7A (1.75W.. rises to 1.815W after 10min)
ASI camera USB-cable pointing vertically down gives horizontal inner-electrode images
with modified imaging-objective (0.5" slimmer lens-tube, to fit outer shield cimney), focus position is ca 1mm back from previous setting
then move in Z 425um up. (upper electrode appears ca 850um up from lower)
focus on lower electrode, then move focus inside 750um/2.
removed also mid and inner shield (camera side) in order to find electrode images
remove outer shield, to align camera to electrodes
no signal in 11mins... check camera alignment to electrodes?
09:00Z try to load, with 1.7A (1.74 W)
RF monitor 2.74 Vpp
A 461nm, 250 uW (fiber was blocking IR-beam, and temp/current adjusted after moving driver-boxes)
422nm A 0.9uW, B and C 1.0 uW, detuning 185 MHz
405nm A adjusted to 3 mW
Check 461 driver settings: T = 4.642 kOhm , I = -195.23 mA (dispenser 5A and ca 2min as usual)
Adjusted 422 input polarization.
Temperature rise with outer top shield installed
14:45Z close cavity-doors after adding 3rd minus-K clamp and aligning laser to cavity
SRS shutter test:
Quiet: 10:06 - 10:19 (STDEV: CH1: 26.4 Hz, CH2: 26.4 Hz), SRS shutter (1 ms probe): 10:21 - 10:34 (STDEV: CH1: 27.4 Hz, CH2: 27.4 Hz).
14:47Z close cavity doors
Clamp down minus-K so it doesn't move
turn off cooling-laser, ASE.
no ion.. Storage time from Monday 25.1 to Friday 4.2, ca 11 days?
try large 422-detuning, and ramping back to 208 MHz
try 461nm laser
try 405nm laser
try to re-lock to Rb-line, but no ion..
Shutter noise tests:
Nothing happens: 15:14 - 15:19
Dummy clock run with 1000 ms probe: 15:20 - 15:25
Ion seems to been lost.
ID11687 NPL uMotion search, 100kPhotons. Result: [ 0.24382725 26.82098208 48.19374483]
ID11686 Peak search with OD2+OD3 based on C4 linecenter and B estimates from clock run. Probe = 190 ms, 2021-02-05-search-1
0.4 Hz sampling, 13 Hz range. 2 Hz expected AOM FWHM.
Switch to OD2+OD3. Try to operate experiment remotely.
10:16Z Abort clock-run
ID11684: Clockrun with OD1+OD3, 60ms probe, based on C1 estimates of linecenter and B field. 2021-02-04_clockrun1
New compared to last clock-run: Continuous drift-compensation with ARTIQ-DDS, adjusted in real-time to 2h drift-estimate from comb data. 2nd-order servo in peak-lock algorithm.
ID11669: OD1+OD3 peak search with probe = 60 ms. 2021-02-04-search18-. Initialized based on C1 linecenter and B from previous measurement.
C1+ 75 965 302.4 Hz, FWHM 5.6 Hz, h33
C1- 75 959 600.5 Hz, FWHM 7.0 Hz, h43
C1center = 75 962 451.4 Hz, B = 1.016029 uT
ID11678: Rabi flopping on C1+ with OD1+OD3. ~60ms still ok.
ID11669: OD1+OD3 peak search with probe = 60 ms. 2021-02-04-search17-. Initialized based on C1 linecenter and B from previous measurement.
C1+ Center visually 75.965301
ID11673: Rabi flopping on C1+ with OD1+OD3. Excitation only ~16% at max.
ID11671: OD2+OD3 peak search with probe = 190 ms. 2021-02-04-search15-. Initialized based on C4 linecenter and B from previous measurement.
Sampling 0.4 Hz, range 12 Hz. Aborted during C2+.
ID11669: OD1+OD3 peak search with probe = 60 ms. 2021-02-04-search14-. Initialized based on C1 linecenter and B from previous measurement.
C1+ 75 965 311.5 Hz, FWHM 7.4 Hz, h30
C1- 75 959 609.5 Hz, FWHM 6.8 Hz, h44
C2+ 75 968 143.3 Hz, FWHM 6.9 Hz, h38
C2- 75 956 774.3 Hz, FWHM 7.9 Hz, h26
C4+ 75 976 676.1 Hz, FWHM 8.4 Hz, h39
C4- 75 948 237.5 Hz, Unreliable fit, noise in the data.
C1center = 75 962 4605.2 Hz, B = 1.016047 uT
ID11667: OD1+OD3 peak search with probe = 60 ms. 2021-02-04-search13-. Initialized based on C1 linecenter and B from previous measurement.
C1+ fit failed. Visually 75.965312
Linecenter 75.9624608364173 MHz.
ID11664: OD1 peak search with probe = 1.9 ms. 2021-02-04-search11-
C1center = 75.96246910765049 uT
B = 1.017961 uT
Abort after C1 pair.
ID11648: OD1 peak search with probe = 1.9 ms. 2021-02-04-search2-
Drift comp seems stabilized.
ID11643, ID11646: NPL searches with wider range. Trying to debug contrast fitting errors.
uMotions: A: 4% in 12s, B: 1% in 10s, C: 8% in 17s
ID11637: Peak search with OD1. Probe 1.9 ms. 2021-02-04-search1-
+C1 75 965 329.2 Hz, FWHM 240 Hz, h31
-C1 75 959 617.7 Hz, FWHM 212 Hz, h56
C1center = 75 962 473.5 Hz, B = 1.0177363 uT
B-field estimate from the day before (1.018202) seems good enough to predict transition frequencies for OD1.
Adjust comb f0 lock. Restart drift-comp.
07:02Z Abort remaining runs. Drift-comp stopped working 2020-02-03 14:20Z.
leave OD5, 190ms scan over night... may or may not find something?
clock-laser was unlocked for some time at the start of C1+ scan. from pt 33/1000 onward it should be ok
Go to OD3 and 20ms, to find lost peaks..
4 Hz step, 200 Hz span, 51 pts/peak
C1 fit fails... left in the 200Hz span - can we have this much drift?
scan around 75 965 498, to find C1+?
find only low-height peak, try rabiflop there? get only low counts, so we are off-resoncance?
ID11627 Try instead peak-search with OD4 and 60ms probe
C1+ far left in span, maybe 75 965 498 ?
C1- also left? 75 959 780 or lower?
(B seems to have decreased? moves C1+ left and C1- right?)
190ms scan is possible only with well-compensated cavity-drift!
ID11619 repeat OD5, 190ms peak-search (search9)
Reset estimates based on C4 linecenter and B-field from OD1+OD3 peak search.
Estimate tau_pi = 190 ms with OD2+OD3
ID11618 Peak search with OD2+OD3, 190 ms probe 2021-02-03-search8-.
Sampling 0.4 Hz, range 12 Hz.
+C1 75 965 513.20 Hz, FWHM 2.2 Hz, h26
-C1 75 959 798.60 Hz, FWHM 2.1 Hz, h41
C2 fit failed. Visually: 75 968 350.4 h37
C1center = 75 962 655.90 Hz, B = 1.018282389 uT
ID11613: Rabi-flopping on C1+ OD1+OD3. tau_pi ~60 ms seems ok.
Reset B-bield and linecenter estimates based on previous C1 peak search.
ID11612 Peak search with OD1+OD3, 60 ms probe 2021-02-03-search6-.
+C1 75 965 513.9 Hz, FWHM 6.5 Hz, h32
-C1 75 959 799.2 Hz, FWHM 7.1 Hz, h41
+C2 75 968 352.7 Hz, FWHM 7.0 Hz, h34
-C2 75 956 959.5 Hz, FWHM 7.4 Hz, h29
+C4 75 976 905.5 Hz, FWHM 7.2 Hz, h36
-C4 75 948 404.0 Hz, FWHM 7.5 Hz, h
C1center = 75 962 656.6 Hz, B = 1.0182969 uT
C4center = 75 962 654.8 Hz, B = 1.0182016 uT
ID11610 Peak search with OD1+OD3, 60 ms probe 2021-02-03-search5-. Aborted after C1 pair.
+C1 75 965 513.6 Hz, FWHM 7.5 Hz, h33
-C1 75 959 798.8 Hz, FWHM 7.0 Hz, h38
C1center 75 962 656.2 Hz, B1 = 1.0183112 uT
ID11607 Peak search with OD1, 1.9 ms probe 2021-02-03-search3-.
+C1 75 965 505.5 Hz, FWHM 196 Hz, h42
-C1 75 965 505.5 Hz, FWHM 228 Hz, h49
C1center 75 962 650.3 Hz, B1 = 1.0175206 uT
uMotions at new 'optimum': A: 2.5% in 11s, B: 1% in 10s, C: 7.7% in 15s
ID11959: NPL search again. Optimum out of bounds again. Selected manually: [0.28, 27.0, 48.0]
ID11594: NPL uMotion search around new 'optimum'.
uMotions at new 'optimum': A: 6.5% in 10s, B: 7.2% in 10s, C: 3.1% in 14s
ID11589: NPL uMotion search. Optimum out of bounds. Manually selecting: [0.29, 26.6667, 48]
uMotions: A: 7.4% in 11s, B: 24.5% in 10s, C:22.0% in 15s
06:50Z Stop clock run.
ID11583 try to start OD4, 60ms clock-run
Cycle duration is ca 133 seconds. This should roughly match 6 peaks * 2 R/B sides * 100 shots * (100ms probe + 40 ms dead) = 120 s (check what our actual programmed dead-time is?)
For clock-run results see ion.
ID11578 peak-search, OD4, 60 ms (Fourier limit 6.7 Hz AOM)
1.3 Hz step, 30 Hz span, 23 pts/peak
+C1 75 965 645.2 Hz, FWHM 7.4 Hz, h29
-C1 75 959 936.7 Hz, FWHM 7.1 Hz, h39 (right in span)
C1center 75 962 790.9 Hz, B1 = 1.017194 uT
initialize clock-run with these values. (initializing with a C4 peak search might be better? more accurate abs(B) value)
ID11577 peak-search, OD3, 19 ms
2.5 Hz step, 50 Hz span, 21pts/peak
50/100Hz sidebands not resoled? this is not a good probe length?
ID11576 peak-search, OD1, 1.9 ms
now better peak-height with correct tau_pi
+C1 75 965 647, FWHM 202 Hz AOM, h38
-C1 75 959 931, 208Hz, h49
C1center 75 962 789, B1 = 1.0185 uT
+C2 75 968 488, 227 Hz, h44
-C2 75 957 102, 229 Hz, h34
+C4 75 977 030, 221 Hz, h47
-C4 75 948 563, 229 Hz, h27
OD1 rabiflop on center gives 1.9ms tau_pi.
674nm power in A-beam seems reduced? re-align?
ID11569 peak search, OD1 and 3ms - but get low peak-height for +C1, so probe length is wrong?
ID11568 rabiflop on +C1 (75 965 644), with OD3,
ID11567 peak search, A-beam, OD3, 30ms probe
2.5 Hz step, 100Hz span, 41pts/peak
ion dark moment? abort
ID11563 peak search, A-beam, OD3, 30ms probe
uMotion at old voltages 100kphot: A 7.3%, B 19%, C 13%
ID11557: OD1 peak search with adaptive Artiq drift comp. 3 ms probe, 2021-02-02-search4-
uMotions: A: 6% in 13s, B: 22% in 10s, C: 18% in 15s
ion still stored (now 8 days?)
Leave 100ms peak-search over night. 0.8 Hz step, 200Hz span, 251 pts/peak
A-beam, 18 dB Zatt., OD4, 100ms probe
Quite difficult to get a clock-run to lock to the peaks with so much uncompensated cavity-drift...
100ms peak search again. 0.8Hz step 20Hz span
f_beat_center = 75 962 869
+C1 far left in span ?
Try clockrun initialized with (centers, FWHMs AOM, gains) [75.96589229 75.95992971 75.96885552 75.95696648 75.97778133 75.94804067] [4.e-06 4.e-06 4.e-06 4.e-06 4.e-06 4.e-06] [0.5 0.5 0.5 0.5 0.5 0.5]
100ms probe, 6 peaks gives cycle time of ca 181 s
somehow we again have +10 mHz/s of uncompensated cavity drift. This gives 1.8 Hz @ 674nm drift during the 180s servo-cycle. For a 100ms probe, 8 Hz Fourier limit, this is 1/4th of the FWHM!
ID11534 peak search with 100ms
+C1 far left in span? 75 965 892 (or even lower?)
-C1 right in span 75 959 929
C1 center 75 962 911 and B1 1.06247 uT
predicted line-center from beat-note 75 962 884
Try clock-run with 100ms probe, but servos seem not to find the peaks..
Based on OD3 results, switch to OD4 and 100ms probe.
ID11528 peak search A-beam, 18 dB Zatt, OD4, 100ms probe, 0.8 Hz step, 20 Hz span, 25 pts/peak
+C1 seems far left in the span? 75 965 901
-C1 more centered in span (why?),and higher, 75 959 922
C1center 75 962 912 and B1 1.06532 uT
With OD3 and 30ms, found C1 peaks:
+C1 at 75 965 910
-C1 at 75 959 920
C1center 75 962 915 and B1 1.0673 uT
ID11525 A-beam peak search. 18 dB AOM att, OD3-filter, 30ms probe
2 Hz step, 200 Hz span. 101 pts/peak
C beam 106/3 counts in 10ms
B beam 94/3 counts in 10ms
A beam fluorescence 173/4 bright/dark counts in 10ms
422 power A 1.5uW, B 1.9uW, C 1.5 uW
13:05 after many hours of working with the 422-laser (no cooling), ion re-appeared!
repeat measurement of the cavity power-shift (Friday result was 190Hz/V @ 1348 nm)
With reduced FALC-gain, photodiode signal is slightly more than before.
Cavity drift-compensation should also match true cavity drift better than before
13:45Z set-point 0.8 V, 5min beat-note 23 164 432.26 (stdev 30.39 Hz)
13:52 set-point 0.7V (BOA 430 mA), 5 min beat-note 23 164 433.33 (valid data only)
14:01 set-point 0.8V (BOA 480 mA), 5 min beat-note 23 164 434.19 (31.19 Hz) - not much evidence of power-shift?
14:24 set-point 0.9V (BOA 520 mA), 5 min beat-note 23 164 434.41 (32.57 Hz)
14:56 set-point 1.0V (BOA 580 mA), 5 min beat-note (this hits the BOA current limit of 600 mA)
again dark image after power-measurements? takes a long time to cool and get an image?
422nm 0.6uW in A-beam
15-20 uW before ABC-beamsplitter
A-beam fluorescence 32/12 bright/dark in 10ms gate. B and C weaker
09:47Z ion re-appeared!? maybe weaker image than before, lower 422nm power after fiber-adjustments
09:40Z seems we lost the ion?, while adjusting polarization (no cooling-laser)
422nm Polarization extinction ratio improved from 7 dB to 10 dB, on ion-breadboard after 422-laser-fiber, before ABC-beamsplitter
Note on cavity doors closed (Friday):
Coarse heater-power dropped from 90W to 30W - as before. time-constant is 'fast', below one day
Inner shield free-running temperature sensor seems to settle at to a temperature +4...6 mK than before. Very slow settling, maybe a week?
09:24Z adjust comb f0-beat SNR to >25 dB
Updated drift-comp to current observed value of 12.6 mHz/s
Reducing FALC-gain (trimpot) gives lower servo-bump sidebands and more power transmitted through cavity, now 0.85 V
'Heating' fiber by hand, both laser -> breadboard and breadboard -> ion.
ID11492: A-beam, samples: 0-200: laser-> breadboard, samples: 200-350 breadboard->ion, 350 ->: laser to breadboard. Changes mostly in first case.
ID11490: C-beam, samples: 0-200: laser-> breadboard, samples: 200-350 breadboard->ion, 350 ->: laser to breadboard. Changes in both cases.
ID11488: B-beam, samples: 0-200: laser-> breadboard, samples: 200-350 breadboard->ion, 350 ->: laser to breadboard. Changes in both cases.
ID11486: A-beam, PMT count changes with heating laser -> breadboard, but not with breadboard -> laser.
Check 422 beam polarizations.
Cancelled long experiment from ID11481 forwards.
Leave B-beam, OD5, 60 ms peak search over night
Maybe 50/100 Hz sidebands are smaller, when t_probe is a multiple of 10 or 20 ms?
ID11477 (narrow span), 11478(search5), 11479(6), 11480(7) (aborted), 11481(8) (cancelled), 11482(9) (cancelled), 11483 (10)(cancelled)
Try B-beam instead?
rabiflop with OD3 suggest maybe 10ms tau_pi?
peak-search with 10ms and OD3
OD2 tau_pi looks like 3ms
OD3 maybe 10 ms
OD4 no clear oscillation?
OD5 rise tp 60ms, then flatish
recalculate peak-positions based on beat-note and latest B1-estimate.
try peak-search with OD1 and 100ms probe
+C1 at 75 965 640. try rabiflop there.
no clear signal, try OD3 and 30ms instead.
ID11467 peak-search with OD3 and 30ms
rabiflop on 75 965 669, with OD3. no higher than 25 counts?
abort long run.
ID11461 OD3+OD2 and 290ms peak search, long, over night. 290ms is a guess of tau_pi, might be too long?
18dB Zeeman AOM atteunation.
0.3 Hz step, 150 Hz span, 501 pts/peak
duration ca 30s per point = 4 hours per zeeman peak! done in ca 24 hours..
Fourier limit 2.8Hz optical, 1.4 Hz AOM. ca 5 steps of 0.3 Hz AOM each
ID11460 peak search with OD3 and 29ms probe
49pts/peak. 150 Hz span, 3 Hz step
Fourier limit is ca 25 Hz - so the 50/100 Hz sidebands are resolved.
ID11458 peak search with OD1 and 2.9ms probe
+C1 75 965 242, FWHM 174, h26
-C1 75 959 314. FWH; 153, h34
C1center 75 962 279, B1 1.0562 uT
+C2 75 968 195, FWHM 159, h35
-C2 75 956 362, FWHM 169, h31
+C4 75 977 060, FWHM 152, h27
-C4 75 947 479, FWHM 149, h18
Estimated line-center from comb 75 962 255
here get: A 6.7%, B 13.3%, C 8.4 %
algorithm suggests [ 0.27459138 26.48031258 49.01349212]
old voltages: 0.2693, 26.9714, 49.4475.
ID11446 Try short uMotion-search 3x3x3 grid with 30kphotons
uMotion 100kphot: A 15.3%, B 14.7%, C 9.6%
15:56Z stop collition monitor
cavity output-power now more than on Friday? ca 0.7 V
comb f0 lock seems very unstable
fiber was blocking ion-breadboard 674nm power-detector...
turn on comb, comb-branch, clock laser
15:50Z ion still stored
16:27Z leave lab.
16:20Z start collision-monitor with 10ms gate, 65/3 bright/dark. DB tag '2021-01-29-run1'. NO clearout.
Fluorescence levels 66 / 2.8 in 10ms
Switch to B-beam.
Turn off clock-laser and beat-note branch
Comb f0 unlocked. (maybe lab-temperature changes when cavity-box was closed?)
Cavity drifts too much to leave any useful clock-laser runs
ID11439 re-start OD3/29ms search, with wider span, since we have so much uncompensated cavity-drift.
ID11438 peak search with OD3 and 29ms probe (Fourier limit 13 Hz AOM)
Note we have 50-60 mHz/s uncompensated cavity drift!
C1 Gaussian fit fails...
ID11437 repeat peak-search (since something jumped??)
Cavity drift seems to be 70 mHz/s - since we touched the cavity when testing power-shift
C1 75 964 506, FWHM 97 Hz, h24 and 75 958 555, FWHM 107Hz, h36
C1center 75 958 555, B1 1.0604 uT
C2 75 967 440, FWHM 107 Hz, h38 and 75 955 605, FWHM 123 Hz, h21
C4 75 976 334, FWHM 94 Hz, h32 and 75 946 679, FWHM 89 Hz, h21
ID11436 peak search, OD2, 9.9ms probe (Fourier limit 40 Hz AOM). 8 Hz step, 170 Hz span
C1 75 964 528, FWHM 93 Hz, h21 and 75 958 587, FWHM 103 Hz, h31
C1center 75 961 557, B1 1.0587 uT
C2 75 967 452 FWHM 112 Hz, h32, and jump??
C4 only zeros?
ID11435 Repeat OD1 peak search for C1 pair. 3.3ms probe (Fourier limit 121 Hz AOM), 25 Hz step
C1 FWHMs 134 Hz and 150 Hz
C2 FWHMs 150 Hz and 164 Hz
C4 FWHMs 133 Hz and 137 Hz
(measurement time 688s)
ID11422: Rabi flop with OD1, att. 18 dB. tau_pi 3.3 ms. -> use 10 ms for OD2
ID11418 Line scan over C1+ with 18 dB att. and OD1.
ID11415 Rabi flopping on C1+ with 18 db att. and OD1. Measurement was off-peak.
ID11411 peak-search with OD1 and 3ms probe. 15 dB att. C4+ fit wrong. Corrected dataset manually.
13:08 close cavity doors
Try to measure cavity power-shift.
ca12:20Z: 0.40 V power-servo set-point (servo measures cavity output-power with photodiode, and steers BOA 1348nm amplifier current)
12:24 feedback off
12:24:30 0.3V set-point, feedback ON
1s comb-data noise is ca 25 Hz at IR frequency of 222.5 THz (1e-13, the maser noise level at 1s)
5min = 300s average should give 1/sqrt(300) of this, or 1.4 Hz @ IR = 2.2e-14
12:32 adjust power-servo, set-point 0.4V (servo on causes BOA current variations 450-490 mA)
5min beat-note 23 161 465.41 (1s stdev 25.8)
12:37 servo off
12:39 servo set-point 0.3V (BOA current 350-380 mA) - servo-bumps on beat-note SA look much smaller
5min beat-note 23 161 492.17 (1s stdev 27.2)
12:44 servo off
12:46 set-point 0.35V (BOA current 410-440 mA)
5min beat 23 161 508.94 (1s noise 25 Hz)
12:52 set-point 0.45V (490-530mA)
in-loop photiode scope-signal at mu=451, 450.9, 451.0, 451.0, 450.9, 451.0, 451.2 451.1 451.0 451.1 mV
5 min beat 23 161 522.1 (25.3 stdev)
Power-shift slope is 190 Hz / V of frequency shift in IR, relative to cavity-output photodetector voltage
This is 8.5e-13 fractional per volt.
If we want a 1 Hz 674nm laser-stability, we need 0.5 Hz in IR, or 2e-15 fractionally
This leads to a requirement of 2.6e-3 for the fractional voltage stability of the photodetector
Closed-loop we see 0.3mV/450mV which is 7e-4 fractional
uMotion contrasts with new voltages [ 0.26927269 26.97143157 49.44745276]: A: 2% in 11 s, B: 5% in 12 s, C: 2% in 17 s
ID11403: NPL search for umotion minimum: 4x4x4. Result: [ 0.26927269 26.97143157 49.44745276] (optimum now within search-grid?)
-0.0203 2.7627 2.2088
0.3076 -1.4601 0.2457
0.2695 1.1614 -3.8900
Previous voltages: (0.3821,26.7839,51.0685)
ID11400: NPL search for umotion minimum: 4x4x4. Result: [ 0.22398756 27.24888464 49.35180688]
uMotion 100kphot: A 30% in 10s, B 46% in 11s, C 23% in 13s
Peak search terminated (ID11392). Partial +C1 data in DB.
(rabi-sinc with transform-limited FWHM shown for reference)
ID11391, ID11392: Overnight: Peak search with OD2+OD3 filter, 225 ms probe
DB-tag: 2021-01-28-search14-, -search15-.
Fast drift comp. Estimated length of one search: 2.2 hrs per peak = ~13 Hr
100 Hz span per peak. Step: 0.3 Hz, 333 pts/peak. Fourier limit 1.78 Hz AOM
ID11387, ID11388: Rabi flop with OD1+OD3. tau_pi closer to 75 ms. -> use 225 ms probe for OD2+OD3.
ID11386: Scan over +C1 with OD1+OD3 filters, 72 ms probe, DB-tag: 2021-01-28-run8. With fast drift-comp.
ID11385: Peak search with OD1 filter, 3 ms probe, DB-tag: 2021-01-28-search12-. Fast drift comp.
+C1: height 28, FWHM: 173 Hz
-C1: height 42, FWHM: 178 Hz
+C2: height 42, FWHM: 172 Hz
-C2: height 27, FWHM: 147 Hz
+C4: height 53, FWHM: 152 Hz
-C4: height 30, FWHM: 163 Hz
ID11382: Scan over +C1 with OD2+OD3 filters, 216 ms probe, DB-tag: 2021-01-28-run7. With fast drift-comp.
ID11381: Scan over +C1 with OD1+OD3 filters, 72 ms probe, DB-tag: 2021-01-28-run6. With fast drift-comp.
Disable drift-comp in crontab, and update drift-DDS every second:
ID11378: Scan over +C1 with OD1+OD3 filters, 72 ms probe, DB-tag: 2021-01-28-run4. With fast drift-comp.
ID11377: Scan over +C1 with OD1+OD3 filters, 72 ms probe, DB-tag: 2021-01-28-run3. With normal drift-comp.
ID11376: Scan over +C1 with OD1+OD3 filters, 72 ms probe, DB-tag: 2021-01-28-run2. With normal drift-comp.
ID11375: Scan over +C1 with OD2+OD3 filters, 216 ms probe, DB-tag: 2021-01-28-run1
ID11373: Peak search with OD1+OD3 filters, 80 ms probe, DB-tag: 2021-01-28-search9-. Aborted after C1-
ID11372: Peak search with OD1 filter, 3 ms probe, DB-tag: 2021-01-28-search8-. Data looks OK now.
+C1: height 34, FWHM: 151 Hz
-C1: height 47, FWHM: 162 Hz
+C2: height 44, FWHM: 164 Hz
-C2: height 27, FWHM: 151 Hz
+C4: height 44, FWHM: 160 Hz
-C4: height 26, FWHM: 170 Hz
Switch back to A-beam
It turns out, that C-beam was on?!?!
ID11363: Rabi flopping without OD filters
ID11351: Peak search with OD1, peaks seems really broad.
re-lock comb f0
uMotion 100kphot: A 18.2% in 10s, B 38% in 9s, C 9% in 18s
08:40Z ion still stored. Clock-laser locked. Comb f0 unlocked sometime during the night
60s stack of 1s exposures. old objective, ASI camera
ID11337 (run4), 11338 (run5), (18 local)
ID11339, 11340 (19 local)
ID11341, 11342 (20 local)
ID11343 (21 local)
ID11344, 11345 (22 local)
ID11346 (run13), ca 23 local?
OD3+OD1 filters, guess 72ms tau_pi (Fourier limit 5.6 Hz AOM). scan C1+ many times. 400 Hz span, 1Hz step
each scan takes 4900 seconds = 1.4 hours
10 scans in the queue, run3 thru run13
ID11322 peak-search with OD2 and 7.2ms probe (Fourier limit 56 Hz AOM). 400Hz span, 10Hz step, 41 pts/peak
C1+ h 24, FWHM 109 Hz
C1- h 38, FWHM 99 Hz
B1 1.0585 uT, linecenter 75 962 451
C4+ h 32
C4- h 18
Try OD2 filter.
ID11319 rabiflop on +C1, tau_pi 7.2 ms?
ID11315 rabiflop on +C1.
ID11314 peak search with OD3, 20ms probe (guess of tau_pi), Fourier limit 20 Hz AOM, 200 Hz span, 5 Hz step, 41 pts/peak
+C1 75 965 411, h 13, FWHM 89 Hz but Gaussian fit does not fit central peak
-C1 75 959 471, h 21, FWHM 76 Hz. Central peak ca 25 Hz AOM wide
B1 1.058 uT
linecenter 75 962 441 Hz
C2+ 75 968 369, h 22, FWHM 96 Hz
C2- 75 956 527, h 18, FWHM 79 Hz
C4+ peak is far left in the span, B-field estimate too high?
C4- peak is far right in span
Repeat OD1 search (ID11313), Influx tag: 2021-01-27-search6-
all peaks roughly transform-limited with 3 ms probe
Rerunning peak search with OD1 (ID11312), probe 3 ms, Fourier limit 133 Hz AOM, Influx tag: 2021-01-27-search6-:
C1+ heigth 33 FWHM 154 Hz
C1- height 40 FWHM 166 Hz
C2+ height 34 FWHM 191 Hz
C2- height 24 FWHM 166 Hz
C4+ height 49 FWHM 155 Hz
C4- height 29 FWHM 152 Hz
Problem with saving Zeeman-peak data in peak-search program. Problem corrected.
Try OD3 filter, should work with 27ms?
ID11286 peak search, A-beam, OD3 filter, 27ms probe. Fourier limit 15 Hz AOM. 150 Hz span, 3 Hz step, 49 pts/peak
(db tag 2021-01-27-search5-)
+C1 central peak looks OKish, but Gaussian fit is to all of data. no clear 100Hz sidebands?
Other peaks looks similar
Clock probe maybe too long?
Try OD2 filter, should work with 9 ms probe
ID11285 peak search, A-beam, OD2 filter, 9ms probe (Fourier limit 44 Hz AOM). 250 Hz span, 10 Hz step, 25 pts/peak
(Note Fourier limit close to AC frequency of 50Hz - could interfere?)
(db tag 2021-01-27-search4-)
+C1 h 22, FWHM 136 Hz AOM - looks noisy, Gaussian not a great fit (span is a bit narrow)
-C1 h 34, FWHM 115 Hz - B1 1.065 uT - linecenter 75 962 441 Hz
+C2 h 33, FWHM 140 Hz - strong sidebands at 5 steps = 50 Hz AOM = 100 Hz optical from center! - Gaussian is fit to everything, not only central peak
-C2 h 21, FWHM 153 Hz - also sidebands
+C4 h 37, FWHM 104 Hz - highest peak in this run! - not any wider than C1/2, so isn't limited by B-field?
-C4 h 79, FWHM 110 Hz - Gaussian covers central peak and side-bands..
Try OD1 filter, this should work with 3ms probe
ID11283 (data file stored as ID11284?) peak search, A-beam, OD1 filter, 3ms probe (Fourier limit 133 Hz AOM). 500 Hz span, 25 Hz step, 21 pts/peak
(db tag 2021-01-27-search3-)
+C1 h 33, FWHM 157 Hz AOM
-C1 h 46, FWHM 157 Hz - B1 1.066 uT - linecenter 75 962 434 Hz
+C2 h 45, FWHM 142 Hz
-C2 h 25, FWHM 175 Hz
+C4 h 45, FWHM 164 Hz
-C4 h 26, FWHM 183 Hz
ID11275 NPL uMotion search over lunch. 4x4x4 grid
algorithm suggests Optimal voltages (Common mode, diff): [ 0.38212489 26.78389499 51.06850713] - both first and last value are outside scan range.
Try suggested voltages.
uMotion 100kphot: A 4.7% in 14s, B 27.3% in 12s, C 12.9% in 14s. (overall better than before, but not optimal)
ID11274 peak-search with 1000us probe (Fourier limit 400 Hz AOM). 1.5kHz span, 80 Hz step, 19 pts/peak
+C1 height 43, FWHM 384 Hz
-C1 h 52, FWHM 423 Hz - B1 1.064 uT - liencenter 75 962 455 Hz
+C2 h 50, FWHM 405 Hz
-C2 h 41, FWHM 356 Hz
+C4 h 51, FWHM 432 Hz
-C4 h 30. FWHM 415 Hz
(duration 607 seconds)
ID11272 rabilfop on +C1, to 5 ms probe. tau_pi ~ 1ms (this might give better peak-heights than previous peak-search)
ID11271 rabiflop on +C1
Peak search (ID11270, 18 dB att., 500 us probe):
+C1: Center = 75.9654 MHz, height 20, FWHM: 890 Hz.
-C1: Center = 75.9594 MHz, height 27, FWHM: 942 Hz.
+C2: Center = 75.9684 MHz, height 35, FWHM: 873 Hz.
-C2: Center = 75.9565 MHz, height 22, FWHM: 861 Hz.
+C4: Center = 75.9773 MHz, height 22, FWHM: 862 Hz.
-C4: Center = 75.9476 MHz, height 12, FWHM: 840 Hz.
B estimate = 1.07137 uT.
Coil currents set to (-8,0,0).
-3 floor A/C setting is 20 Hz
NOTE uMotion contrasts not optimal, all beams ca 30% now!
ID11260 compare to non-zero B-filed by flipping I1 sign, currents: (+4.4719, -1.0115, 1.0309) mA. lower 18 kcounts/s at max
ID11258 repeat A-beam detuning scan, 1000ms gate. Far detuned 2 kcounts/s, near resonance max 21.3 kcounts/s
ID11257 A-beam detuning scan - no Dork-resonances apparent..
ID11254 clock-spectrum with zero B-field, 500us probe (Fourier limit 800 Hz AOM), 3 kHz span, 50Hz step
Gaussian fit FWHM 908 Hz
center 75 962 438
ID11252 rabiflop on center, in zero B-field, visually tau_pi ca 500us
ID11248 clock-spectrum with zero B-field?
A-beam, 600us probe (Fourier limit 667 Hz AOM), 100Hz step, 5 kHz range. zeeman AOM attenuation 18dB from y-day
Gaussian fit FWHM 760 Hz AOM
center 75 962 481 Hz
Try the zero B-field currents I1 = -4.4719, I2 = -1.0115, I3 = 1.0309
uMotion 100kphot: A 36.3% in 9s, 30.4% in 10s, 28.7% in 13s
08:19Z ion still stored, clock-laser locked to cavity, but comb f0 was unlocked.
Initial coil results:
B = B0 + A I
0.2649(2) 0 0
0 0.41962(14) 0
0 0 0.2713(2)
% current giving zero field
IO = [-4.4719 -1.0115 1.0309] mA
A-beam, 600us probe, 500Hz step (note this is quite sparse sampling... Fourier limit 667 Hz AOM), 50 kHz span, 101 points per spectrum
ca 5s per point, so only 500s per spectrum. We have 3*6 spectra, so 3*6*500 = done in 3 hours.
step each current: +2, -2, +4, -4, +8, -8 mA.
2021-01-26-coil-0-0-0 is the first run, with zero currents
0, 0, 0
+2, 0, 0
-2, 0, 0
+4, 0, 0 leave lab around 17:10 local, artiq got to this point.
-4, 0, 0
+8, 0, 0
-8, 0, 0
0, +2, 0
0, -2, 0
0, +4, 0
0, -4, 0
0, +8, 0
0, -8, 0
0, 0, +2
0, 0, -2
0, 0, +4
0, 0, -4
0, 0, +8
0, 0, -8
Note the last coil characterization results from 2020-08-19: (coil sensitivities very similar on 2020-06-12)
A2 = [0.216895938211581 0 0;
0 0.325633046718816 0;
0 0 0.208504965524669];
B02 = [1.085916150083249, 1.291554002640864, -0.100996485194311]';
orange: zero coils
blue: I1 varied
red: I2 varied
green: I3 varied
Remove OD-filters, for overninght coil characterization
A-beam, no OD filters, rabiflop shows 600us tau_pi
Spectrum with 4.5ms and 18dB attenuation. Fourier limit 89 Hz AOM
Gaussian fit FWHM 124 Hz AOM
center 75 963 825 Hz
Try increasing Zeeman AOM attenuation, to only slightly lower probe intensity. Attenuation 18dB.
Rabi-flop on center gives 4.5ms tau_pi
ID11197 scan with OD2 and 3.2ms. Fourier limit 125 Hz AOM
Gaussian fit FWHM 144 Hz AOM
Center 75 963 826
ID11195 rabiflop on center, OD2 filter. tau_pi 3.2ms
OD3 filter is too hard? go back and try OD2 filter?
re-run rabiflop on center
re-run +C1 scan with 9400us probe (Fourier limit 42 Hz AOM). A-beam, OD3 filter
Gaussian fit FWHM 102 Hz AOM
Rabiflop, to find tau_pi?
OD3 filter, 15ms probe, scan around +C1
OD2 filter, 4500us probe, scan around +C1
Try coil currents (-4, 0, 0) mA, scan with 450us
Peak-search with A-beam and 450us probe:
but doesn't find C1.
rabiflop on -C1, visually tau_pi ca 450 us
P674_A = 0.8 uW, no OD-filters (note B-beam had about 5uW!)
Try A-beam, it may have better 422nm polarization?
ID11174 scan with 600us probe pulse
C1 peak search OD3+OD1 filter, 15 ms probe, 5 Hz step, 100 Hz span, Fourier limit 27 Hz AOM
+C1 not a great peak..
-C1 75.960007, FWHM 70 Hz AOM, h 22 (Gaussian fit not so great)
C1 peak search, OD3 filter, 5400us probe pulse, 300 Hz span, 20 Hz step, Fourier limit 74 Hz AOM
+C1 75.964675, 129 Hz, h 19
-C1 75.960007, 124 Hz, h 35
B1 = 0.832 uT
C1 peak search, OD2 filter, 1800us probe pulse, 500 Hz span, 50 Hz step. Fourier limit 222 Hz AOM
+C1 75.964672, 272 Hz, h 19
-C1 75,962336, 246 Hz, h 48
B1 = 0.832 uT
C1 peak search
+C1 75.964650, 703 Hz AOM FWHM, height 30
-C1 75.962345, 753 Hz FWHM, height 32
B1 0.821 uT
Coil sensitivity roughly dB/dI1 = 0.24 uT/mA ?? (compare to previous results)
abs(B0) field at zero coil currents 1.3 uT?
Lineshape scan, OD1: ID11132
Change coil currents to (-2,0,0)
Lineshape scan, OD1: ID11130. C2 peaks (?): 75.97232, 75.95232
Change coil currents to (2,0,0).
Install the OD1 filter, this should increase tau_pi 3-fold, to 600us.
ID11121 scan spectrum with OD1 filter, 600us probe, 40 kHz range, 200 Hz step. Duration 800s
Comb f0 had unlocked over lunch, fixed by adjusting waveplates in GUI
All peaks seem to fit within 40 kHz AOM
ID11120 wider spectrum scan with 200us probe
C1 separation looks like roughly 15 kHz AOM ?
ID11119 rabiflop on +C1(?), tau_pi ca 200us
ID11118 initial clock transition scan with 15dB AOM attenuation, 600us probe pulse.
674 power in B-beam: 5 uW
Detuning scans: A: ID11108, B: ID11110 C: ID11112. Gate 20 ms. Counts: ID11115 (dark approx. 10)
Manually trying: (0.28, 27.3333, 53.3333): 8%, 6%, 3%
Automatic NPL search (ID11098) gave weird result [ 0.08607781 32.11213181 52.31081692]
Try (0.3571, 26.8571, 53.5714), got 3%, 31%, 3%
-0.05B step to (0.1334, 26.77556, 54.59336) gives (3%, 2%, 39%)
Repeating search with (0.2489, 26.6999, 54.6666) gives now 7%, 25%, 29% (repeat 10%, 29%, 36%)
NPL search finished with [ 0.33088186 27.33750859 54.74849937], but ion seems to have disappeared during the night.
Last measured contrast seems ok.
Ion reappeared after switching on all beams.
Contrast with new voltages: 33%, 32%, 16%
Leave NPL-algorithm running, 8x8x8 grid
By scanning V34diff, found B and C optimal point:
Contrasts: 30 1.4 3.9 %
Voltages: 0.2489, 26.6999, 54.6666
Found rough optimum for B beam, by stepping along the B-vector
Contrasts: 71, 13, 69 % at voltage of (0.841, 26.692, 47.1523)
From 2020-10-05 we have some notes on the cardinal directions:
0.2372 2.3100 1.9737
0.4131 -1.5132 -0.0264
0.6644 1.4648 -3.8875
uMotion 100kphot: A 70% in 10s (B and C similar ca 70%)
Coil currents from December were (127.7, 6.1, 18.4) mA. Now set them all to zero
Adjust alignment into 844nm amplifier - seems to improve 422nm power
422nm power seems halved compared to start of day - fiber coupler between 422-laser and ion-breadboard?
uMotion 100kphot: A 73% in 58s
Install mid shield
Install inner shield
Install pt100 sensor on IR side window flange
uMotion 100kphot: A 82 % (anharmonic) in 38s, B 73% in 56s, C 76% in 75s (Note no shields!)
C ca 70 counts/50ms
B ca 100 counts/50ms
A-beam and ASE alignment optimization gives ca 140 counts in 50 ms
Found PMT signal, in upper left corner of camera image, as before.
ion is still stored, Z-focus of camera was wrong..
Align PMT-splitter with help of 422-fiber-end placed at PMT position. This allows aligning the PMT-splitter so 422nm-light is bright at analog camera on IR side
Lost ion (?) while swapping cameras etc.
Install the PMT-splitter next
Switch to the ASI camera. Remove intensifier/PCO.
Align B-beam (thru-transmitted diffraction pattern) -> get fluorescence (slightly weaker) with only B also.
Align C-beam -> we get fluorescence with only C now also.
Try camera (z) focusing - but it was already quite optimal. (thus aligning the camera to the electrodes, and defocusing by 750um/2 is a good way to z-focus the camera)
set detuning to 208 MHz - more fluorescence
optimize A and ASE directions, but no major fluorescence changes. PCO camera to 200ms exposure
11:53:15 see single(?) ion? load time 5min 15s
11:48 try to load single ion. 1.7A dispenser
throw away cloud by shutting off RF
A cooling shows cloud, B and C quite dark (do not cool much?)
11:43:37 dispenser off. cloud remains
11:42 see a cloud! (intensifier gain higher, PCO exposure up to 1.2 seconds). A-beam adjustments while dispenser on finally revealed the cloud.
stopping down the imaging objective with the iris reduces a vertical strange 422-reflection in the camera image
11:27:15Z try load, dispenser 1.7A. Intensifier gain 2.7V, PCO camera 200ms exposure
Align ASE beam with the analog camera, A-beam with the thru-transmitted diffraction pattern from the electrodes.
Optical power in the beams:
422nm A 0.9 uW, B 1.1uW, C 1 uW. 185 MHz detuning
461nm A 180 uW
clearout 65 uW
repumper 10.8 mW
405nm A 3.6 mW
RF 14.424 MHz, 10Vpp generator, 2.759 Vpp monitor (scope in averaging mode)
Latest bias-voltages are from 2020-10-29 (?). They are -6.2, 0.2635, 27.0703, 46.7861. Apply these.
This gives individual voltages: V1 = -6.4635, V2 = -5.9365000000000006, V3 = -19.715799999999998, V4 = 73.8564
Analog camera images IR/ASE side. Intensifier + PCO images ion.
Remove mid top-shield, and inner shields - to allow aligning beams. Mid bottom-shield(s) need not be removed?
Try to trap again?
Temperature rise with thermal compound between helical-resonator and alu-mount.
Adding thermal compound changed helres heating from 120 mK to 90 mK.
Installed longer antenna: monitor voltage now 2.78Vpp! f = 14.422 MHz
re-tune SWR close to 1.0. Monitor 60 mVpp with 10Vpp drive from siggen
Install heatsink-compound between helical-resonator and alu-mount
Install inner and mid-shields. Seems to fit OK. Beams go through shield-holes OK.
Note back feet of mid-shield bottom left/right half did not fit the designed M3 holes - they loose between bottom shields and pump-chamber
Tickler voltages measured after coupling capacitor (at eletrode) with 10x probe. Voltages are peak-to-peak
Measured voltages seem low. Capacitors of the DC filter (fig. from OC18 guidelines) equal to approx. 0.5 or 0.25 ohms at 1 MHz and 2 MHz respectively.
This seems to load the DDS output significantly
f (MHz), DDS att., V
1.0 5 distorted..
1.0 10 7.4 mV
1.0 15 4.4 mV
1.0 20 2.9 mV
2.0 5 distorted..
2.0 10 3.0
2.0 15 2.0 mV
2.0 20 1.5 mV
20 dB attenuator
f (MHz), DDS att., V
1.0 3 1.97 mV
2.0 3 ~500 uV (noisy)
Mu-Shields delivered. Inner and mid-shield unpacked - looks ok. Quite heavy duty material thickness!
Tomorrow: Check ASE-beam alignment (analog camera), RF trap drive OK?
15:03 no signal, dispenser off
Intensifier gain turned up, PCO exposure ca 100-200ms
14:51Z try to load with 2.0A. Now using image-intensifier and PCO camera.
Field of view now larger, both top and bottom almost in same view
NOTE: Image-intensifier input side is closer to the electrical wires. 3 wires: Orange (+5 to +15 V supply, 30 mA), White (Control voltage 0-5 V), Grey (Ground).
Use the old PCO-camera, to image the intensifier output plane
Install the (new) image-intensifier, this may help detect weak signals?
12:56 no signal, dispenser off
12:46 try load with 2.0A dispenser (2.3W initially)
check Rb-lines, they look ok
check 461nm with dispenser: 198.2 mA, 4.645 kOhm
12:32Z no signal, dispenser off
12:17Z try to load, 1.7A dispenser
remove PMT-splitter, to get more signal to camera. Use old imaging-objective. re-align objective to electrodes
(O-ring inside PMT-splitter-cube was loose?) blocking light thru lid is important
11:07Z no signal, dispenser off
10:53Z try to load, 1.7A dispenser
09:46 no signal, dispenser off.
09:32Z try to load, 1.7A dispenser
with old objective A-beam window reflection is quite strong, and exits the objective-tube towards B-beam post
new objective may have issues with beam window-reflections? Brass tubes used as spacers may need to be painted black?
test new imaging objective - but no huge improvement in image-quality of electrode image with flashlight illumination
realign camera to electrode image
08:52 no signal, dispenser OFF
08:37Z load with 1.7A dispenser
trap RF as before. monitor waveform seems jumpy!?
Optical power (slim sensor)
422nm A-beam 0.8uW, B 1.1 uW, C ~1uW
461nm A 200uW
405nm adjust to 3.7 mW
clearout 64 uW
repumper 10.6 mW
align A-beam with help of electrode reflections
remove inner shield (outer shield was removed yesterday).
win machine tilted? reboot.
13:28 no signal, dispenser OFF
adjust A-beam, based on output beam shape...(??)
13:15 try 1.7A dispenser, 1.7W
13:13 no signal, dispenser OFF
13:01Z try again with 1.6A Dispenser. 422nm Rb-lines checked, 405nm power checked. A-beam output doesn't look very clean (hits some aperture somewhere?)
12:50 no signal, dispenser OFF
12:47 no signal in 10min..
12:37:30Z dispenser on, 1.6A. 1.45W initially, rising slowly to 1.5 W
Try to load with 1.6A dispenser current
Trap RF from signal generator 10 Vpp, 14.424 MHz
monitor 64 mVpp
Preparing for trapping.
461 dispenser check, OK at I = 198.3 mA, T = 4.645 kOhm.
(New Thorlabs slim power sensor, model S130C)
422 power: A exit: 0.85 uW, B: 1.1 uW, C: 1.0 uW. Detuning at 185 MHz
461nm A-beam 195 uW
405nm A-beam adjust to 3 mW
repumper 13.5 mW (repeat 12mW) at ASE beam output
clearout 65 uW (seems low?)
(AW paper notebook clearout values: 75 uW on 2019-12-04, 77uW on 2019-11-25, 78uW on 2019-11-04)
New Kubuntu linux install
Installed new OS on new disk. Migrated all software.
Artiq version now: v5.7144.9ed8f663. Conda enviroment: artiq-ion-v5-6.
Ionclock ARTIQ now ident 5.7144.9ed8f663;mikes (autogenerated by M-Labs).
New drift-DDS ARTIQ now ident 5.7144.9ed8f663;mikes-drift (built manually).
Drift-DDS ARTIQ seems to work. Configured for external ref for both Kasli and Urukul.
(slight improvement with lid-adjustment and heatsink)
13:50Z add a heatsink to the alu end-plate
13:34Z improve slightly contact between antenna-lid and end-alumount. No visible reduction in monitor-voltage when loosening lid M3 screws (impedance matching still ok?)
With added clamping, resonator heating is reduced 140 -> 120 mK. RF-feedthru heating ~few mK lower.
(note the new year 2021 AD...)
Install 2pcs hose-clamps that push the Cu-resonator into contact with the Alu-mounts
12:13Z install clamps to breadboard on both sides of Alu helres-mounts
Heating of the RF-feedthrough flange (10Vpp from signal-gen, 6+ hours settling time):
|Old mount, no sink
|Old mount, Cu bars as heat-sink
|New mount, 2 mounts clamped
|New mount, mounts clamped, 2 hose-clamps around cu-helres
||66 - ~few mK (results 2021-01-14!)
Results with two alu-mounts clamped to breadboard, and no clamping-force between cu-resonator and alu-mounts.
Install new alu-mounts that imrpvoe heat-ocnduction from Cu-resonator to breadboard
11:26Z remove temporary copper heatsinks from helres
(TLi) New vented rf shield (0.1 mm copper) with 6-mm slots top and bottom:
ca 12Z move Pt100 ch. 6 (106) from rf shield to helical resonator cylinder, above adapter piece.
With old shield:
Signal generator (Vpp) Monitor (mVpp, average 32)
10 66.4 (without averaging ca 70)
New shield installed. RigExpert: min SWR 1.02 @ 14.4245 MHz, R = 52.1 Ohm, Q ~ 870 (SWR 2) ... 810 (SWR 3).
Signal generator (Vpp) Monitor (mVpp, average 32)
Both Q and monitor ratio as before within uncertianty.
ca 13Z ready
14:15Z RF OFF. will cool down by midnight, when computer turns ON.
RF siggen to 10Vpp, turn ON for testing
lab lights (partially)
ca 09Z Turn on stuff after x-mas break
For 2020 results see index_2020
For 2019 results see index_2019