ID10082 scan for peaks
fbeat 23 166 671 Hz
20:55Z no collision monitor running(?)
13:45Z leave collision-monitor, ID1 '2020-11-10-run1'. B-beam, no clearout
B-beam fluorescence levels 66.1 / 1.6 bright/dark in 10ms
Rabi flopping, B-beam:
ID10078: Start delay 500 ms.
ID10077: Start delay 100 ms.
ID10076: Start delay 1 ms.
Code cleanup day.
06:45Z: Stop collision monitor.
15:15Z leave collision-monitor, ID10005 '2020-11-09-run1'. B-beam, no clearout
B-beam fluorescence levels 50.5 / 1.8 bright/dark in 10ms
NOTE: this may not work, probing-cycle doesn't seem to get longer as more delay is added? ('start_delay' in the rabi v3 program)
ID10002, 200ms start delay
ID9998 +C1 rabi-flop, 100ms start delay
ID9996 +C1 rabi-flop, 50ms start delay
ID9994 +C1 rabi-flop, 1ms start delay
-C1 75 957 774, FWHM 166(?), h24.6
+C1 75 972 228, FWHM 270, h28.7
Peak-search with 3ms (Fourier FWHM 133 Hz AOM) probe and OD1
Rabi-flop on +C1, no OD-filter gives 1.1 ms tau-pi
B1 2.578 uT
C1 75 965 039 (line center)
-C1 75 957 805, FWHM 536. h36.6
+C1 75 972 273, FWHM 501, h40.3
14:26Z peak-search 800us probe
Turn off 4 GHz scope and extra PSUs causing air-currents at the setup
tests of 422/674 polarization.
-C1 h = 29.2
+C1 h = 35.4
ID9963 peak-search, 2 kHz range, 100Hz step, 800us probe
Initialize to these values, try again
B1_estimate = 2.646 uT
linecenter: 75 965 020
-C1 75 957 595 (957 685 initial guess), FWHM 519, h= 33.0
+C1 75 972 446 (971 829 initial guess), FWHM 561, h= 32.8
ID9959 try peak-search
set coils to zero (0,0,0) mA
12:07Z: Stop collision monitor.
ca 2500s per sweep
delete the stuck 9942 -> now started running 9944
clock laser unlocked - relock
ID9940 done, ID9942 clock-cycle v2 "running"-state, but nothing is happening? where did 9941 disappear?
15:57Z ion seems still stored. artiq stuck Friday evening?
ID9955 collision monitor, '2020-11-07-run1', starts automatically on 2020-11-08, B-beam, clearout off
Started automatically 2020-11-08 20:10:12.807841
14:50Z: Start coil characterization (see O:-drive readme_2020-11-06-B-field.txt).
Rabi flopping heating rate measurements on +C1 with B-beam.
No OD filter, 0.8 ms pi pulse, varying start delay.
ID9921: 1 ms
ID9923: 100 ms
ID9925: 500 ms
Strange noise in beat-note spectrum (at approx 21.5 MHz.)
leave a few wide scans running
also leave Toptica Pressure Compensation on (but it gives abrupt jumps in the piezo voltage?)
no narrow peaks seen - end.
23:12Z peak-search, now 6ms probe (rabi-flop indicated this might be tau_pi)
expect 133 Hz optical, 67 Hz AOM FWHM!?
+C1 FWHM 174 Hz
go back to ARC factor -0.5 V/V
clock-laser unlocked :(
23:02Z try again peak search
ion went dark for a while!?
22:39 +C1 spectrum again..
+C1 FWHM 205
ca22:36Z check spectrum of +C1. OD2, 8ms probe, 800 Hz span, 20 Hz step, 41 points
gaussian fit FWHM 217 Hz
22:21Z peak search
+C1 75 971 829, FWHM 224 Hz
-C1 FWHM 207
+C2 FWHM 240
-C2 FWHM 252
uMotion 100kphot: A 7.9% in 10s, B 7.5% in 21s, C 4.6% in 16s
22:12Z ID9874 peak search
+C1 FWHM 199 Hz - still too wide.
-C1 FWHM 179 Hz
+C2 FWHM 254 Hz
22:07Z Rabi-flop on +C1 seems to give incoherent result - still metro?
21:53Z run peak-search again
+C1 FWHM 233 Hz - still much too wide - metro noise!?
+C2 FWHM 245 Hz
Based on found C1-peaks, predicted positions are: [75.97184693 75.95769907 75.97887799 75.950668 76.00005692 75.92948908]
Repeat this until see Fourier-limited peaks?
21:41Z last westbound metro seems to leave Otaniemi on liikenne.live map
OD2 filter, 8 ms probe, DB '2020-11-05-OD2'
+C1 75 971 846, FWHM 199, h 18.2 (FWHM should be 100 Hz optical 50 Hz AOM?)
-C1 75 957 700, FWHM 191, h 26.5
C1 75 964 773
B1 2.521 uT
+C2 FWHM 235
-C2 FWHM 279, at 75 950 649
+C4 (no peak found?)
-C4 (no peak found?))
No ND filter, 800us probe. 4 kHz span, 150 Hz step -> 27 points per peak
ID9864 search peaks
+C1 75 971 849, FWHM 527, h 34.6
-C1 75 957 700, FWHM 524, h 54.2
C1 75 964 775
B1 2.52 uT
+C2 75 978 912, 474, h 42.5
-C2 75 950 705, 468, h 24.2
+C4 75 998 875, 280(?), h 3.14 (3) (maybe peak-fitting failed?)
-C4 75 930 811, -31(??), h0.16?? (fit fails?)
21:16Z relock clock laser. Set ARC Factor -1.0 V/V
For fun, leave servo running
ID9861, DB '2020-11-05-run1'
-C4 75 930 138, FWHM 591, h 33.2
+C4 75 999 442, FWHM 584, h 42.3
-C2 75 950 929, FWHM 521, h 25.0
+C2 75 978 636, FWHM 450, h 40.3
(TODO: improve program to output linecenter-estimate and B-estimate after finding +C1, -C1)
-C1 75 957 866, FWHM 550, height 43.9
+C1 75 971 714, FWHM 505, height 39.7
ID9860 Try peak-search, B-beam, 800us probe, to check everything works
4 kHz span and 150 Hz step, 27 points per peak
5 min fbeat 23 166 296 (26) Hz. ch1+ch2 sum check 100MHz (0.7 Hz) OK
Try lower ARC Factor -0.3 V/V instead?
uMotion 100kphot: A 7.4% in 10s, B 7.1% in 22s, C 4.7% in 16s
No!? jumped quickly to TEM01...
Try even more ARC Factor, to keep in lock? set to -1.6 V/V
decrease piezo-voltage to go to higher wavelengths
14:07 clock laser now in TEM00 at 1348.049 nm (two 500 MHz cavity-modes from the right mode?)
camera-program crash. restart
Leave C-beam, it has higher fluorescence than B.
ion still stored
06:10Z end clock-cycle run. clock-laser was unlocked at some point, now in TEM00, but wrong mode?
ID9852 new try, v3-clock-servo program, DB '2020-11-04-run2' with 800us probe. (previous clock-run was 2020-06-15)
Cycle time 64s, so for 6 peaks 10.7s per peak
100+100 shots per peak, so 53 ms per shot
NOTE: the 'all2' figure shifts the +/-C peaks closer to center, for better visual effect. peak-pair-centers not shifted.
Abort run, it was probably with 40ms probe...
ID9851 try clock-servo run with the found peaks, DB '2020-11-04-run1'
First clock-run with B-beam and the new helical resonator
B-beam, t_probe 800us, AOM att. 15 dB,
cycle time is ca 111 s, or 18.5 s per peak.
100 B + 100 R shots per peak, so 18.5/200 = 93 ms per shot - something wrong? maybe using t_probe = 40 ms now??
-C4: 75 930 045
+C4: 75 999 371
-C2: 75 950 875
+C2: 75 978 577
-C1: 75 957 766
+C1: 75 971 639
ID9850 Try again with improved values: center 75 964 751, and B = 2.479 uT
-C1 found at: 75.957 794
+C1 found at: 75.971 707
From first B-beam spectrum of the day, -C1 75.9577, +C1 75.97165
this gives center 75.964675 and B = 2.46uT (? morning estimate was 1.8 uT ?)
Second: 'Search for transitions v1'
First run: 'Calculate peak positions from B field v1'. This takes line-center and B-field as input.
try automatic peak search?
re-lock. try even more ARC Factor = -1.3 V/V
clocklaser unlocked gain, in TEM01 mode
A-spectrum peaks very low, max 30%
1 hour cavity-drift vs. maser ca +17 mHz/s. Driftcomp fixed at 2*(-10.45 mHz/s) = -20.9 mHz/s
note: two DDS-channels and splitter still connected to the Zeeman-AOM PA. Probing with DDS 3rd channel
11:44Z ID9834 run4, clock-spectrum with A-beam. Same settings as with B and C: 80 kHz span, 150 Hz step. 800 us probe
Fluorescence 139.8 / 2.0 bright/dark in 10ms
09:48 ID9831 run3, C-beam spectrum, same settings as ID9820/run2
C-beam counts 81.4/1.9 bright/dark in 10ms
Try a C-beam clock spectrum!?
uMotion 100kphot: A 8.9% in 10s, B 8.4% in 36s, C 6.7% in 15s
re-lock, try even more ARC Factor: -1 V/V
09:27 clocklaser jumped to TEM01
Note +Cn and -Cn peaks unequal height, so B-beam linear polarisation isn't so linear?
09:22 scan completed ok, 422 and 674 lasers stayed locked
08:25 ID9820 re-try the clock-transition scan.
duration 2400s = 40 min
Visually B-field ca 1.8 uT
B-counts 46.2/1.4 bright/dark in 10ms
08:23 re-lock Rb-line -> strong fluorescence again!
weak fluorescence with all beams ABC?
08:16Z ion dark moment during scan!? This gives 100% jumps...
Note AHM3 prediction updated to -3.1e-12 from latest atomic bulletin
ID9815 clock-transition scan around predicted center 75.964 MHz. span 80 kHz AOM, step 150 Hz. probe 800 us. randomized order
533 frequency points, duration ca 2000s
B-beam fluorescence a bit weak 44.3/1.4 bright/dark in 10ms
Try more slow gain on Toptica, to keep it in lock? ARC Factor set -0.7 V/V
set Zero coil currents: (0, 0, 0) mA
5 minute fbeat = 23 166 211 (31) Hz
Toptica ARC Factor = -0.5 V/V. Slew Rate Enable OFF. Set current ILD=167 mA. Piezo actual voltage 60.5V
clock-laser now in TEM03 mode, maybe ~1348.054 nm on wavemeter.
coil runs from the night failed, clocklaser probably unlocked early in the evening Tuesday
07:50Z ion still stored.
14:00Z: Start coil characterization (see O:-drive readme_2020-11-03-B-field.txt).
All experiments until now with zero coil currents.
ID9784: B-beam Rabi flopping in +C1, longer range.
ID9778: Detuning scan, B-beam
ID9776: B-beam fluorescence: bright: 506.0 dark: 13.5 in 100 ms
ID9775: B-beam Rabi flopping on +C1.
ID9759: Rabi flopping with B-beam.
674 nm power at A output: 160 nW
422 nm power at A output: 0.6 uW
ID9747: Clock sweep over +C1, A-beam with optimal tau_pi=0.8 ms as clock probe (Fourier 1000 Hz). Gaussian fit: center:75.9717 MHz , FWHM: 438 Hz AOM.
ID9744: Detuning scan, A-beam.
ID9743: A-beam fluorescence: bright: 1509.91, dark: 14.49 in 100 ms.
ID9740: Rabi flopping on +C1, A-beam.
Beat note 23.166161 MHz.
ID9738: run2, A-beam, wide scan. 1ms probe, 15 dB AOM att.
9:22Z Stop collision monitor.
15:29Z ID9729 collision monitor, '2020-11-02-run1'
leave collision monitor, B-beam, no clearout. Counts 55.38 / 1.3 bright/dark in 10ms
shut down clock-laser and comb
no peaks found. also clock-laser jumped out of lock...
ID9726. run6, A-beam, wider scan. 2ms probe, 15 dB AOM att.
fbeat = 23 167 582 Hz
ID9722 scan with A-beam (bright 120 counts/10ms). 1 ms probe
1/100 jumps across the spectrum, but no clear peak
B-beam gives ca 60 counts/10ms
ID9720 clock-transition search, now with some small 674nm power in B-beam
B-beam alignment needed after aligning the 674nm mirrors (why?)
ID9706 run2, clock-transition search, now with AOM att 15 dB - no quantum jumps, something basic is wrong.. no 674nm in the B-beam
Note the DDS-connections are now for bichromatic probing, with two DDS-channels via splitter to the AOM
2020-04-16 has notes om 674nm Zeeman AOM-efficiency vs. DDS attenuation
coil currents I1, I2, I3, all zero
13:31Z re-lock clocklaser
slow PDH servo, aka "ARC Fine In 1" is set to -0.5 V/V. Disabling the slew-rate limit possibly improves the servo lock?
Laser again jumped to TEM01 mode...
ID9705 try to find clock transition
1ms probe pulse, 23 dB AOM att. 75.965 MHz center, 10kHz span, 100 Hz step, random sampling
10:00Z stop collision monitor
old AHM3-model predicts linecenter at 75965976 Hz AOM
start clock-laser. Ilaser 170.9mA. I_amp 400 mA
Vpiezo 60.13 V gives 1348.052 nm on wavemeter
fbeat 23.2 MHz - but jumped to TEM01 mode.
SDR ch2 mixing-note is at 78.5 MHz
relock, fbeat = 23 167 608Hz
09:36Z ion still stored, collision-monitor running
09:06Z re-start collision monitor. B-beam, no clearout
uMotion 100kphot: A 5.9% in 8s, B 8.3% in 13s, C 5.8% in 13s
B-beam counts 95.95 / 1.47 bright/dark in 10ms
B-beam alone does give fluorescence, but weaker than A
09:00Z stop collision monitor, re-lock 422 to Rb-line. ion appears with all beams, but not with B?
Leave collision-monitor on. B-beam, no clearout
Fluorescence levels 54 / 1.6 bright / dark, in 10 ms
14:41Z start monitor.
Weird noise problems with comb line measurement...
About 07:20Z 422 unlocked. Stopping collision monitor.
Leave collision-monitor on. B-beam, no clearout
Fluorescence levels 74 / 1.4 bright / dark, in 10 ms
15:13Z start monitor.
ID9674 A-beam 3200 kphot (in 195 s) - fit fails again (now first bin overpopulated?)
ID9673 A-beam 1600 kphot (in 147 s) - note modulo(1/RF) histogram-fit fails, because last bin is overpopulated? why?
ID9672 A-beam 800 kphot (in 105 s)
ID9671 A-beam 400 kphot (in 53 s)
ID9670 A-beam 100 kphot (in 13 s)
use only A-beam where fluorescence is highest for this test
More photons should give better statistics for the g2-histogram?
ID9666 C-beam fluorescence levels: 619.63 bright, 13.89 dark in 100ms
ID9665 C-beam time-stamps, 100kphot
ID9664 C-beam detuning, 100ms, -23 +10 MHz
ID9662 B-beam fluorescence levels: 523.26 bright, 12.02 dark in 100ms
ID9661 B-beam time-stamps, 100kphot
ID9660 B-beam detuning, 100ms gate, -23 +10 MHz.
ID9657 A-beam fluorescence levels: 1001.29 bright, 13.69 dark in 100ms
ID9655 A-beam uMotion 100kphot, timestamps saved. t_pulses variable should have time-stamps
ID9655 A-beam detuning scan, 100ms gate, -23MHz +10MHz from nominal 208 MHz - no drop in fluorescence at the end? (218MHz is mid-resonance)
Note new RF-drive is 14.424 MHz and 10 Vpp to the tuneable resonator
Contrasts 100kphot: A 2.6%, B 3.6%, C 3.3%
Back to: -6.2, 0.2635, 27.0703, 46.7861
result is Optimal voltages (Common mode, diff): [-0.27877341 23.45517975 42.63258276] - but this seems wrong!?
ranges [0.2, 0.3], [26, 28], [45.5, 47.5]
12:07Z ID9649 run uMotion optimizer again, 100kphot, 4x4x4 grid
12:01Z uMotion with doors, 300kphot: A 2.1% in 38s, B 2.7% in 57s, C 3.1% in 59s
B-beam again required alignment after working with the objective-door...
install outer shield doors
uMotion with 2 shields, 100kphot: A 2.6% in 12s, B 2.5% in 18s, C 3.7% in 18s
Outer shield B-holes work OKish - maybe not worth changing the old shield.
B-beam required alignment after installing the outer shield
inner-shield tends to push the imaging objective, not all plastic M6 screws attached
11:30Z Install old shields
uMotion 100kphot: A 3.1% in 11s, B 3.5% in 13s, C 2.7% in 18s
ID9627 Optimal voltages set to (Common mode, diff): [ 0.2635054 27.0703494 46.78614235]
-6.2 V is the new V12cm value.
422 power at A exit is 1 uW! Reduced to 0.60 uW.
umotions all around 35% now!
Ion re-appeared after re-locking 422.
07:52:00Z: Stop collision monitor. 422 unlocked during the night.
15:26:00Z: Leave collision-monitor with B-beam with clearout OFF
63 bright, 1.3 dark counts in 10ms, without the holy-blanket
uMotions contrasts at optimum: A: 1.6%, B: 1.7%, C: 2.2%
ID9560: NPL uMotion minimization. Optimal voltages (Common mode, diff): [ 0.26023989 26.98180953 48.72613037]
Seculars (correlation): lower radial: 0.94373, upper radial: 0.98916, axial: 1.97352, ax/rad: 2.042042
Seculars (coarse): lower radial: 0.94372, upper radial: 0.98932, axial: 1.97368 ax/rad: 2.042047
umotions: A: 7%, B: 15%, C: 28%
Voltages: -6.2, 0.2214, 26.3703, 49.7532
Adjust V34com by + 1.3*0.1
V12cm to -6.2 V
Seculars (correlation): lower radial: 0.9407, upper radial: 0.98625, axial: 1.97595, ax/rad: 2.0509
Seculars (coarse): lower radial: 0.9406, upper radial: 0.98632, axial: 1.97632, ax/rad: 2.05127
umotion: A: 6%, B: 10%, C: 23%
Compensation voltages: -6.3, 0.2214, 26.3703, 49.7532
Adjust V34diff by -18e-3 * 0.8 V
A: 6%, B: 22%, C: 22%
umotions: A: 6%, B: 21%, C: 22.3%
Adjust V34cm by -1.3*0.8 V
Set V12cm to -6.3 V
Seculars with correlation (B-beam): lower radial: 0.95028, upper radial:0.99539, axial: 1.96657965, ax/rad 2.02149
Seculars after adjustment with old tickler: lower radial 0.9504, upper radial 0.9955, axial 1.96736, ax/rad 2.0220566318927
umotions: A: 9%, B: 8%, C: 8%
Adjusted V12_common to -6.0V, and compensated accordingly by decreasing V34_common by 1.3/2 = 0.65 V
Seculars using old tickler program: 0.9654, 1.0098, 1.9519, ax/rad = 1.975, had to use a lot of RF power (3 dB) on upper radial???
Adjust A beam 422 power to 0.3 uW.
uMotion: A: 4.1%, B: 4.4%, C: 5.5%
ID9435: NPL optimization over lunch: comm-diff: [ 0.25898505 27.87242426 49.79449256]
Optimized beam directions after umotion compensation
Coarse micromotion compensation done by moving iteratively in cardinal directions.
New voltages: [0.26711, 28.05231, 49.42779] A = 8%, B = 13%, C = 17%
uMotion contrasts: A: 50%, B: 34%, C: 47%
Previous compensation voltages (comm-diff): [-5.5, 0.3043, 27.7037, 47.2835]
Optimize beam positions. All beams ok after adjustment.
Blink test: 2 levels.
Detuning set to 208 MHz.
08:18:00Z start load, dispenser 1.60 A. Ion! after 3 min 30s.
08:03:00Z start load, dispenser 1.60 A. No ion after 13 min.
422 detuning set to 185 MHz for loading.
15:17:30Z ion! in only in one minute!? at least two ions, at roughly center of camera (so camera-alignment to electrodes worked ok)
15:16:30Z try again, after re-aligning camera without inner B-shield
Image of inner electrodes is with higher magnification than before. Top and bottom electrodes found after some
14:28:30Z start load, dispenser 1.60 A. No ion after 19 min.
461 dispenser checked and adjusted to 198.65 mA, 4.645 kohm
14:09Z start load, dispenser 1.60 A. No ion after 15 min.
12:36:00Z start load, dispenser 1.60 A (1.468W). ASE on at 12:40:40Z. No ion after 14 min (12:54:15Z).
405 spontaneously switched off during warm-up for couple of seconds!
12:05: 422 power approx 0.6 uW in A beam exit. 461 power approx 150 uW.
12:00Z: Check 461 dispenser. Ok with I = 199.21 mA, R = 4.645 kohm.
Resonator adjusted to 14.424 MHz
08:55Z KSK 105 AC increased from 20Hz to 25 Hz - to make lab cooler
ion disappeared again while installing the inner shield..
new B-holes in the outer shield seem ok, beam enters and exits ok
after some dark moments the ion re-appeared suddenly... (Rb-lock checked, ionization lasers tried, etc.)
detuning was at 185MHz, but now back to 208 MHz as ion re-appeared
camera focus-micrometer screw position at 6mm + 300um ("30") gives sharp ion image. removing the objective requires defocusing by 4-6 turns (2-3mm)
lost ion (?) while installing old inner-shield (now with new B-beam holes)
Cardinal voltage directions do not seem very stable or repeatable!? Why?
Maybe voltage-ranges for uMotion scan needs to be optmized further?
0.2372 2.3100 1.9737
0.4131 -1.5132 -0.0264
0.6644 1.4648 -3.8875
0.0655 1.7460 1.6995
0.2458 -0.9212 0.3168
0.3429 1.0064 -2.8921
inv(Cfit_2020-10-01) = (new B-beam direction)
0.0159 0.2947 0.2067
2.0259 -1.1843 1.3154
1.9228 -0.1466 -3.5292
at this point uMotion(100kph): A 5.2% in 16s, B 13.4% in 22s, C 3.7%% in 23s
Optimal voltages [ 0.30428835 27.70370948 47.28350979]
08:56Z ID9327 ranges: 0.25-0.35, 27.4-27.8, 47.1-47.5 V
Try again around the new optimum?
At this point uMotion 100kphot: A 2.4% in 23s, B 11.1% in 27s, C 0.8% in 27s
08:53Z done. Optimal voltages (Common mode, diff): [ 0.29125006 27.67571599 47.28898777] (V12cm -5.5 V as usual)
08:42Z point 56/63
07:32Z ID9320 uMotion optimization, 4x4x4 grid, ranges: 0.2-0.3, 27.25-27.75, 47.9-48.5
Duration (22+25+25) s/point, total duration ca 77 minutes (finish by 11:45 local?)
uMotions (100kphot): A 17.8% in 21s, B 23.5% in 24s, C 23.5% in 27s
currently voltages: 0.2495, 27.5755, 48.2091
B-beam bright/dark counts now 44/1.2 in 10ms
07:26Z stop collision-monitor (this one started 20:00local on Friday, there was a second collision-monitor in the queue - intentional?)
07:24Z ion still stored (9+ days, longest so far?)
07:23Z ASILive camera-program on win-machince crashed. restart.
Scheduled to start 17:00Z:
Leave collision-monitor with B-beam with clearout OFF
52 bright, 1.3 dark counts in 10ms, using the holy-blanket
422 unlocked from Rb-line? lights on or working in the lab causes this?
10:10Z Stop collision monitor.
07:10Z ion still stored, collision-monitor running.
This ion was loaded 2020-09-25 Friday at 11:54:50
14:42Z leave collision-monitor with B-beam and both ASE-sources
56 bright, 1.3 dark counts in 10ms, using the holy-blanket
ID9301 DB '2020-10-01-run-1'
(compare to A-beam result from 2020-05-04)
ID9280, A beam 1000 ms gate ASE on/off at 208.08 MHz
ID9284, A beam cooling lineshape scan, 1000 ms gate, starting point 208.08 MHz
ID9286, B beam 1000 ms gate ASE on/off at 208.08 MHz
ID9287, B beam cooling lineshape scan, 1000 ms gate, starting point 208.08 MHz
ID9289, C beam 1000 ms gate ASE on/off at 208.08 MHz
ID9290, C beam cooling lineshape scan, 1000 ms gate, starting point 208.08 MHz
A beam power: before chamber 0.47 µW, after chamber 0.38 µW
B beam power: before chamber 0.50 µW, after chamber 0.45 µW
C beam power: before chamber 0.48 µW, after chamber 0.42 µW
Vopt [ 0.24950152 27.57550502 48.20908736]
at this point 100kphot contrasts: A 1.6 % (20s), B 1.6% (26s), C 1.1% (30s)
The columns of inv(Cfit) are the voltage directions VA, VB, VC, that change only one beam-contrst by 100%.
(Compare to 2020-04-23 and 2019-12 results)
When manually searching for uMotion-null, maybe 10% or 20% of these vectors is a useful step.
For the least-squares optimization, maybe 50% of these vectors is a good range?
inv(Cfit_2020-10-01) = (new B-beam direction)
0.0159 0.2947 0.2067
2.0259 -1.1843 1.3154
1.9228 -0.1466 -3.5292
0.1235 0.3759 0.4353
2.1165 -1.2290 1.1596
1.8881 0.4130 -2.7941
0.1239 0.4418 0.4723
2.4101 -1.6631 1.0045
2.4619 0.5661 -3.6602
08:25Z progress at 42/63. duration per point is (20+27+30) s -> will finish in ca 27 minutes
uMotion-optimization, to find the new VA, VB, VC directions. Currently at [0.2555, 27.2649, 49.0708]
smaller scan range 0.2-0.3, 27.0-27.5, 48.75-49.25, 4x4x4 pts scan, 100 kphotons/pt
optimize beam-directions - but only slightly improved fluorescence levels
07:20Z stop collision-monitor
16:09Z leave lab
16:04Z leave collision-monitor with the new B-beam. 50/3 bright/dark counts in 10ms gate. Both ASE-sources on
Note first start of program gave ARTIQ-error, second run OK
axial might be at 1.9485 MHz.
1.947-1.950 MHz, 27 dB att, 50 pts, A, B, C is ID9244, 9245, 9246
measure secular frequencies and FWHMs, with new B-beam direction and good micromotion
Lower radial seems lower than before around 0.9606 MHz
0.960-0.962 MHz, 15 dB att, 40 pts, 40 kphot/point. A, B, C is ID9208, 9209, and 9210
1.003-1.007 MHz, 13 dB att, 50 pts, 40 kphot/point. A, B, C is ID9211, 9212, and 9213
1.941-1.945 MHz, 27 dB att, 50 pts, 40 kphot/point. A, B, C is ID9214, 9215, and 9216
The Cfit matrix, and inv(Cfit) (whose columns are the va,vb,vc voltage directions) is similar to previous results, but maybe not correct, since the optimum voltage is outside the scan range
Optimum at: [ 0.2337378 27.2069111 49.21050508]
Contrasts at new optimum: A: 2% (22s), B: 5% (25s), C: 13.2% (32s)
Move along Vc to [0.255465, 27.26488,49.070795]
Now A: 2.1, B: 4.1, C: 8% !
08:53 still running, point 37/63... duration per point (20+24+29) s = will finish in about 30 mins...
Try automatic search around manually found (roughly AB-optimal) point 0.3876 27.653 47.4867
ID9129 search 4x4x4 grid 0.28-0.48 26.5-28.5 46.5-48.5, 100kphot
Back to A-optimal point [0.2601, 29.1, 46.0387]
From 2020-04-16 we have voltage directions that change only one beam-contrast
These vectors change contrast by 100%, so use only a fraction of them
VA04 0.1235 2.1165 1.8881
VB04 0.3759 -1.2290 0.4130, useful change is maybe 10% of this, 0.1*VB = vb = (0.04, -0.12, 0.04) V
VC04 0.4353 1.1596 -2.7941, 0.1*VC = vc = (0.04, 0.12, -0.28)
If we move along VB or VC, this should not change the A-contrast!?
at [0.2601, 29.1, 46.0387] A 5.8%, B 66%, C 70% - good, we are at A-optimal point
move +vb to 0.29859 28.9771 46.08 - get: A 5.0%, B 62%, C 69% - good, A did not change
move +vb to 0.3362 28.8542 46.1213 - get: A 5.1%, B 56%, C 70% - B-seems to change, so more of the same
move +2vb to 0.4114 28.6084 46.2039 - get A 4.0%, B 39.6%, C 70.6 % - more vb...
move +2vb to 0.4865 28.3626 46.2865 - get A 1.6%, B 17%, C 70.4 % - more vb
move +2vb to 0.5617 28.1168 46.3691 - get A 0.6%, B 8.6%, C 70.6%
The larger than previously V12diff is probably wrong, so the right move to optimize C is now -vc!?
move -2vc to 0.4747 27.8849 46.9279 - get A 5.1%, B 2.0%, C 62% - seems to be the right direction?
move -2vc to 0.3876 27.653 47.4867 - get A 10.6%, B 3.5%, C 47%
C is largest, so optimize it?
no clear minimum found by scanning small ranges around current set-point..
at [0.2601, 29.1, 46.0387] get(100kphot): A 5.8% in 23s, B 66% in 25s, C 69% in 27s
try scanning V34diff, with A-beam, only 10-20kphot to make it faster
A-optimum at ca V34diff 29.1V
Try older values from 2020-09-24, optimal values [ 0.26014207 28.08051673 46.03867207]
Here contrasts: A 39% in 24s, B 57% in 25s, C 59% in 28s
At the new optimal point, contrasts:
100kphot: A 55.5 % (23s), B 70%, C 66% in 35s(blurry weak image)
scan-range was 0.15-0.35, 26.5-29.5, 47.5-60.5, 6x6x6 grid
uMotion optimizer suggests Optimal voltages (Common mode, diff): [ 0.20202339 24.60426097 57.7988483 ]
07:06Z ion still stored.
ID9133: Overnight NPL uMotion optimization search 6x6x6 grid, 100k photons.
Optimize B-beam intensity and measure seculars.
Axial: 1.942-1.945 MHz, 27 dB att, B ID9121
Higher radial 1.0055-1.0085 MHz, 13 dB att, B ID9128
Lower radial 0.961-0.964 MHz, 15 dB attenuation, B ID9127
uMotion contrasts now 65-75%!!!
Relocated B-beam 10deg closer to camera axis, increased height 160 -> 185 mm
Comparison Tuesday (Monday):
Axial FWHM: A, B, C: 228, 294, 270 Hz (135, 361, 412) Hz
Upper radial FWHM A, B, C: 995, 371, 425 Hz (899, 434, 514) Hz
Lower radial FWHM A, B, C: 447, 671, 379 Hz, (445, 1023, 353) Hz
Axial 1.948-1.961 MHz, 27 dB att, ABC ID9039 9040 9041
Higher radial 1.006-1.009 MHz, 13 dB att, ABC ID9036 9037 9038
Lower radial 0.963-0.965 MHz, 15 dB attenuation, ABC ID9033, 9034, 9035
Repeat tickler secular sweeps, to compare with results from Monday
06:45Z stop collision-monitor. ion still stored. lots of collision events during the night?
14:12Z start collision-monitor
A-beam 65 bright / 1 dark counts per 10 ms. Both ASE-sources on
ID9028 C-beam, 15 dB att,
ID9027 B-beam, 15 dB att,
ID9026 A-beam, 15 dB att, maybe slight blur, diffraction-rings visible at resonance,
repeat lower radial with more attenuation, 40 kphotons/point, 0.962-0.965 MHz, 50 points
ID9025 C-beam 13 dB att, c-max 23%, FWHM 739 Hz, Lorentzian fit: [2.23711492e-04 1.00740753e+00 3.69715782e-04]
ID9024 B-beam 13 dB att, c-max 17%, sharp image, FWHM 727 Hz, Lorentzian fit: [1.50025783e-04 1.00741820e+00 3.63464740e-04]
ID9023 A-beam 13 dB att, 1.006-1.009 in 50 steps. 13 dB att. c-max 13%, FWHM 1327 Hz, Lorentzian fit: [2.35154904e-04 1.00739944e+00 6.63353354e-04]
ID9022 A-beam 13 dB att, higher range 1.006-1.008 MHz. this is the broadest resonance. c-max only 13%. Image stays sharp through resonance. Direction might be along the camera axis??
ID9021 A-beam, 13 dB att
Now higher radial 1.005-1.006 MHz scan
ID9020 C-beam 13 dB att, slight blur (more than A-beam?), c-max 39%, FWHM 471 Hz, Lorentzian fit: [2.63558885e-04 9.63508909e-01 2.35579404e-04]
ID9019 B-beam 13 dB att, c-max 22%, FWHM 696 Hz, Lorentzian fit: [2.31554074e-04 9.63493373e-01 3.47986538e-04]
ID9018 A-beam 13 dB att, 0.962-0.965, 50 pts. maybe slight blur at resonance, c-max 28%, FWHM 619 Hz, Lorentzian fit: [2.37744064e-04 9.63548187e-01 3.09470795e-04]
ID9017 lower radial 0.963-0.969 40 points, 13 dB att, A-beam
Run all three beams with 27 dB attenuation, and more photons for better statistics
ID9014 A-beam. 27 dB, FWHM 134 Hz
ID9015 B-beam. 27 dB, FWHM 361 Hz
ID9016 C-beam. 27 dB, FWHM 412 Hz
Summary of 9005, 9008, 9011 Axial mode FWHM comparison:
A-beam 27dB att, faint blur, c-max 15%, FWHM 222 Hz
A-beam 25dB att, still blurry, c-max 18%, FWHM 274 Hz
A-beam 23dB att, 660 counts / 100 ms gate. Blurry! c-max 21%, FWHM 288 Hz
B-beam: 23dB att, 660 counts / 100 ms gate. c-max 23%, FWHM 439 Hz
C-beam: 23dB att, 350 counts / 100 ms gate. c-max 30%, FWHM 477 Hz
ID9013 A-beam, 27 dB att, slight blur remains, diffraction-rings faintly visible through resonance
Lorentzian fit: [5.21626436e-05 1.94964964e+00 1.10884548e-04]
ID9012 A-beam, try 25 dB att, still a blurry image at resonance
Lorentzian fit: [6.73627956e-05 1.94959299e+00 1.37258450e-04]
ID9011 A-beam, 23dB att. max contrast 21%, image is blurry at resonance! asymetric shape.
Lorentzian fit: [1.01563463e-04 1.94961471e+00 1.44198864e-04]
ID9008 B-beam, 23dB att. max contrast 23%.
Lorentzian fit: [1.43360487e-04 1.94959719e+00 2.19437360e-04]
ID9005 C-beam, 23dB att (same as ID9003). Max contrast 30%, 1s exposure camera image stays ~sharp during resonance.
Lorentzian fit: [1.68185078e-04 1.94965074e+00 2.38378880e-04] FWHM 477 Hz (why higher than 9003 below?)
Compare beam fluorescence-levels, and FWHM at a low tickler level. Axial mode.
ID9003 23 dB att, [1.74523068e-04 1.94961955e+00 1.95524956e-04] FWHM 391 Hz
ID9002 22 dB att, [1.78054036e-04 1.94969403e+00 1.87024488e-04] FWHM 374 Hz
ID9001 21 dB att,
ID9000 20 dB att
ID8999 19 dB att
ID8998 18 dB att
ID8997 17 dB att (still asymmetric resonance) Lorentzian fit: [3.62632810e-04 1.94959744e+00 2.47898271e-04]
ID8996 16 dB att (blurry image at resonance, asymetric/hysteretic shape?) 1.9485-1.951 MHz, 40pts. Lorentzian fit: [3.86085904e-04 1.94972414e+00 2.58231443e-04] FWHM 516 Hz
ID8995 C-beam, axial tickler 1.948-1.951, 50pts, 20kphot. 20+15dB att. Lorentzian fit: [4.28908909e-04 1.94968264e+00 2.58508669e-04] FWHM 517 Hz
Does tickler drive-level affect the observed FWHM?
08:35Z stop collision-monitor.
15:13Z leave lab.
15:09Z try to start collision monitor. Now it seems to start ok!? (see Friday below for data)
A-beam, 10ms gate, 70bright/1dark counts. 2020-09-26-run1 DB tag. Both ASE-sources are on
Lower radial FWHM (A, B, C): (572, 809, 468) Hz (Friday result 737 Hz for A-beam)
Upper radial FWHM (A, B, C): (1295, 633, 604) Hz (Friday 1414 Hz A-beam)
Axial FWHM (A, B, C): (315, 508, 656) Hz (Friday 259 Hz A-beam)
Axial mode, 15 dB att, scan 1.947-1.951 MHz
ID8989 A-beam (at resonance camera image blurry - too high tickler level?)
print:Lorentzian fit: [2.04808225e-04 1.94986872e+00 1.57526180e-04]
ID8990 B-beam, axial mode
print:Lorentzian fit: [3.34966326e-04 1.94984179e+00 2.54115110e-04]
ID8992 C-beam axial mode. (at resonance camera image blurry)
print:Lorentzian fit: [4.57426818e-04 1.94979897e+00 3.27820460e-04]
Higher radial mode
ID8984 1.005-1.007, A-beam. too low range? try ID8985 1.006-1.008 MHz with 11 dB attenuatio, prelim Lorentzian fit: [2.31810846e-04 1.00751707e+00 5.25566911e-04] but scan-range still too low
ID8986 1.006-1.008, A-beam, 11dB, 40 pts
A-beam upper radial Lorentzian fit: [2.59266870e-04 1.00753200e+00 6.47697943e-04]
ID8987 B-beam, 11dB att, 1.006-1.008 MHz
print:Lorentzian fit: [1.68944514e-04 1.00753194e+00 3.16734569e-04]
ID8988 C-beam, 11dB att, 1.006-1.008 MHz
print:Lorentzian fit: [2.57580818e-04 1.00752836e+00 3.01935747e-04]
parameters are [A, f0, g] in the lorenzian A*(1/(g*np.pi))*( g**2 / ( ( f-f0 )**2 + g**2 ))
FWHM = 2g
print:Lorentzian fit: [2.39217324e-04 9.63620925e-01 2.43054054e-04]
print:Lorentzian fit: [2.20101447e-04 9.63596411e-01 4.04463962e-04]
ID8981 tickler-sweep, A-beam, lower radial mode 0.962 to 0.965 MHz, 50 pts. 13 dB att. center. 20 kphotons/point
print:Lorentzian fit: [2.22614323e-04 9.63654977e-01 2.85932046e-04]
From Friday expect the modes to be at 0.963 (13 dB att), 1.006 (11 dB att), and 1.948 MHz (15 dB att)
14:06Z ASI camera program crashed. re-start. ion seems still stored.
No collision monitor active in ARTIQ? error message
12:50: Overnight secular sweeps around 0.962 MHz: 40kPh, 13 + 20 dB att. Ca 10minutes per sweep. ID8879-8979 (approx. 18hr). Save Lorentzian fits to DB (desc 2020-09-26-weekend).
Afterwards dark-state monitor on over the weekend (Starts automatically around noon on Saturday):
Fluorescence level for A-beam ca 70 counts/10ms
Correlation seculars (20kPh, 20 + 13 dB): Radial: ~0.9626 MHz: ID8873
(20 kPh 20 + 11 dB) Radial: ~1.0064 MHz: ID8874
Axial: ~1.948 MHz: ID8875, 15 + 20 dB; ID8877: 18 + 20 dB.
Drop 422 power to 0.25 µW
Seculars with old sweep program (A-beam): Radials: 0.9625 MHz, 1.0064 MHz. Axial: 1.9482 MHz
11:39:00Z Load ion. Idisp 1.6A (1.44W). Ion! at 11:54:50.
Keysight 33522b CH2: 14.404 MHz, 10Vpp (23.8 dBm on Siglent SA), monitor voltage Vpp: 67-68 mV (previously 71-73 mV)
11:30Z Switch RF drive to straight from generator
08:52Z still 3 runs of the secular-sweep in the queue... last one is ID8837.
Results of the overnight monitoring of the lower radial secular frequency
Secular frequency is anti-correlated with lab-temperature, because of RF PA gain-variations?
14:00Z: Overnight secular sweeps around 1.019 MHz: 40kPh, 13 + 20 dB att. Ca 10minutes per sweep. ID8737-8837
Seculars with correlation:
Correlation seculars (20kPh, 20 + 13 dB): Radial: ~1.0187 MHz: ID8718
(20 kPh 20 + 11 dB) Radial: ~1.0606 MHz: ID8727
Axial: ~2.0 MHz: ID8735: 18 + 20 dB more sampling; ID8734: 23 + 20 dB; ID8733: 18 + 20 dB; ID8732: 15 + 20 dB
Seculars vs drive amplitude:
These four drive-levels correspond to trap parameters
az = 0(1)e-4
trap asymmetry epsilon = 0.0190(8)
trap voltage V0 =
V0/Vgen = Vratio = 12680(20)
From HelRes-characterization (Q, kappa), and PA-gain -> Vratioexp = 12283
Vratio/Vratioexp = 1.03
Timestamps saved at secular 1.019 MHz: 10k: ID8711, 200k: ID8712, 400k: ID8713, 800k: ID8714, 1M: ID8715
try lower 422-power, 0.2 uW
422 power in A-beam ca 450nW
ID8708: Attenuation: 3 + 13 db, 20 kPh per point
ID8706: Attenuation: 3 + 13 dB
10:45Z: Photon timestamp correlation seculars (10kPh per point): Radial around 1.02 MHz: ID8705. Attenuation: 3 dB + 20 dB.
Broad peak, too much power?
08:50Z ID8676 uMotion-optimization over lunch
4x4x4=64 pts. 100kphot
Optimal voltages (Common mode, diff): [ 0.26014207 28.08051673 46.03867207]
10:00Z: uMotion with new optimal values (100kphot) A 3.5%, B 1.4%, C 1.7%
Previous values: [-5.5, 0.2788, 28.0406, 44.7618]
08:42Z: uMotion (100kphot) A 23.9%, B 14.6%, C 25.1%
axial secular freqs with old tickler program:
60 mVpp, 2.0757 MHz (25 dB attenuation, 2.074-2.078 sweep/200pts, 50ms gate). Visible motion on camera (0.5s exposure)
61 mVpp, 2.11372 MHz (26 dB att)
62 mVpp, 2.15234 MHz (26 dB att)
63 mVpp, 2.19086 MHz (23 dB att)
and now radials: (prediction is ca 49% and 51% of axial?)
63 mVpp, 1.07364 MHz (13 dB, broad resonance) and 1.11328 MHz (10 dB att, A-beam cooling stronger, so need more tickler?)
62 mVpp, 1.0558 MHz (13 dB), 1.0959 MHz (10dB)
61 mVpp, 1.0376 MHz (16 dB), 1.0785 MHz (9dB)
60 mVpp, 1.020 MHz (15 dB), 1.0614 MHz (11dB)
uMotion 100kphot (note RF 14.404 MHz), A 21.5%, B 13.6%, C 21.8%
Repeat: A 20.1%, B 12.1%, C 24.6%
06:17:00Z stop collision monitor
2020-09-23 Wednesday, new Magic Helical-Resonator
14:30:17Z start collision monitor with clearout on!
Fluorescence level for A-beam ca 76 counts/10ms
60mVpp: 2.072 MHz axial, with 21 dB attenuation
60mVpp: 1.019 (scan 1.01-1.03 = 20 kHz and 401 points) and 1.0608 MHz (7dB DDS attenuation)
Lower RF-drive to 60mVpp from siggen (this was predicted drive-level for maintained secular freqs)
Seach for secular freqs
Reapeat uMotion after optimizations. 100kphot
A 7.6%, B 2.5%, C 8.7%
B gives no signal, needs alignment, after alignment 120counts/10ms
C first 40counts/10ms, optmize to 60counts
uMotion 100kphot (note RF 14.404 MHz), A 8.2%, B 3.6%, C 15.2%
optimizing ASE-direction increases to 80counts/10ms! further align, to 100counts/10ms
PMT signal: move camera in X/Y, got ca 50counts/10ms
Focus camera in Z-direction, to get similar ion-image as before (still far left and up in image)
increase RF siggen to 90mV, now a stable and bright image? Detuning 208MHz now.
11:03Z try to load ion. Idisp 1.6A (1.4W)
All beams ABC on
11:18Z weak signal (far left in camera image)? disappears?
Previous bias-voltages set: (-5.5, 0.2788, 28.0406, 44.7618) V
B-coils zero (0,0,0) mA
405nm, adjusted to 2.6mW (after warmup)
Cooling-laser Rb-signal OK 2Vpp
ca 0.4 uW in A-beam. 185 MHz detuning
461nm checked against dispenser. R=4.646 kOhm, I=199.21 mA
ca 187 uW in A-beam
New tuneable helical-resonator on magic-RF 14.404 MHz
Monitor signal (SMA-connector stub) is weak, 60 mVpp from signal-gen, thru RF-amp, gives 73 mVpp monitor
Monitor components: SMA-connector, SMA L-piece, SMA-BNC adapter, probe-adapter, Keysight 10x-probe
shut down lasers
13:14Z ion lost while placing holy-blanket?
11:50Z stop collision-monitor, to test 3d printed shield
08:10Z camera software on win machine crashed? restart.
13:26:15Z start collision monitor with clearout on!
Fluorescence level for A-beam ca 70 counts/10ms
Tried to optimize B-beam fiber input coupler to align polarization axis. All beams aligned OK for power. Need to check polarization.
09:45:00Z Stop collision monitor.
08:34Z (visit lab to check all is ok), collision monitor running without B-shields (3D printed inner shield fitted)
14:36:40Z start collision monitor with clearout on!
Fluorescence level for A-beam ca 70 counts/10ms
B-beam aligment lost! A and C ok.
3D printed inner shield, camera side: all beam-holes ABC need to move closer to camera hole
reduce 422nm in A-beam to 200nW, this gives bright counts 70/10ms
A-beam uMotion 6.7% (100kphot)
with the high loading intensity, A-beam bright counts 172/10ms (0.5/10ms dark)
set detuning to 208 MHz
12:25Z start load, dispenser 1.60 A
RF settings not changed, 70 mVpp from siggen, 1.67 Vmonitor
422nm detuning 185 MHz, maximize power to ca 0.8 uW in A-beam
12:21Z load ion, before removing B-shields (in order to test-fit new design 3D printed dummyshields)
06:05Z: Graphics card driver crash during weekend, stop collision monitor.
13:23:10Z start collision monitor with clearout on!
Fluorescence level for A-beam ca 70 counts/10ms
ID8292, ID8293, ID8294: Bichromatic line center scan, tau_pi = 2300 us, AOM att 23 dB, center: 75.96446 , split: 0.0052616 MHz
Center: 75.9644626 MHz, FWHM: 211 Hz AOM, height 63 (%)
Select tau_pi = 2300 us.
ID8290: Linecenter scan on -C1, FWHM Hz AOM
ID8288: Linecenter scan on +C1, FWHM 207 Hz AOM (optical Fourier limit 348 Hz).
Rabi flopping tests with increasing attenuation.
ID8284: att 18 dB
ID8286: att 23 dB
AOM amplifier goes to compression with two signals -> reduce RF power.
ID8251: AOM att 18, tau_pi = 1100 us
ID8250: AOM att 15, tau_pi = 1100 us
ID8249: AOM att 15, tau_pi = 800 us
ID8248: Bichromatic line center scan. Linecenter: 75.964429 MHz, separation: 0.005395883 MHz (2020-08-21-run7), AOM att 6. Tau_pi = 800 us
ID8244: Linecenter scan on -C1 (2020-08-21-run6)
ID8243: Linecenter scan on +C1 (2020-08-21-run5)
ID8241: Rabi on +C1:
ID8237: Bichromatic line center scan with 2.5 kHz separation
ID8236: Bichromatic line center scan. Linecenter: 75.964397 MHz, separation: 0.005428 MHz
ID8224: Rabi on +C1, corresponding to AOM att 18 dB. tau_pi = 1100us:
ID8229: -C1 linescan with tau_pi = 800 us
ID8227: +C1 linescan with tau_pi = 800 us
ID8224: Rabi on +C1:
ID8211: I: (0,0,5) mA (green), ID8213: I: (0,0,-5) mA (red).
08:09Z Stop collision monitor.
13:21Z start collision monitor with clearout on!
Fluorescence level for A-beam ca 70 counts/10ms
13:30Z Start coil characterization. Description on O:-drive (2020-08-19-coil-characterization).
10:03Z Stop collision monitor.
13:21Z start collision monitor with clearout on!
Fluorescence level for A-beam ca 70-90 counts/10ms
ID8078: Rabi on +C2:
ID8076: clock-spectrum scan. C1s are maybe very low, C2s visible. B-field direction has changed over summer?
Fit to previous spectrum gives 1.7 uT
C-contrast 1.3 %
B-fluorescence changes from 30 to 110 counts / 10ms when heating the B-fiber with one finger
A-beam uMotion 3.7% (much better than before). So this is a single ion!?
detuning to 208 MHz -> much brighter image
12:38Z load with 1.60A
422-detuning to 185 MHz
check 461nm against helper-dispenser, 199.18 mA, 4.646 kOhm.
seem to have trapped one bright one dark ion? uMotion correlation-histogram not sine-shaped. 60% contrast. Bright ion can sometimes suddenly move sideways on camera image
12:12:30Z try to load ion, dispenser 1.65 A, 1.6W, Ion! after 5 min.
12:01:00Z try to load ion, dispenser 1.6 A, 1.5W, no ion after 10 min.
11:49:30Z try to load ion, dispenser 1.6 A, 1.5W, no ion after 10 min.
100kphot: A 68%, B 66% in s, C 68%. Maybe multiple ion? Blink test shows only 2 levels. Maybe dark isotope?
Check 461 dispenser, OK, I = -198.93 mA.
10:26:00Z try to load ion, dispenser 1.65 A, 1.6W, Ion! at 10:35Z (9 min)
Maximize 422 power (now 0.4 uW at A exit) and try again:
10:12:00Z try to load ion, dispenser 1.60 A, 1.50W, no ion after 10min.
09:58:00Z try to load ion, dispenser 1.60 A, 1.50W, no ion after 10min.
Ion lost around 8:00Z
ID8025 No ND line-scan with 600us probe
50 kHz span, 150 Hz step, duration ~1500s
06:34Z stop collision monitor
2020-08-17 Fall-season opening
13:55Z start collision monitor
422-amp to 1.8 A, fluorescence level for A-beam ca 70-90 counts/10ms
13:30Z: ID8015: Clock sweep at expected center, 25 kHz, no filter, 150 Hz step. C1 separation: 75.9747 - 75.954 = 20.7 kHz AOM, 41.4 kHz optical.
B field = 7.4 uT ?!?!? (6-fold larger than old field of 1.3 uT)
13:00Z: Micromotion contrasts:
Compensation voltages: V1 = -5.7788, V2 = -5.2212, V3 = -16.7212, V4 = 72.8024, Com-Diff: (-5.5, 0.2788, 28.0406, 44.7618)
100kphot: A 2.9% in 12s, B 2.3% in 14s, C 3.3% in 14s
12:32:30Z try to load ion, dispenser 1.60 A, 1.50W
12:37:00Z ion! only 4min 30s loading time
ASI-camera usual settings 968x548 pixels (Bin2), High-gain, 1s exposure, histogram stretched/zoomed
bias-voltages not touched
coil-currents not touched, all zero (I1, I2, I3)
RF-drive 70mVpp 16.770 MHz, monitor 1.67 Vpp
12:29Z check A-beam power levels.
405nm 2.7 mW (reduced from max 6-7 mW)
461nm 200 uW
422nm 0.4 uW
07:40Z Current driver changed for 461 nm. Good parameters: I = -198.68 mA, T = 4.646 kohm.
13:05Z stop experiment for summer break
300kphot: A 7.1% in 85s
300kphot: C 107s, last bin of histogram overpopulated!? why?? can't fit sine-wave to histogram
100kphot: A 7.6% in 28s, B 4.7% in 71s, C 11.4% (!) in 42s
Check micromotions over lunch - did not work? - low counts only 40/10ms for A-beam?
09:30Z ion still stored. clock-laser in TEM01.
|no filter||709 Hz||
|ND1||256 Hz||2.1 ms||0.807 Hz||0.598 mK
|ND2||109 Hz||5.8 ms||12 Hz ?||2.3 mK ?
|ND3 ID7959||28 Hz||20.6 ms (note 10x ND1 result)||11 Hz||1.15 mK
|ND3 ID7961||31 Hz||20.5 ||2.6 Hz||2.1 mK
|ND3+ND1 ID7967||10 Hz||61.8 ms||0? Hz||2.2 mK
leave scans with ND3+ND2 and 180ms over night
(ion has dark moment now - let's hope it recovers...)
ID7970 ND3+ND2 flop-on-center
low Pexc, not hitting the resonance??
ID7969 ND3+ND2 line-scan with 180ms probe
40 Hz span, 1Hz step, 41pts * 22s/100shots = duration 900s
only 24/100 jumps at most?
ID7967 ND3+ND1 flop 2nd try at lower freq 75.968039, 100shots
ID7965 ND3+ND1 flop-on-center 75.968045, from 0.05 to 300ms in 20 steps
low excitation? not hitting the carrier frequency?
ND3+ND1 predict 21ms*sqrt(10) = 66 ms tau_pi
ND3+ND1 line-scan with 66ms probe, Fourier limit 12 Hz
100Hz span with 1 Hz step, 101pts * 11s/100shots = duration 1100 s = 18 minutes
optimistically central feature ca 6-7 Hz AOM wide!? lots of side-bands.. at 50 and 100 Hz?
ID7961 ND3 flop to 150ms, 20pts
ID7959 ND3 flop-on-center, 0.05 to 40ms in 20 steps (maybe too dense sampling, maybe longer pulses needed?)
ID7957 ND3 line-scan, 15 ms probe
300 Hz span, 5 Hz step, 59 pts
now get side-bands, central feature maybe 5pts = 25 Hz AOM wide. Fourier limit for 15ms is 53 Hz
Gaussian fit does not work for this data.
ID7954 ND2 flop on center 75.968055, to 40ms
ND2 predicted tau_pi = 1800*sqrt(10) = 5.7 ms
ID7949 line-scan with ND2 and 5 ms probe (Fourier limit 160 Hz)
500Hz span 16 Hz step, 31 pts.
Fit FWHM 95 Hz AOM, center 75.968055
ID7947 ND1 flop on center 75.968052, 50us to 30 ms
(23:35 local, no trains?) ID7945 +C1 with ND1 1800us probe
1 kHz span, 30Hz step, 33pts
Gaussian fit 222 Hz FWHM AOM (at Fourier limit)
rabi-osc-fit now with fixed A=0.5 (?)
possibly to be continued after 23:30 local...
ID7942 wide scan with 1.8ms probe
duration is ca 8000s = 2+ hours
50 kHz span with 30 Hz step, 1667 points
ID7917 flop-on-center, ND1-filter, to 3 ms probe pulse
predicted rabi-freq is 709Hz/sqrt(10) = 224 Hz, fit gives 258 Hz
ND1 filter, now tau_pi maybe 1800us?
line-scan to check +C1 center frequency:
ID7910 flop to 12ms, 300 shots.
adjust 422-power down to get ca 60counts/10ms
Previous flops with ca 120 photons/10ms. Try a lower cooling-laser power, to get colder ion? (but fit-function now limits lowest Tion to 0.5mK...)
ID7906 probe pulses to 12ms, with 300shots per pulse for better statistics
ID7904 probe pulses to 6 ms
behold, the new rabi-osc-fit-figure:
ID7897 +C1 line-scan
Fit 75.967965, h46, FWHM 719 Hz
ID7896 Try C1/2/4 peak search, with B=1.31uT and linecenter=75.964275
+C1 75.968047, h44, FWHM 710 Hz AOM
-C1 75.960564, h37, FWHM 739 Hz AOM
+C2 75.971737, h41, FWHM 618 Hz
-C2 75.956842, h50, FWHM 611 Hz
+C4 75.982886, h16, FWHM 743 Hz (low peak height)
-C4 75.945704, h16, FWHM 870 Hz (low peak height)
From the 30 kHz scan over lunch:
C1 pair split by 7.35 kHz AOM -> B1 = 1.310 uT
C2 pair split by 14.76 kHz AOM -> B2 = 1.319 uT
C3 pair splot by 29.7 kHz AOM -> B3 = 1.325 uT
C3-pair just about fits into the 30 kHz span
No ND-filter, 600us probe, search for clock-transition
ID7894 predicted linecenter 75.9642, scan 30 kHz span with 150 Hz step
201 pts to scan, ca 4.6 s per point
reinstall the 2nd shield. install the doors.
uMotions seem ok..
re-run B contrast: 1.2% in 72s
B-beam histogram fit fails for some reason? B gives only 20 counts/10ms gate - reposition the holy-blanket
uMotion 300kphot: A 1.8% in 62s, B ?? % in 107s, C 3.7% in 72s
re-lock clocklaser, fbeat now ca 23165894 Hz (stdev 24Hz), 100 MHz checksum stdev 0.0003 Hz
ion is still stored
ASILive camera program had crashed
Comb f0 lock had unlocked during the night. Clock-laser in TEM01 mode
08:03Z stop collision-monitor
14:35Z start collision monitor
stopped collision-monitor at some point.. (sometime before 10:51Z)
13:40Z start collision monitor. A-beam, no clearout, 70/2 bright/dark in 10ms. Note without 2nd B-shield lab lights need to be OFF to get low dark counts.
13:28Z uMotion 300kphot A 0.8% in 46s, B 2.7% in 64s, C 2.5% in 75s
07:20Z stop collision-monitor
07:19Z ion still stored
ca 14:30Z ID7832 2020-06-29-run1 is collision monitor over night
Only A-beam. no clearout. With lab-lights off and holy-blanket over trap: 70 bright, 2 dark counts in 10 ms
Leave collision-monitor over night
investigate 2nd B-shield DC/AC shielding with B-meter. overnight measurement in Z-direction with 2nd-shield and mu-metal-plate bottom.
12:51Z ion signal!? when adjusting 422-power slightly up!? (should use max 422-power allways when trap loading?)
12:31:30Z try dispenser 1.60A
12:46 no signal in 10+ minutes...
check Rb-lock, check 461nm current 197.25mA
12:12Z dispenser 1.55 A, 1.36 W
12:25 no signal in 10+ mins...
Load ion before removing 2nd B-shield for measurements
A-beam power levels
422nm 0.35 uW (208 MHz detuning)
461nm 143 uW
405nm 2.7 mW
11:18Z: Stopped B-field monitor
13:24Z: Starting over-the-weekend magnetic field monitor:
09:00Z: Ion lost. Switching off lasers.
08:00Z: Clock laser was spontaneously unlocked during the night. Relocking.
run27 in progress 2020-06-17 ca 08 local
fancy 3d plot of all the spectra?
15:15Z leave lab.
15:13Z start Bmeter on win-machine
With 500Hz span and 2 Hz step (251pts) each scan takes ca 2100s
run3 is the first one, queue about 30 scans overnight
Leave this as a repeated scan over night to monitor noise?
Rabi-flop maybe shows good Pexc with a long 35ms pulse?
ID7787 scan with 2Hz step and 35ms probe
Try ND3 and 10ms, only 16/100 jumps at most
try to get a narrow linewidth?
ND2 and 4ms gives 111 Hz AOM for +C1
seems hard to find a coil-setting where C1,2,4 are strong..
ID7775 rerun peak-search with B=1.436 uT and wider 5 kHz scan, hopefully to find C4
5 kHz span, 150 Hz step, 33 pts/peak
+C1 h35, 719 Hz
-C1 h35, 753 Hz
+C2 h48, 607 Hz
-C2 h47, 670 Hz
+C4 weak 20/100 central peak with two sidelobes..
-C4 strange, multiple peaks
ID7773 peak-search with (0,0,0) currents
+C1 75.967400, h48, 624Hz (moved closer to linecenter?)
-C1 75.959360, h35, 706 Hz. From C1-pair B=1.436 uT now
+C2 75.971367, h47, 668 Hz (peak at far left of scan-range)
-C2 75.955355, h46, 707 Hz (far right in scan-range), ca 1.43 uT from C2-pair
+C4 no peak found, program aborts at Gaussian fit..
Try zero coil-currents (0,0,0), so far we had (+1,0,0) mA from yesterday
ID7771 re-run peak-search
+C1 75.967936, h48, 627Hz (now at position where ID7765 found it!)
-C1 75.958822, h46, 666Hz
+C2 75.972437, h43, 609Hz (this one falsely identified as +C1 below in ID7768...)
-C2 75.954234, h48, 672Hz
+C4 low peak-height 16/100 jumps, with big sidelobes, no good gaussian fit
-C4 center lower than sidelobes??
B-estimate below is wrong (by 2x), field should be 1.625 uT ?
ID7768 auto-search for peaks. 600us probe (Fourier 667 Hz AOM), 3kHz span, 150 Hz step, 21 pts per peak
+C1 75.972372 (?), h55, 618 Hz (this may search for the peaks in the wrong place now??)
-C1 75.954280, h49, 628 Hz
+C2 75.981584, h40, 703 Hz
-C2 75.945219, h36, 727 Hz
-C4 no peak found
+C4 no peak
ID7765 line-scan for transition, no ND-filter, 600us probe
the mu-plate may have changed the field a lot!?
+C1 at 75.968
-C1 at 75.9589
From these peaks, B-field now 3.25 uT
Improved B-meter data-collection program on win-machine. Probe is now on top of old trap, rouglhy vertically
Note that meter has an arbitrary DC-offset applied now - absolute field not measured
13:29 Try a mu-metal plate on top of the setup!?
15:49Z leave lab.
15:44Z start clock-servo (v3 program), 4 ms probe
Cycle time ca 67s
ID7763 peak search, 4ms probe (Fourier 100 Hz AOM), 400 Hz span, 20 Hz step, 21 pts
+C1 75.966944, h44, 117 Hz (better height now!)
-C1 75.959637, h34, 110 Hz
+C2 75.970567, h27, 137 Hz
-C2 75.956002, h36. 132 Hz
+C4 75.981502, h24, 201 Hz (broad and noisy)
-C4 75.945069, h32, 195 Hz (also broad and noisy)
ID7762 rabi-flop on +C1. visually tau_pi ~4ms
ID7759 peak-search 6 ms probe (Fourier 67 Hz AOM), 300 Hz span, 15 Hz step, 19pts
+C1 75.966938, h17, 302 Hz (low height, check probe-length with rabi flop)
-C1 75.959640, h17, 208 Hz
+C2 75.970573, h22, 124 Hz
abort and do rabi-flops instead
Switch to ND2 filter
ID7758 peak search, 1800 us probe (Fourier limit 222 Hz AOM), 800 Hz span, 40 Hz step, 21 pts
+C1 75.966949, h48, 202 Hz
-C1 75.959645, h36, 198 Hz
+C2 75.970598, h37, 222 Hz
-C2 75.956008, h51, 233 Hz
+C4 75.981529, h30, 308 Hz (looks noisy..)
-C4 75.945104, h28, 279 Hz (better looking peak)
Switch to ND1 filter
ID7757 peak-search with 600us probe (Fourier limit 667 Hz AOM), 3000 Hz span, 150 Hz step, 21 pts
+C1 75.966912, h50, 668 Hz
-C1 75.959570, h45, 579 Hz
+C2 75.970544, h40, 707 Hz
-C2 75.955929, h52, 717 Hz
+C4 75.981499, h27, 802 Hz
-C4 75.945003, h50, 601 Hz (much better looking peak than +C4)
set coil-currents (+1,0,0) mA
ID7755, run7, carrier, same 16 kHz span
ID7754 red secular sideband at 75.481 MHz
ID7753 larger 16 kHz span, around expected blu radial (lower) sideband
ID7752 spectrum around 76.445, predicted upper secular sideband?
span is 8 kHz - so we may only see one radial mode?
scan spectrum at carrier + 0.964 MHz
Tickler with 10dB attenuation shows radial (lower?) mode at 0.964 MHz
(on 2020-05-04 we have secular modes at 0.9644 and 1.0067 MHz)
ID7748 run3, smaller span around carrier. 600us probe, 200 Hz AOM step
Fit 75.963303, h 64, FWHM 1995 Hz
ID7745 'run2' spectrum
abort... correct signs are (-4.6061, +0.7881, +1.3572) mA
ID7743 'run1', spectrum in zero field!?
For fun, try a zero-B-field setting of I = (-4.6061, -0.7881, -1.3572) mA
Notes for possible big box around experiment, 3rd Bshield:
Trap Breadboard: 900mm x 600mm x 55mm
C-beam post is highest component on breadboard, height 370mm
oscilloscope-shelf is ca 660mm from optical table, instrument-shelf is 860mm from table
06:43Z stop Bmeter logging
22:08Z (?) servo seems stalled at iteration 535?? why?
htop shows 12.5/15.5 G RAM use (Monday morning 07 local)
gnuradio-companion uses 25% of RAM, and Firefox ca 13% (WebExtensions?)
diskspace seems ok, 350 G free
09:45Z leave lab.
09:43Z start the Bmeter logging on the win-machine. Same probe position as before, ca. vertical
09:29 run the clock-servo (lock-to-transition v3 program) for C1,C2,C4, with 600us probe
Not sure how dark/bright threshold now set - find out!?
This is ID7740 or clockrun3
Clock-cycle duration is ca 64s
ID7739 rerun peak search
+C1 75.966709, 53, 624
-C1 75.959544, 40, 650
+C2 75.970336, 41, 688
-C2 75.955943, 52, 642
+C4 75.981083, 37, 634 (stronger sidelobes than C1/C2)
-C4 75,945122, 41, 586
recompute predicted peak-positions for all 6 peaks - this should prime the peak-search correctly
FIXME: make this prediction use actual beat-note (cavity-frequency) and AHM3-model
Try to find all thre pairs C1,C2,C4
ID7736 peak search for three pairs, 600 us probe, 3kHz span, 150Hz step
(center / MHz, height / jumps, FWHM / AOM Hz) - from Gaussian fits
+C1 75.966766, 47, 634
-C1 75.959584, 48, 614
+C2 75.970399, 40, 673
-C2 75.955970, 50, 656
+C4 (program aborted... length 4 vs length 6 dataset vector problem?)
ID7733 peak-search, 4 kHz span, 200Hz step
+C1 center 75.966768, height 58, FWHM -622 (negative!? FIXME)
-C1 center 75.959613, height 45, FWHM 599
+C2 center 75.970382, h 42, FWHM 760 Hz
-C2 center 75,955975, h 44, FWHM 779 Hz
set coil currents (+1,0,0) to also see C4 peaks
ID7731 (restarting an artiq-program works) run peak-search with 660us probe (Fourier limit ca 1300Hz optical)
+C1 height 52, FWHM 653 Hz
-C1 height 46, FWHM 635 Hz
08:34Z Chinesium-camera software had stalled (win machine suggests updates need to be applied..). Restarting cam-software worked. Ion seems still stored. Clocklaser seems locked
However cavity-camera on website shows a dark image - why?
ARTIQ seems to have stalled? Last successful run is 7728. Now stalled on 7730 (0-0-5)? why?
13:27Z: Long B-field run for the week-end. No filters, clock probe 600us, 34 measurements, 80 kHz span, 200 Hz step, approx. 17.5 h total
Data saved to DB. Currents, RIDs and DB description fields: 2020-06-12-B-field.txt
06:40Z check uMotions 300kphot: A 1.6% 42s, B 2.3% 51s, C 4.3% 63s
06:37Z turn off B-field logging and the peak servo
servo ran for ca 1207 iterations, ca 47 s per C1+C2 pair
Peak servo run starting ca 15:00Z
+/-C1 and +/-C2 and centers in same figure, peaks offset for visual effect:
Qauntum jumps for the peaks:
ID7603 restart with PMT-counts set to 90 bright, 10 dark
ID7602 servo-lock run to C1, C2 peaks (C4 was weak or weird, not used)
OD2 filter with 4 ms probe
B-field monitoring start: 14:55Z. Bartington-probe about vertically on IR-window-side of trap
Meter voltage output seems to be (uT)/100 V, so 0.29 V corresponds to meter-reading of 29 uT
Sample at 200 S/s and store in text-file on Win machine
ID7599 Automatic peak search OD2, C1+: 119 Hz, C1-: 98 Hz, C2+: 148 Hz (?), C2-: 135 Hz (?)
ID7591 line-scan with OD4 and 40 ms, 2 Hz step and 125 Hz range 63 steps, AC off
ID7590 line-scan with OD4 and 40 ms, 2 Hz step and 125 Hz range 63 steps, AC off
Note: AC-off reverses air-flow, now cold air and dirt comes down from the ceiling...
ID7588 line-scan with OD3 and 12 ms, 7 Hz step, AC off, FWHM = 81 Hz, two side bands, not a good fit.
ID7586 line-scan with OD3 and 12 ms, 7 Hz step, FWHM = 72 Hz
ID7585 flop, out to 30ms - noisy slope upwards..
Switch to OD3 filter
ID7584 flop-on center, OD2, 20 steps to 10 ms. tau_pi 4..5ms visually
ID7582 +C1 line-scan, 5ms probe
FWHM 100 Hz AOM (Fourier limit 160 Hz optical)
ID7581 peak-search with 6ms probe, 20 Hz step - takes too long, abort
Switch to OD2. Try 6 ms probe
ID7580, A-beam, auto-peak-search, OD1, 2ms probe. 1kHz range 50 Hz step, 21pts/peak
Gaussian fit FWHM AOM:
+C1 202 Hz height 48.5, -C1 167 Hz h49.2
+C2 193 Hz h44.2, -C2 175 Hz h49.5
+C4 Gaussian fit failed - program abort.
(+C4 spectrum noisy with no clear peak?)
at new voltages uMotion 300kphot: A 1.9% 42s, B 2.2% 43s, C 4.5% 54s - A and B improved, C same or slightly worse
Set the new optimal voltages (ID7566 result): V1 = -5.7788, V2 = -5.2212, V3 = -16.7212, V4 = 72.8024
uMotion at old voltages, 300kphot: A 6.2% 42s, B 3.4% 48s, C 4.1% 54s
(422-detuning kept at 208 MHz throughout loading and experiments)
11:01Z flashing the 405 on/off a few times restored ion fluorescence!? (one of the first times this has worked!?)
check 422 Rb-lock, seems fine
10:55Z Ion seems to have vanished..
NPL results: Optimal voltages (Common mode, diff): [ 0.2787832 28.04061488 44.76179374]
(old voltages in use so far: [0.2763 27.9914 44.5533] )
09:00Z over lunch: ID7566 NPL uMotion-optimizer with old scan ranges 4x4x4, 200kphot per beam
B-beam was realigned to get some signal again
ID7561 repeat auto-peak-search, now with fiber-noise and breadboard-intensity servos on
Fit AOM FWHMs (2ms Fourier limit 400 Hz optical). 50 Hz AOM-sampling is only 4pts/FWHM
+C1 183 Hz, -C1 189 Hz
+C2 209 Hz, -C2 222 Hz
+C4 720 Hz, -C4 1996 Hz (??)
(-C2 far right in scan-range, because overnight cavity-drift not accounted for in the starting-values?)
(+C4 double-peak or very strong rabi-sidelobe? low max Pexc 22/10 )
(-C4 no clear peak found)
peak-search run duration 690 s
ID7560 try auto-peak-search (old starting values from yesterday). scan range is 1 kHz AOM around old peak value. 50Hz step, 2ms probe
+C1 found to the left of scan.
Gaussian fit FWHMs:
+C1 198 Hz, -C1 (fiber-noise and trap-breadboard-intenstiy servos switched on mid-way...)
abort and re-run
fbeat 23164431 Hz
Clocklaser. Toptica 170mA, piezo 58.638 V, amplifier 350 mA
Fluorescence levels at 90 bright and 10 dark in 10ms seem still valid
08:20Z stop collision monitor
08:19Z ion still stored
15:44Z Turn off clocklaser + comb amp(s)
meh, no clock-run today, just leave collision monitor.
15:36Z start collision monitor. 10ms gate, 90 bright, 10 dark photons. Only A-beam on. No clearout.
ID7556 peak-search again after re-locking clocklaser. Leave intra-cavity servo off, it may cause unlocking?
AOM FWHMs from fits:
+C1 211 Hz, -C1 179 Hz
+C2 168 Hz, -C2 205 Hz - again clock laser in TEM01!!??? why?
+C4 , -C4
ID7555 try the automagic peak-search program with 2ms probe
AOM FWHM outputs:
+C1 192 Hz, -C1 202 Hz
+C2 184 Hz, -C2. Clock-laser jumped to TEM01 mode!!?? argh. abort run
ID7552 run11, line-scan, OD1, 2ms, 40Hz step 1500 Hz range
Fourier limit 400Hz optical, observe FWHM 186 Hz AOM (Gaussian fit)
ID7551 try OD1, flop shows tau_pi ca 2ms
ID7550 flop with comb-amp OFF
ID7546, run10, 4ms probe, OD2 filter. Wider 500Hz scan with 20 Hz sampling
FWHM 187 Hz AOM - but Fourier limit is 0.8/4ms=200Hz??
Try OD2 and 4ms
ID7545 run9, OD3 with 13ms probe, same scan as before
Noisy peak maximum with Pexc=26/100
Expected FWHM is (0.8/13ms)/(2*2Hz) = 15 points wide
61 points in 356 s
try OD3, if 40ms<->OD4, then OD3 should have 40/3=13 ms probe with 61 Hz Fourier limit.
ID7544, run8, repeat with 120 Hz AOM range, to also see +/-100Hz optical sideband? Re-center on previous scan presumed carrier 75.965777
61 points in 521 s
ID7543 line-scan with 80 Hz AOM range, 2 Hz step, run7
Now see 50Hz side-band peaks!? They are spearated by 50Hz/(2*2Hz) = 12.5 steps(datapoints) from the carrier
Try intra-cavity power servo. set-point 1.25V (1348nm Iamp=350mA)
(fiber-stabilization, and probe-power-at-beamcombiner are already ON, from the start of the day)
Turn comb-amp back on
ID7542,run6, same line-scan 120 Hz AOM range, try switching OFF the Comb 1348nm beat-note branch amplifier
61 frequency points, duration 521s
OD4, 40ms probe, 2Hz AOM step, only 80 Hz AOM range (should see +/-50 Hz optical sidebands if they are present)
Fourier limit is 0.8/40ms = 20 Hz optical, 10 Hz AOM, so with 2 Hz AOM sampling we should get 5 points/FWHM.
Scanning 41 frequency points, 100shots/point, takes roughly 350 seconds, so 8.5 s/100shots
ID7536 line-scan with 40ms probe
ID7535 rabi-flop on center, 20 steps to 100ms. oscillating rising curve to 30% @ 60ms, then smoother??
ID7534 repeat line-scan with 2Hz step 55ms probe. Two peaks at Pexc=30/100 spaced only 12 Hz apart?
ion had a dark moment... ID7530 aborted
ID7530 line-scan with 2Hz step, 55ms probe
ID7529 rabi-flop on center - low Pexc and strange shape
ID7528 first OD4 line-scan, +C1
Try OD3+OD1 = OD4 with a long probe
ID7525 repeat peak-search with OD2, 4ms probe, now 20 Hz AOM step. AOM FWHMs:
+C1 127 Hz, -C1 110 Hz
+C2 149 Hz, -C2 129 Hz
+C4 248 Hz, -C4 273 Hz, (noisy looking spectra..)
Note C4 Zeeman shift is 5-fold more than for C1.
ID7524: Automatic peak search. OD2, 4ms probe. FWHMs: +C1: 115 Hz, -C1: 129 Hz, +C2: 154 Hz, -C2: 125 Hz, +C4: 241 Hz, -C4: 325 Hz
ID7520: +C1 line scan, OD2 ND filter, clock pulse 4000 us. AOM FWHM = 102 Hz. Fourier limit = 200 Hz.
ID7517: +C1 Rabi OD2 ND filter. tau_pi ~ 4000 us
ID7516: +C1 line scan, OD1 ND filter, clock pulse 1900 us. AOM FWHM = 228 Hz. Fourier limit = 421 Hz.
ID7513: +C1 Rabi OD1 ND filter. tau_pi ~1900 us
ID7511: +C1 line scan, no ND filter clock pulse 550 us. AOM FWHM = 700 Hz. Fourier limit = 1454 Hz.
ID7510: +C1 Rabi no ND filter. tau_pi ~550 us
Problems with B-beam alignment: almost no signal
uMotion contrasts 300kphot: A 6.5%
461 parameters check: 197.2 mA, 4.645 kohm
11:03Z Ion at 8min!
10:55Z Load with I=1.55A (1.35 W), 422-detuning at 208 MHz (because this worked OK yesterday)
06:01Z collision monitor off.
Ion seems lost, possibly due to Rb-lock problems (lab temperature induced?) with coolinglaser
14:06Z collision monitor on. 10ms gate, 90 bright 3 dark counts.
ID7426 repeat beam shutter on trap bread board again. line-scan repeat with comb-amp on. Both ACs off. now FWHM 109 Hz AOM
ID7424 beam shutter on trap bread board again. line-scan repeat with comb-amp on. Both ACs off. now FWHM 117 Hz AOM
ID7422 repeat line-scan repeat with comb-amp on. Both ACs off. now FWHM 104 Hz AOM
ID7420 line-scan repeat with comb-amp on. Both ACs off. now FWHM 108 Hz AOM
ID7419 rabi-flop on center. Visually tau_pi ~5 ms.
ID7417 line-scan repeat with comb-amp on. now FWHM 89 Hz AOM
ID7415 line-scan repeat with comb-amp off. now FWHM 130 Hz AOM
ID7414 line-scan 5 ms probe. FWHM 96 Hz AOM - Fourier limit is 160 Hz optical.
ID7412 try OD2 filter
ID7411 rabi-flop on center. Visually tau_pi ~2ms.
ID7408 line-scan of +C1, 1900us probe. FWHM 205Hz AOM - Fourier limit is 421 Hz optical.
Try 422-shutter further away from the ion. Spare shutter now mounted on optical table at 422-breadboard fiber-launch
ID7399, ID7400-7401: +C5, clock probe 1900us. FWHM: 258 Hz AOM.
ID7396, ID7397 (7403-7404 high res.): +C4, clock probe 1900us. Two peaks ?.
ID7392: +C3, clock probe 1900us. FWHM: 250 Hz AOM.
ID7390: +C2, clock probe 1900us. FWHM: 220 Hz AOM.
ID7387: +C1, clock probe 1900us. FWHM: 210 Hz AOM.
ID7386 OD1 wide scan 70 kHz, clock probe 1900 us. 2020-06-09-run14
ID7383 OD2 line-scan 4500us, repeat, Gaussian fit 116 Hz FWHM AOM - Fourier limit is 178 Hz optical. Both ACs off. 461 drivers off. 2020-06-09-run10.
ID7382 OD2 line-scan 4500us, 2 peaks ?!, Both ACs off. 461 drivers off. Comb amp off. 2020-06-09-run10.
ID7380 OD2 line-scan 4500us, Gaussian fit 124 Hz FWHM AOM - Fourier limit is 178 Hz optical. Both ACs off. 461 drivers off. 2020-06-09-run10.
ID7378 OD2 line-scan 4500us, Gaussian fit 140 Hz FWHM AOM - Fourier limit is 178 Hz optical. Both ACs off. 2020-06-09-run10.
ID7377 OD2 rabi-flop on line-center - AC off -3k (KSK?), AC off behind the wall -2k (TK203).
ID7376 OD2 rabi-flop on line-center - AC off
ID7375 OD1 rabi-flop on line-center - AC off
ID7373 OD1 line-scan 1900us, Gaussian fit 205 Hz FWHM AOM - Fourier limit is 420 Hz optical. AC off. 2020-06-09-run9.
ID7372 OD1 rabi-flop on line-center -
ID7369 OD1 line-scan 1900us, Gaussian fit 240 Hz FWHM AOM - Fourier limit is 420 Hz optical
OD1 ND-flip-filter gives 30 nW
674nm 270nW -> tau_pi = 600 us. FWHM 661 Hz - agrees with Fourier limit
uMotion contrasts 300kphot: A 5.1%, 0.8% B %, 3.4C %
run1: scan for +C1 peak, 1ms probe
07:12Z stop collision monitor
14:05Z collision monitor on. 10ms gate, 90 bright 3 dark counts.
Dark state monitoring: DB id: 2020-06-08-run1.
ID:7334 Line scan, +C1, clock 10 ms. OD2 ND filter. Comb amplifier off. FWHM: 110 Hz.
ID:7333 Rabi at +C1. OD2 ND filter. Comb amplifiers off.
ID:7332 Rabi at +C1. OD2 ND filter.
ID:7330 Line scan, +C1, clock 10 ms. OD2 ND filter. FWHM: 130 Hz.
ID:7328 Line scan, +C1, clock 2.7 ms. Center: 75.650 MHz, FWHM: 160 Hz (AOM). OD1 ND filter.
ID:7327 Rabi at +C1. OD1 ND filter.
ID:7325 Line scan, +C1, clock 3 ms. Center: 75.650 MHz, FWHM: 154 Hz (AOM). OD1 ND filter.
ID:7322 Rabi at +C1. No ND filter.
ID:7320 Line scan, +C1, clock 1 ms. Center: 75.650 MHz, FWHM: 373 Hz (AOM). No ND filter.
Line scan, clock pulse 1 ms, 2020-06-08-run2.
uMotion contrasts 300kphot: A 6.5%, B 3.9%, C 4.2%
12:05:15Z try to load again. 1.55A
12:08:20Z signal! single ion
11:51Z Idisp=1.55A (1.3W) - try to load again
12:01Z signal! two ions.
Glitch from signal-gen at 36mV - ions lost
RF-monitor 1.69 Vpp - 70mVpp from signal-gen
11:36Z Idisp=1.55A (1.3W)
No signal in 10 mins
461nm check dispenser - 198.07mA - a bit lower than before (199mA)?
Check pulse-stretcher. Output is 25ns long
-2 floor air-compressor seems to give strong 85 Hz noise component
07:19Z reduce KSK105 VFD setting to 20 Hz
Cavity heating power slightly higher than before 20.5 W + 3.2 W (heat-loss thru steel I-bar to the now colder A/C channel?)
A/C VFD was on 25 Hz setting overnight. lab-temp shows slow oscillations around +23C
07:09Z ISS to further reduce input-air temp. Biltema sub-floor shows 18.1C
11:17Z Biltema temp sensor: 18.2C
08:25Z lower A/C airflow to 25 Hz setting
08:20Z lab-temp falls below +23C, biltema-meter sub-floor +18.5C.
Biltema-meter at back of lab shows 19.2C under the floor
06:50Z ISS to lower temperature of incoming air, setting at +17.5C
A/C 30Hz and original temp gave ca 23.5C lab-temperature (Vaisala DMP248 sensor)
Now 5pcs D160mm holes in the floor completely open
10:35Z Try a setting of 30 Hz
Various tests with different A/C-VFD frequencies..
Opened 4pcs D160mm holes in the floor. This lowers lab-temp by ca -0.5C with A/C-VFD at original RPM of 48 Hz
With EOM PA outside the box the top coarse heater servo now works
Ca 19W bottom heater and 2W top heater now (rightmost data in figure)
17:19Z Turning off the experiment... ion storage time was 7+ days
17:12Z A/C VFD back to 48 Hz
ID7292 flop to 20ms, 100shots - compare with previous 7291, but this one is longer. seems noisy after 8ms or so
ID7291 flop on center. 100shots, good oscillations out to 10ms
ID7289 line scan 1.4ms probe
fit FWHM 310 Hz AOM
ID7288 flop with high probe power, first max ca 1400us
for fun try a high probe power, ca 180 nW now
not so great flop
Adjust probe power with waveplate.
previously 18dB att was ca 25 nW.
now go to lower value, with 18dB att ca 5 nW
ID7284 line scan 20dB att, 4ms probe
FWHM 106 Hz AOM
ID7283 flop 20 dB att, 100shots - max around 4ms
ID7282 flop 24 dB att, 100shots - rises to 6ms but then incoherent/noisy?
ID7281 flop-on-center 18dB att, 100shots - seems ok - tau_pi now maybe shorter than 4ms? compare to ID7276 with high 422-power
ID7279 scan for line, 4ms probe, 18dB attenuation 20Hz step
mis-align 422 input-fiber so we have 0.4 uW and get 100 counts in 10ms
422-amp to 1.8A, ca 0.8 uW in A-beam, 160 counts in 10ms (previously 'low' cooling was 80 counts?)
Clock-laser jumped to TEM01 mode? but maybe flop-run was done?
We now have a 'high' cooling power 1.24uW, so next ty a lower value?
ID7276 flop, 18dB att, 500 shots
ID7274 flop, 26 dB AOM att - slow incoherent rise
Note: comb beat-note stdev now at 23 Hz (lower than before?)
ID7272 line-scan 8ms probe, 23dB AOM att, 10 Hz step.
gaussian not a good fit
ID7270 flop, 23 dB AOM att. tau_pi maybe 8ms
ID7268 line-scan, 20dB AOM att, 6ms probe, intra-cavity servo on
fit FWHM 87 Hz AOM
ID7267 flop, 20dB AOM att, intra-cavity servo on (1.4V setpoint), tau_pi ca 6ms
ID7266 flop, 20 dB AOM att
ID7264 line-scan 6ms probe, 10 Hz AOM step
Fit FWHM 96 Hz AOM
ID7262 line-scan, 6ms probe, 20 Hz AOM step
FWHM 103 Hz AOM
ID7261 flop-on-center, tau_pi maybe 6ms now(!)
ID7260 line-scan 4ms probe
15:30Z turn down aircon VFD to 10 Hz (it was at 48..49 Hz)
ID7252 line-scan with 2ms probe
seems noisy, Pexc(max)=16/100. Fit FWHM 245 Hz AOM
ID7250 repeat line-scan
Fit FWHM 167 Hz AOM
ID7248 line-scan, 18dB AOM att, 4ms probe. 180 counts bright-state (we have more 422-power than before)
better blanket, but still a noisy spectrum!?
Fit FWHM 186 Hz AOM
ID7246 flop on center, 18dB AOM att
ID7244 flop on center. 20dB AOM att
ID7242 try positioning a blanket over the trap breadboard
4ms probe, 18 dB att, 20 Hz step
fit FWHM 159 Hz AOM (but Gaussian not a good fit)
ID7240 line-scan without blanket. 18dB probe att, 4ms probe. 20 Hz AOM step. (this is not necessarily a pi-pulse - test rabi-osc also later)
Gaussian fit gives FWHM 178 Hz AOM (Fourier limit 200 Hz optical) - so we have 1.78-fold FWHM
Try a bigger holy blanket that minimizes air-currents around the trap breadboard?
NPL-optimization 200kp, ID 7231.
Optimal voltages (Common mode, diff): [ 0.36018227 28.15357435 44.74236194]
Optimum outside scanning range!
uMotion contrasts at the old bias voltages. 300 kphot
Old: 0.2763, 27.9914, 44.5533
A 7.2% (26s), B 7.6% (40.4s), C 6.5% (39s)
11:10Z doors closed
09:45Z doors open. Move the EOM PA outside the box.
09:32Z collision monitor off.
14:15Z collision monitor on. 10ms gate, 80 bright 2 dark counts.
ID7208 line-scan, 20 dB AOM att, 4 ms probe
FWHM 171 Hz AOM. noisy?
ID7207 flop, 20 dB AOM att, visually tau_pi 4ms
build a finnfoam box for the ECDL
ID7197-98 repeat, clock laser now repeatedly unlocks??
Gaussian fit 140 Hz FWHM AOM (This was 79 Hz before lunch)
ID7196 repeat 6ms scan, with 10 Hz step - clock laser unlocked
ID7193 repeat 6ms line-scan, to compare aircon off vs on
6ms probe, 23dB AOM att. 20 Hz AOM step
Fit FWHM 56 Hz AOM
Go back to the old voltages
New uMotions 300kphot: A 14.6%, B 12%, C 8.4%. These are worse than before!?
set the new voltages
uMotion contrasts at the old bias voltages. 300 kphot
Old: 0.2763, 27.9914, 44.5533
A 10% (35s), B 9.6% (39s), C 7.4% (65s)
Vopt found: (Common mode, diff): [ 0.31609226 28.105798 44.69042412]
ID7174 NPL uMotion over lunch. 4x4x4 grid. C-beam gives quite weak fluorescence compared to A and B
ca 12 local - aircon back ON
ID7172 line-scan 12ms probe, 24 dB attenuation
Gaussian fit FWHM 84 Hz AOM (not a good fit)
ID7171 flop on center, 24 dB AOM att, tau_pi 12ms visually
ID7170 line-scan 14ms probe 25 dB AOM att.
ID7169 line-scan, 14ms probe, 27 dB AOM att.
ID7168 try 27 dB AOM att.
ID7167 flop on center, 29 dB AOM attenuation. taus out to 40 ms - no clear osc.
ID7165 relock clocklaser, try again. same as ID7164, 14 ms probe 26 dB AOM att.
FWHM 60 Hz AOM, but Gaussian is not a good fit
ID7164 repeat same scan - clock laser unlock!
ID7161 same scan, but smaller span
Fit gives FWHM 59 Hz AOM
ID7160 line scan, 14 ms probe, 26 dB AOM attenuation. 6 Hz AOM step
Fourier limit now 57 Hz optical
meh.. abort the long run
ID7159 flop-on-center, now 26 dB AOM attenuation.
ID7157 line-scan 6 ms probe, 23 dB AOM att, 20 Hz AOM step
Fourier limit now 133 Hz optical
Fit gives FWHM 79 Hz AOM
ID7156 flop. increase to 23 dB AOM att. taus out to 20ms now. now tau_pi maybe 6ms
ID7155 flop-on-center, seems ok tau_pi around 4ms
ID7153 line-scan again, 4000us probe, 20 dB AOM att.
FWHM 102 Hz AOM, close to the Fourier limit!
08:12Z aircon OFF
ID7151 line-scan 4000us probe, 20 dB AOM att.
FWHM 178 Hz AOM
ID7150 flop on center, try 20 dB AOM att. visually tau_pi 4ms
ID7148 line-scan again. 3200us probe
FWHM 175 Hz AOM
ID7144: Line scan with 461 current, temp and dispenser powers off (acoustic noise/vibration?).
FWHM AOM 154 Hz
ID 7146: Rabi
ID7143 Rabi at line center.
ID7140 Line scan with doors closed, same settings as the day before.
FWHM 178 Hz AOM (similar to last line-scan y-day!?)
With the AA AOM PAs (ca 9W each) moved off the cavity-system and attached to the Thorlabs rails we still overshoot the cavity-top temperature
Previously bottom-heater was at only 5-6 W, now at 14 W.
Probably need to move the AOM PAs and the EOM PA outside the box?
Top temperature overshoot is now +0.150 K. Previously it was around +0.8 K
Positive Kikusui supply to the box is +24V 2.7A = 65 W
07:23Z: Collision monitor stopped
15:36Z restart monitor with 10ms gate time, 55/2 bright/dark counts
14:51Z start collision monitor
Previous runs had all beams ABC. Now we have only A beam. Bright 18, Dark 1 count in 2ms
line-scan with doors closed, 3200us probe
FWHM 178 Hz AOM
14:34Z close cavity doors
Move AOM PAs from top-breadboard to the inside of the Thorlabs-profile. Still inside the box, but not in contact with the cavity-system
line-scan with 3200us probe, 25 Hz AOM step
FWHM 220Hz AOM
intra-cavity servo on seems to unlock the PDH-lock to the cavity (jumps to a TEM01 mode) - maybe lower gain, BW, or output-range on intra-cavity servo would help?
touching the cavity probably shifted frequency
Try intra-cavity servo on. set-point 1.7V (low?)
ID7122 flop on center, AOM att 20 dB (was 18dB previously)
ID7120: line scan
Fit 75.970640, FWHM 240 Hz AOM
turn on intensity servo
line scan with 40nW, 2600 us (fourier limit 308 Hz optical). 30 Hz AOM sampling
Fit: 75.970705, FWHM 221 Hz AOM
ID7116 flop, 500 shots, 20pts out to 10 ms. visually tau_pi 2600 us
ID7115 flop-on-center, 40nW. taus 0.05 to 5 ms
lower probe to 40 nW, with waveplate
ID7113 rabi-flop on center. ca 300 nW, visually tau_pi= 1300 us
+C1 at ca 75.9708 MHz
Fit 75.970787, FWHM 437 Hz AOM
Fiber-noise servo on, but intensity-servo and intra-cavity servos off for now
Beam powers: 674nm: 320 nW. 1ms probe, scan for peaks
12:53Z: Dark state data collection stopped
replaced detector with ET-2030A
(from length lab. Now with two mini-circ amplifiers
Photodiode-signal after amplification ca -15 dBm
Try to optimize 357 MHz fiber-noise detection chain.
Currently: unknown photodiode -> Filter SHP-300 HP 290MHz -> ZHL500LN +28dB -> SLP-450 -> ZHL500LN +28dB -> 6 dB Att -> ZHL550LN +28 dB
+84dB - 6dB = 78 dB Gain!?
07:08Z open cavity doors, to decrease heat-load (AOM PAs) inside box
07:04Z ion still stored. collision-monitor running. Now back-of-lab lights on, hopefully doesn't affect PMT counts
14:20Z no dark states seen so far (12 minutes). Leave this running.
14:08Z start new collision-monitor run, now with also Clearout ON. We expect to only see short events now!?
This is 2020-05-17-run3
14:07Z end the random telegraph, ca 2k events collected. Turn OFF 405-laser.
Try driving D5/2 state with 405nm laser. Visually turn-down 405-power to only 30uW to get nice quantum-jumps. A high 405-power results in completely dark ion (at 2ms resolution)
'random telegraph' method
2ms PMT gate. 28 bright, 1 dark counts. Threshold is geometric mean, rounded to an integer, so 5.
Letchumanan thesis FIG5.6 has 97k events and lifetime of 389.3(2.1) ms
We collect ca 60 events per minute, so 97k events in 27 hours
200ms get 2800/20, or SNR 140 (21.5dB). Scaling this to 2 ms gives 28/0.2 !
20ms get 280/2
2ms gate, 28/1 bright/dark counts still roughly OK.
13:33Z stop collision-monitor run. It ran from around 15:23Z on Saturday until 13:33Z on Sunday, and recorded ca 110 events. Ca 70 of these (64%) are short-lived around 2-6 ms maybe.
13:31Z ion still stored and collision-monitor running
Try to plot Likforman2016a FIG4 figure:
15:23Z vacuum collision-monitoring restarted with 2ms gate (bright 28 counts, dark 1 count), tag 2020-05-16-run4
NOTE: influxdb needed re-starting (systemctl restart influxdb) after disk-space was OK
Dispace was running low, about 360 GB SDR IQ data... deleted.
ion still stored
C1 not resolved, problem with fiber noise cancellation. Rerun:
wide scan. center 75.963 MHz, span 2.2 MHz, step 200Hz. 11 kpoints, duration 58 ks = 16 hours
Overnight run failed due to memory leak. Problem solved, rerun:
wide scan. center 75.963 MHz, span 2.2 MHz, step 200Hz. 11 kpoints, duration 58 ks = 16 hours
wide scan overnight. center 75.963 MHz, span 2.2 MHz, step 200Hz. 11 kpoints, duration 58 ks = 16 hours
fbeat 23164789 Hz. predicted line-center 75,963 MHz
Back to high probe-power of 370 nW. tau_pi=800us. probe-intensity servo now off
FWHM 668 Hz AOM, Fourier limit 1kHz optical
and now a 60s image-stack:
Determining the Zernike polynomials that lead to these aberrations is left as an exercise for the reader
12:55 Center 75.966346 MHz, FWHM 361 Hz AOM (if line-center now has moved a lot, then we cannot compare ID6998 and ID6995 fairly!)
ID6998 and ID6995 were run at 75.966328 MHz (line-center measured ca 12:10)
12:39 ID6998 is comparison against ID6995, now with aircon recently started
12:35Z aircon back ON
12:32 ID6997 AOM-att 18dB, out to 20ms
12:30 ID6996 AOM-att 24 dB (probe W?), 100shots, incoherent..
12:20 ID6995 repeat with 500 shots. Pexc(tau_pi=2.5ms)=163/500 = 33%
12:17 ID6994 flop on center - clear oscillation, tau_pi 2..3 ms (similar to 2020-05-08 result with 63 nW?)
12:10Z ID6992 line-scan 3ms 30nW, 25Hz AOM step
clock-laser unlocked (door-slamming...)
FIT 75.966328, FWHM 249 Hz AOM
ID6991 aircon OFF flop!?!
12:06Z aircon OFF
ID6990 flop on center. incoherent rise to 25%
check probe power, still ca 30 nW
try with airconditioning off!?
ID6989 flop on center, reduced 422-power
get ca 80 counts/10ms with PMT
Reduce 422-power to ca 0.3 uW
ID6989 line-scan again, 3ms 30nW probe
low incoherent Pexc level?
fit 75,966275, FWHM 379 Hz AOM
ID6985 flop on center, 30pts to 10ms, 200shots - no clear oscillations
ID6983 line-scan, 30nW 3ms probe
FIT 75.966267 MHz, FWHM 271 Hz AOM
ID6982 flop, to 10ms, 100 shots
reduce to 30 nW, with the waveplate
probe was ca 300 nW
ID6980 rabi-flop on center, to 5ms, 100shots. tau_pi ca 1 ms visually
scan for +C1 around 75.966
Fit: 75,966291 FWHM 493Hz AOM
scan for clock transition
1ms probe. fiber-noise servo ON, probe-intensity servo ON, intra-cavity servo OFF
with beat-note of 23.1647 MHz line-center should be around 75.963 MHz
at Vopt uMotion 100kphot: A 7.4%, B 6.9%, C 6.0% - not much better than before?
Optimization result is: Optimal voltages (Common mode, diff): [ 0.27629906 27.9913947 44.553304 ]
09:05Z 39/64 pts done. ion still alive and stable..
ID6965: Try the NPL uMotion optimization. 4x4x4 grid.
uMotion 100kphot: A 6.9%, B 7.3%, C 6.5% (weakest)
detuning to 208 MHz
08:20Z now suddenly a stable ion image!!?
stacked 10s image:
trapping too unstable to optimize anything?? image comes and goes randomly, just few seconds bright, then dark for 10s or 1 min or whatever..
strange blinking. try relocking to rb-line
07:30Z signal! diffuse,single ion?
07:24Z dispenser 1.55A (1.3W)
try to reload again
detuning to 185 MHz
Try aligning beams onto two ions, with RF 100mVpp
turning off 422 for 30s or 60s does not remove second ion
changing V34cm around moves ion sideways in image, but doesn't get rid of the second ion. ca 10V steps tried. But we have slow RC-filters on the bias voltages..
turning off cooling for 10s or 30s does not remove the second ion
RF-drive can be adjusted 40mVpp to 100mVpp without glitches. At 40mVpp ions move clearly further apart.
stacked image, 40x 1s exposure:
07:00Z signal! again two ions? now stable side-by-side for longer
06:54Z dispenser 1.55A (1.3W) (lower is better?)
cycle RF OFF/ON, and try to reload
diffuse blinking image? briefly saw two sharp images side-by-side
06:44:20 ion! diffuse image? went dark, then reappears
06:37:30Z dispenser 1.6A (1.40W)
422-detuning 185 MHz
all beams on
camera: 1s exposure, high gain, bin2
405-power to 2.5mW
tickler-cable terminated 50R
Trap RF 16.77 MHz, 70 mVpp, monitor 1.69 Vpp
beat-note around 23.1644 MHz now, maybe 1 kHz change from before closing the doors
Bottom heater power dropped from 40-45 W to ca 6 W. Time-constant for coarse heaters around 4-6 hours maybe
However cavity-box top is still too hot by +0.8 C.
The AOM-amplifiers, 2pcs bolted to side-breadboards, consume about 9W each. Could be moved outside cavity-box?
EOM-amplifier is on stone, so heats air inside box.
Photodetectors, camera, ion-pump should dissipate only small amounts of heat?
14:47Z shut down coolinglaser. Leave clocklaser, tracker, and comb running
Overnight might see large cavity-frequency excursions as temperature settles?
go back to the old Vbias values (-5.5, 0.2772, 27.9984, 44.4698)
Try 405+461, but doesn't seem to help - ion gone/dark?
Try to re-lock Rb-line. All beams on.
Program only got to point 8/64, then ion went dark.
Try to run uMotion-optimization in a 4x4x4 grid around previous optimum. 100 kphot.
ID6876. 64-point scan. (A 11s + B 17s + C 19s) * 64 pts= 3000s duration
run3: try again scanning +C1, 2 kHz region around peak in run1
get Pexc of 5/100 as max. Maybe all Pexc is now in uMotion sidebands?
uMotion 100kphot: A 66% !?, B 69%, C 68%
These have changed a lot!
do we have multiple ions? blink-test looks ok?
went dark when starting clock-spectrum scan!? why?
detuning to 208 MHz
13:34:40Z ion! looks good!?
13:29Z dispenser 1.6A (1.4W...1.418W)
405-power to 3 mW...drifts to 2.5mW, try 1.6A. camera 2s exposure
13:28Z no clear signal..
13:16Z dispenser 1.55A. all beams on. 1.315W
Try to reload again, with 2.6 mW 405-power, which worked previously
12:56Z dispenser 1.55A, 1.306 W
13:09Z signal, but flashing? (possibly multiple ions? long loading time and high 405-power?)
reload ion. detuning back to 185 MHz
scan for clock transition around 75.961, 1ms probe
set probe to max 470 nW
437 frequency points at 5.5s per point
ion seems lost sometime after half of this scan!? (12:45Z maybe)
+C1 at 75,96516 MHz
detuning to 208 MHz for operation
11:42:30Z dispenser 1.55A. (1.32W)
ion! at 11:49Z
detuning to 185 MHz for loading. Try max 405nm power of ~6 mW.
RF siggen 70mVpp, monitor 1.7Vpp
Try to load, with new camera. 1s exposure, high-gain, 2x2-binning, auto-histogram
11:18Z close doors to cavity. Expect heater power to drop from 71-72W down. Previously drop was from 90W to maybe 30W.
New wood-fiber floor-tile with hole for input fiber. Air-gaps to cavity-space sealed.
shut down lasers, leave trap RF on, and comb oscillator on.
ca 11:03Z airconditioning turned back on
Clock-laser unlocked! beep from IR amplifier (?) won't turn on again (cable problem?)...
ID6831 scan for peak again, 5ms 10nW, 'run1', get low Pexc, fit (Fourier limit 160 Hz)
ID6830 scan for peak, 5ms 10nW,
ID6829 flop on center P674=10nW, reduce sampling to 20pts, expect sqrt(60/10)=2.4-fold tau_pi at 5.9ms(?). Pexc very low - probably not at line-center?
ion went dark for a while, after adjusting probe power - why?? re-lock 422 to Rb-line, just in case...
reduce probe to 10 nW
ID6823 flop on center. visually tau_pi ~2.4ms. P674 ca 63 nW
This is possibly the slowest rabi-osc with good cotrast we have seen.
ID6821 scan to find the line again. 3ms probe. Gaussian fit center 75.964460, FWHM 142 Hz AOM
Rabi model FWHM 305 Hz
10:16Z airconditioning turned off
With airconditioning on, no contrast at 50nW probe visible
ID6820 P422=200nW, P674=50nW. Rises to Pexc 25% at ca 2.5 ms
Reduce P422 further? to 200 nW. PMT counts now 50/10ms
ID6818 500shots, P674 50nW, P422 300nW. Only see 25% Pexc at expected tau_pi=3ms. Flat 25% after 3 ms
Reduce probe further, to 50 nW, expect sqrt(3)~1.7-fold longer tau_pi (maybe 1.7*1.8 ms = 3 ms)
ID6817 500shots, probe 150nW, P422 0.3uW. Probe reduced 2.4-fold, expect tau_pi increase 1.55-fold. visually tau_pi ~1.8 ms now
reduce probe to 150 nW
ID6816 repeat with 500 shots, P422 0.3 uW, probe 360 nW, visually tau_pi ~1 ms
ID6815 same as ID6812 but with lower 422-intensity
Try lowering 422-intensity to ca 0.3uW, PMT counts down to 80/10ms
Currently 422nm ca 0.6 uW, PMT counts 150/10ms
ID6812 rabi-flop on center freq. tau_pi=1ms, Pmax 50%, oscillation to about 5ms, then 25%
flop with probe 0.05 - 10 ms in 40 steps
ID6810 360nW, 1ms probe scan -> center 75.964383, FWHM 424 Hz AOM
adjust probe to 360 nW - using waveplate for adjustments (Zeeman AOM DDS attenuation constant at 18 dB)
ID6809 scan for peak, 1ms probe 3kHz range (this was at 40nW? probe, very low Pexc)
422-detuning is 208 MHz ( 2*(208-218) = -20 MHz optical)
ID6807 should be repeat of ID6802, slow rise to incoherent 25% Pexc
run9 +C1 peak with 3ms probe 45nW
FWHM 238 Hz AOM (Fourier limit 267 Hz optical)
ID6802 rabi-flop, 45nw, to 10ms probe length, 100shots. slow rise to 25% Pexc..
reduce power to 45 nW
high power was ca 450 nW
Check +C1 peak at 75,963736, 490 Hz AOM FWHM.
ID6799 rabi-flop high power, 100shots
increase intensity-servo setpoint to 0.275 V
turn on intra-cavity servo. signal 2.6 V
peak now at 75.9637
connect PDH-detector + following mini-circ amplifier to batteries. touching the cavity shifts frequency...
ID6792 / run5, +C1 peak, 1ms probe 1 kHz scan 20 Hz step, 52pts, ca 300s
Fit 445 Hz FWHM AOM (Fourier limit 800 Hz optical)
ID6791 0.05 - 10 ms, 40steps, 200shots. tau_pi 1.1ms. damped oscillations.
ID6790 flop, 0.05-2 ms, 25 steps, 100shots. extra faster modulation on the rabi-osc? tau_pi maybe 1ms
go back to highest probe power possible, ca 390 nW
ID6787 +C1 peak, 10ms probe 1kHz span, 10Hz step. ca 600s = 10min scan. run4
Center 75.964994 MHz, FWHM 216 Hz AOM.
Not a thing of beauty...
ID6786 1-20ms, 100shots
ID6785 rabi-flop 1ms-100ms, 100shots. rises to 25% at 10ms, then flat
ID6784 rabi-flop 0.05-5ms, 100shots. slow rise, but low Pexc
Reduce power with waveplate, to 12 nW
ID6781 +C1 peak, 2ms probe 1kHz span 40Hz step, run2. Fit center 75.965011, FWHM 265 Hz AOM. Fourier limit 400 Hz optical
ID6780 turn intensity-servo gain up.
ID6779 flop 0.05 to 5 ms 100 shots with 100nW, visually slow rise to 25% Pexc, then flat.
Reduce power with waveplate, to 100 nW
Rabi-osc on +C1 at 75.96492
ID6776 +C1 peak, 1.2ms probe 1 kHz span, 60 Hz step. Fit center 75.964989, FWHM 368 Hz AOM
ID6774 18dB att, 0.05 - 10 ms, 25 steps, 500 shots. tau_pi=1.2 ms (visually), P 330nW
ID6774 18dB attenuation, 0.05 - 5 ms probe, 25 steps, 100 shots. P674 ca 330 nW
run1: wider scan 40 kHz AOM, 1ms probe, 18 dB probe att.
ID6771 peak-scan, 2 ms, 20 Hz AOM step. Fourier limit ca 400 Hz FWHM optical
fit gives 296 Hz FWHM AOM
ID6770 AOM att 24 dB, rabi-osc.
tau_pi 2 ms maybe
ID6768 same rabi-osc, but now intensity-servo ON, at DC-setpoint 200 mV
ID6767 rabi-osc. intensity servo off, detector DC-level 260 mV
tau_pi ca 1.1 ms
peak at 75.9582 MHz, 1 ms probe, fit gives center 75,958106, 451 Hz FWHM AOM
measured 674nm thru chamber ~450 nW
18 dB AOM-attenuation as usual
Search for clock-transition. Double-pass zeeman AOM config changed, make sure it shifts 2*76 MHz
(+1, +1) is the right diffraction from the AOM. Away from the SMA-connector, after the 45-deg mirror this means the +1 diffraction is towards the trap!
uMotion contrasts A: 1.1% (24s), B: 1.4% (21s), C: 1.4% (22s)
11:50:30Z try reload again. 1.6A (1.38W). all beams on. 422 Rb-lock checked
11:55Z ion! sometimes looks cold, sometimes diffuse/blinking?
strange blinking at first. then stable single ion!?
11:31Z reload. 1.55A. All beams on.
11:43Z no clear signal
11:21Z load with 1.55A (1.27W)
11:27Z see diffuse and weak cloud? Multiple ions!? (camera exposure 1.9s). Signal goes away at 208 MHz detunign - reappears at 185 MHz detuning
only A-beam was on initially (most of loading time), all beams might be helpful?
Built simple servo that measures 674-intensity at trap breadboard. Servo controls 178.5 MHz AOM drive level using variable attenuator.
DC-signal ca 0.25 V (could be higher?), using uncoated UVFS plate as splitter.
Detector is 'one-inch-detector' with S5791 photodiode, OPA657 opamp, and 1 MOhm transimpedance, bandwidth 4 MHz, using 1.9 MHz LPF on output.
Trap RF 16.770 MHz, 70mVpp siggen, monitor 1.70 Vpp.
Power in A-beam (through chamber):
674: 260 nW (intensity servo off)
422: 0.85 uW (185 MHz detuning)
461: 200 uW, I=199.3 mA checked against helper-dispenser
405: adjusted to 2.5 mW
ID6635 SLI position 1+2 (SLI position 1 looks unstable)
ID6634 (servo bumps at -30dB from carrier)
change to 20dB + 10dB attenuation on toptica servo output (it was 20dB + 6dB)
ID6632 rabi-flop (SLI position 2, it was at position 3) servo-bumps 25 dB down from carrier
ID6631 rabi-flop (further increase gain, now see 2nd servo-bump peak also)
ID6630 rabi-flop (slightly again increase PDH, gain)
ID6629 rabi-flop (slightly increased PDH gain, gives clearer servo-bump peaks) - maybe rabi-osc now smoother?
ID6626 rabi-flop, lower PDH gain, no clear servo-bump peaks
flop ID6622m 27dB att - rise to 25% then noise
Try terminating (50R) the tickler-cable
flop ID6621 (27 dB att)- nice rise to 2.5ms, then flat (incoherent?) noise at 25% Pexc
+C2 (?) is strongest peak:
18dB attenuation, tau_pi = 1.3 ms probe gave FWHM 333 Hz AOM (Fourier limit ~600 Hz optical)
21dB attenuation, (ID6615 flop) tau_pi ~2 ms probe, gave 321 Hz FWHM AOM (Fourier 400 Hz optical)
27dB attenuation, (ID6618 flop) not-so-great rise to max at maybe 5 ms, no clear oscillations. (Fourier limit 160 Hz optical).
Noisy spectrum (like the top of the rabi-osc), FWHM 261 Hz AOM
2020-05-04 May the Fourth be with you
run9: leave long and wide run overnight
Start below carrier C5, scan linearly up to and beyond upper axial secular sideband
scan range [75.9379, 77.0629]. 5600 points (ca 200Hz AOM step, 400 Hz optical)
with 600us probe Fourier limit is 1300 Hz so we get 5 points per FWHM
16:41Z expected remaining duration is 25 ks = 7 hours
ARTIQ NOTE: requesting clock-cycle scan of ca 2 MHz width and 200 Hz resolution gave error...
Scan spectrum around expected secular side-bands
ARTIQ NOTE: restarting the 'v2' clock-cycle program cancels the currently running program! ARGH.
ARTIQ NOTE: running the tickler program messed up settings for the fiber-noise-cancelling DDS! ARGH.
Go back to maximum probe power, ca 600us tau_pi now
run6 carrier peaks around 75,9629 MHz, 45 kHz AOM range should capture C5 also?
run7 upper r_lo sideband should be centered around 76,4451
upper r_hi sideband should be centered around 76,4663
upper axial sideband should be centered around 76,9314
Look for secular frequencies with tickler. Helres monitor voltage is 1.71 Vpp
Scans with 4 dB attenuation around 1 MHz
one mode at 1.0067 MHz, lower at 0.9644
Axial, 20dB att, many dips around 1.937 MHz
reduce to 22dB att, still many peaks
try 24 dB att, now maybe a single dip at 1.93692 MHz
Try to reduce probe-power 10x, with the waveplate. Previous reading was 600 nW, now maybe <60 nW
ID6485 100shot rabi-osc, to 5 ms. Now tau_pi = 2 ms maybe (not a good clear maximum)
ID6486 Scan the line with 2ms probe, 40Hz AOM step (3 kHz span is too wide...)
gaussian fit 75,965782 MHz (FWHM 305 Hz AOM), Fourier limit ca 400 Hz optical
Try Rabi-osc on +C1
Locate C1 at 75,965775 (FWHM 1131 Hz AOM)
ID6480 rabi-osc to 1ms probe, 100shots - see only a rise to max Pexc?
ID6481 to 10ms - too few points?
ID6482 to 10ms, 50pts, 500 shots. Tau_pi ca 660 us. see attenuated oscillations out to 6-7 ms or so - good!?
Probe length 400 us - power 'max'. Fourier limit ca 2 kHz optical
scan for clock transition, 70kHz optical span, 400 Hz step
line-center ca 75.9629 MHz (?). The 70 kHz (optical) scan is not wide enough to see C5
NOTE1: run1 and run2 were unsuccessful because the fiber-stabilization servo was unlocked - need a monitor/alert for this!
NOTE2: the (magnetic?) chair is now close to the door
Detung-scan, 1000ms gate
ID6474, bright/dark/bright counts in 1000ms gate
Move closer to target (97mm)
A-beam counts in 10ms bright/dark 170/20 - with some darkening-improvements dark-counts go to 1-5
Move camera closer, obj-to-cube now 102mm. re-focus imaging-optics (small z-movement)
now PMT counts 180-190/10 in 10ms
align ASE: A-beam counts in 10ms 80-90/4-5
set detuning to 209 MHz
A-beam counts in 10 ms ca 70/4 bright/dark
uMotion 100 kphot: A 9.8% (18s), B 6.7% (47s), C 8% (28s)
13:13:30Z Load ion, Dispenser 1.55A (1.26W)
13:20:20Z ion! loading time 6min50s
Image seems weak-ish on camera(?)
Lock toptica to cavity. SDR tracker
fbeat 23164630 Hz
Goal: reposition camera focus-plane and PMT pinhole plane for better agreement, and thus better SNR
Camera objective (front edge), to cube first edge 115mm, should be about 97 mm to have image-plane and PMT-pinhole agree
A-beam power levels after chamber:
405: 2.5 mW after adjustment
422: 0.4 uW, 185 MHz detuning (power is decreased from max possible)
461: 200 uW (4.645 kOhm, 199.32 mA)
674: 660 nW
Bias voltages back to previous values of 2020-04-28
Helical resonator DC-modulator-box removed, N-Short grounding
This raises the resonance frequency and Q. Resonance at 16.770 MHz. 70 mVpp drive gives 1.73 Vpp monitor
Morning: RF investigations with Spectrum Analyzer..
76 MHz drive to the Zeeman AOM
357 MHz drive and 178.5 MHz VCO of the fiber-stabilizer
16.7 MHz trap drive
2020-04-28 Search for EMI
Measured from tickler cable:
Stable 'comb': 64 kHz, 124 kHz, 248 kHz, 372 kHz, 495 kHz, 620 kHz, 743 kHz, 867 kHz
Intermittent noise: ~420 kHz, 320 kHz
Record Vbias voltages before shutting them down
Vdiff = [-5.5 0.2772 27.9984 44.4698]
V(1:4) = [-5.77719 -5.2228 -16.4715 +72.4682]
Carrier peaks fit reasonably well to a sinc-squared model with B=1.52 uT (larger B-field than previously!)
Linewidth is Fourier-limited, ca. 0.8/0.45ms = 1800 Hz
Wide scan shows side-bands at both 525 kHz (why?) and 905 kHz (radial secular mode)
Leave long and wide run overnight, ca14 hours, to maybe see secular sidebands.
18 dB attenuation, tau_pi=tau_probe=450 us
Ion lost mid-way, but some useful data was collected
Shorter scans with sinc-squared visual fit:
24 dB att, 1 ms probe, 516 Hz FWHM AOM (0.8/1ms = 800 Hz)
ID6333 longer rabi-osc run 0.05 - 5 ms, 50pts, 500 shots, 18dB attenuation, high probe power
Try 209 MHz 422-detuning?
With high power now tau_pi = 350 us
350us gives 1071 Hz FWHM AOM (0.8/350 = 2285 Hz)
Test rabi-flop with different attenuations
18 dB -> tau_pi = 400 us, ID6322
24 dB -> tau_pi = 900 us, ID6323
27 dB -> tau_pi = 1400 us, ID6324
30 dB -> tau_pi = ? us, ID6325
Increase intra-cavity feedback set-point to 3.0V - IR-amplifier ca 360 mA
ID6313 scan at 2.5 ms probe, gives 277 Hz FWHM AOM
intra-cavity power stabilizer ON, 2.27Vset
ID6309 18 dB attenuation, 6 ms probe
ID6307 18 dB attenuation, 6ms probe -> 328 Hz FWHM (AOM)
24 dB attenuation, peak with 6 ms probe, ID6305 run5, ca 230 Hz FWHM (AOM) ? wide?
small-power Rabi osc ID6304 0.05 - 20 ms, 50 pts, 100 shots, 24 dB attenuation
Further reduce with DDS attenuation 24 dB
Reduce probe-power with waveplate (works only for A-beam) to approx 10 nW
search again for the peak (run4) ID6302
Try intra-cavity power stabilization, set-point 2.34V - does not seem to stabilize power at trap-breadboard?
back to 18dB attenuation, ca 510 nW probe, seems to fluctuate a lot?
ID6295 0.05 - 20 ms probe, 27 dB DDS attenuation, 100 shots
ID6294 Rabi-flop on 75,966785, 27 dB DDS attenuation, 100 shots
ID6292 run3, search for peak again, 27 dB probe attenuation, 1.6 ms probe
C=75,966785 (311 Hz FWHM AOM)
ID6291 Rabi-flop with 27dB probe-attenuation (35nW?). 0.05 - 5 ms, 100 shots - abort, doesn't look very good..
21dB 350 nW
24dB 120 nW
27dB 35 nW - try this?
Re-measure probe power w. 18dB DDS attenuation: 610 uW (2-fold compared to start of experiments??)
Rabi-flop with more statistics, 500 shots
ID6286 Center 75.966945, probe 0.050 to 5 ms in steps of 0.05 ms
run2/ID6284: 400us probe, 4 kHz span 50 Hz step. 81 frequencies.
This might be Fourier limited? Max quantum jumps as high as 66/100?
Center 75.966945, FWHM 975 Hz (AOM)
re-check the peak-center with tau_pi=400us ID6283 or 2020-04-27-run1, 2kHz span/100Hz step -> only 21pts to scan
ID6282 0.05 - 5 ms with 0.05 ms step (100pts), 100shots - oscillates with better contrast than before, at least out to 3-4 ms
ID6281 rabi-flop 0.05-10ms probe - now we have too few tau_p-points (250us increase per point) -> aliasing?
Now Rabi-oscillations, we should have 330nW/20nW = 16-fold more probe power
ID6280 first rabi-flop. Now visually first max at tau_pi = 400 us
Search for the +C1 peak
Found a peak higher than expected, 75.966911 (FWHM 1283 Hz) with 1 ms probe
uMotion contrasts (300 kphotons): A: 0.2% (45 s), B: 0.6% (90 s), C: 2% (69 s)
Seems OK, no need to re-run uMotion optimization
Set cooling detuning 210MHz
Ion! At 10:54Z, loading time 7 min.
Load ion. I = 1.55 A start at 10:47Z. 1.37 W dispenser power.
RF: 70 mVpp drive, 1.56 Vpp monitor.
405 power 3 mW at A exit.
461 power 175 uW at A exit.
422 power 0.450 uW at A exit. Reduced from maximum - for better temperature.
Optimized 674 nm power on breadboard. Now 330 nW at A beam exit.
Switched to Toptica as clock laser.
Fiber-noise stabilizer on. Intra-cavity power stabilizer not on
2020-04-24 First clock run!
livestream: 128x speedup version
livestream: 12 hour stream
On 2020-04-23 a manual peak-fit gave -1.5(??) Hz, now we get -9.9(5) Hz
Rough estimate of servo-noise, assuming 20 red/blue counts, poisson statistics, G=0.5 nodimensional gain, and 400 Hz FWHM
0.5 * 400/2 * sqrt(40)/40 / SRS = 4e-14
Observed noise is 5-fold larger (C2/C4 have worse poisson statistics)
Y-axis below should be "f - SRS / Hz". The time-series should show aging of AHM3, but this is ca -2e-15/day or 0.4 Hz, not visible
Visualization of the peak servo. In order to reach ion line-center with a downwards-drifting input-light, the servo compensates by
drifting upwards. If we flip the sign of 2*Zeeman_AOM this matches the cavity drift well (assuming small AHM-drift)
Traces offset for visual effect
Note: this shows drift-compensated cavity frequency filtered from 1s counter data with sgolayfilt(f_optical_red,order=2, framelen=61) to ca. 60s,
corresponding to the servo cycle duration.
Magnetic field. Why is there a bias of -3 nT in the C2 and C4 estimates, compared to C1?
Quantum jump counts. Stdev seems to follow Poisson statistics (?) and mean seems 50/50 red/blu distributed
Clock laser plots during the run.
Note1: the drift-compensation overcompensates, so the input to the peak-servo is a (slightly) decreasing frequency
Note2: the predicted zeeman_aom line-center below does not take into account AOM-shift (ca 1200Hz)
Peak servo run starting ca 58962.5
Qauntum jumps for the peaks:
Manual search for C1, C2, C4 to prime the servo
2.5 ms probe, this is tau_pi from the flipflops
with lower 422-power we now have ca 70 counts / 10ms
+C1 75,963436 (FWHM 234 Hz)
-C1 75.956509 (220)
+C2 75.966886 (240)
-C2 75.9530925 (224)
+C4 75.977215 (270)
-C4 75.9427147 (273)
B-field estimated around 1.23 uT
Cycle length approx. 64 s.
Measured 20nW probe in A-beam after chamber
ca 1 uW 422-power
ID6186 rabi-flop on +C1 at 75.96342 (shifted lower over lunch?)
0.05ms to 20ms probe in 40 steps, 500 shots
ID6187 detuning scan, 1000ms gate. background count estimate 350 cps
lower 422-power to 0.35 uW, this halves fluorescence from 12 kcps to 6 kcps @ 210MHz detuning
ID6190 detuning scan, 1000ms gate. background
ID6191 rabi-flop (same settings as ID6187)
Vopt = [0.27719 27.99845 44.46984]
Here 300kphot contrasts: A 0.1% (32s), B 0.5% (54s), C 1.7% (40s) - an improvement!
For previous C-matrices see 2020-04-16
Note this run of the NPL-optimizer produced a C-matrix with opposite sign compared to earlier runs
Cfit =-1*[ [-0.7486428 , -0.25501217, -0.16788016],
[-1.92252664, 0.26163298, -0.16973487],
[-0.86786103, -0.16630517, 0.19853788]]
V0 = [ 0.27719086, 27.99845308, 44.46984 ]
ID6172 NPLv3 scan 4x4x4 points over lunch. 100kphot and voltage ranges around current voltages
uMotion 300kphot: A 3.8% (31s), B 7.4% (76s), C 2.3% (40s)
detuning 210 MHz as before
+C1 peak with 8ms probe, A-beam (damped rabi-osc now shows max at around 8ms)
tried spectra with B and C beams, but Pexc is low. B and C beams might be less focused?
+C1 A-beam 8ms gives FWHM of 192 Hz AOM - not much narrower than previously...
Memory usage during the peak-servo was rising up to 75% - investigate...
Give servo-gain G in units of FWHM/2
Try reversing sign of gain and increasing to G=0.05 (G=1 should be optimal according to theory)
same starting centers/widths as y-day seem to work(?)
add also B-field computation based on peak-separation - field is around 1.3 uT
try G=0.5 (half of optimum, much higher than before) - seems to work
Various gains/settings tried.
Here +C1, -C1, and +C2, -C2 are shifted for visual effect. C1/2-centers are plotted without shifts
Archive 8-hour plots of clock-laser here
When the peak-servo works it should capture the residual cavity drift, here around +1 mHz/s AOM
Long-term cavity drift here +18.8 mHz/s and we (over)compensate at 2*(-10.45) mHz/s = -20.9 mHz/s
Drift compensation once per minute, a 60*(-10.45 mHz) = -627 mHz AOM step.
sign of gain might be wrong here. Low red counts high blu counts should push AOM up!? (not down as we see here?)
422-power was adjusted back to 'high' for these tests
3.5ms probe length, 100shots per frequency. 10ms detection
Scan of 4 peaks, cycle duration ca 79 seconds. 4 peaks * red/blu side = 8 frequencies to probe
leave this overnight. at 80 seconds per cycle we can run about 540 cycles in 12 hours - so program 600
Initial centers and FWHMs: (searching for these should be automated...)
C1+ 75.963553 0.000187
C1- 75.9562818 0.000174
C2+ 75.9671580 0.000149
C2- 75.952675 0.000179
Try to program a peak-lock-servo
While programming, saw mysterious dark disappearance act of the ion. Re-appeared maybe 5 min later... (a bit of 405nm was tried, but not sure if this had an effect?)
The C1 pair takes about 40s to probe (red and blu side of +C1 and -C1)
Try to lower ion-temperature, by reducing 422-power in the A-beam. Detuning kept at 210 MHz (-16 MHz optical)
1 uW after chamber, 120/4 bright/dark in 10ms. Rabiflop ID6044, Detuning-scan ID6046
0.3 uW after chamber, 60/2 bright/dark in 10ms. Rabiflop ID6054, Detuning-scan ID6051
Rabi-flops on +C1 frequency from below 75,96353 MHz
Locate the Zeeman peaks... (3.5ms probe, from yesterday)
+C1 75,963533 (sigma 201 Hz) - run1 (6023)
-C1 75.956234 (sigma 190) - run2 (6026)
C1 center 75,959874
Z-shift 3659 Hz AOM (1.307 uT)
10minute mean driftcorr-freq is -330,4185162 MHz
+C2 75,967154 (sigma 178 Hz) (run3)
-C2 75,952596 (sigma 203 Hz) (run4)
C2 center 75,959875
Z-shift 7279 Hz AOM (1.3002 uT)
+C4 75,978085 (sigma 236 Hz) (run5)
-C4 75,941685 (sigma 178 Hz) (run6)
C4 center 75,959885
Z-shift 18200 Hz (1.3003 uT)
at the end of this 6-peak quest we have drift-compensated 10min mean of -330.4185164 Mhz (a 0.2 Hz change from between C1 and C2 measurements)
AHM3-shift estimated at 1241.1 Hz, ion-frequency at mj_squared=35/12 is 1239.6 Hz. (need to write down more significant digits from measurements in the future)
Difference is -1.5 Hz or -3e-15
10:05Z Check micromotion, 300kphot: A 3.6% (32s), B 7.8% (56s), C 2.0% (41s)
10:00Z ion still stored. Overnight tracking-osc unlocked, and comb f0-signal seems weak
16:40Z end. Leave ion trapped, turn off PMT
And now for something copletely different, a note on the clock-laser performance over the past 4 hours:
ID6013 rabi-flop on -C1 at 75.95584 MHz. 0.1 to 20ms probe. 500 shots
C1 center is at 75.9594955 MHz Zeeman AOM.
Latest value for drift-compensated light is -330.417743 MHz SRS
Thus C1-line center is at: -330.417743 + 178.5 (fiber_AOM) + 2* 75.9594955 (Z_AOM) = +1248 Hz SRS
AHM3 model from Bulletin-Atomique is -2.74565e-12 - 1.98725e-15 * (MJD - 58939)
Now is roughly 58960.6, so we expect y_AHM3 = -2.78857e-12
Naively AHM3_shift = y_AHM3 * SRS = 1240.3 Hz
Remains to be explained (left as an exercise for the reader): observed +1248 Hz - expected 1240.3 = +7.7Hz (1.7e-14)
Gaussian fit to +C1: 75.963145, FWHM 224 Hz (2.5 ms probe)
estimate Z-shift of 3.5 kHz, so look for -C1 ca 7kHz lower?
-C1 seems to be at 75.955846
zeeman-split 7299 Hz AOM = 14598 Hz optical.
2/5*muB*B/h = 5598.5 Hz/uT, so we have a field of 1.304 uT
rabi-osc on -C1 suggest optimal exication at longer probe of 3.5ms
15:12Z micromotion contrasts, 300kphot, A 2.7% (34s), B 5.5% (57s), C 2.9% (54s)
ID5954 Rabi-flop, 500 shots, probe 0.1-20 ms, center 75,96318 MHz (+C1?)
Modified ARTIQ programs: rabi-flop v2 where probe-pulses in order, clock-cycle with 2nd detection window, clock-cycle-plotting + ARITQ-logic-analyzer (coreanalyzer)view
Ion still stored.
No good ideas for overnight experiments..
ID5866 10Hz scan of the line
run29 ID5868 with 8ms probe with 18dB attenuation (too much!)
run30 ID5869 with 8ms probe with 26dB attenuation
run31 ID5870 2.5ms probe, 18 dB att
ID5867 rabi-flop with probe-att 26 dB (maybe 10x less power)
ID5884 rabi flopping at 75,963525MHz (cavity-shifted while closing doors?)
2 ms probe
run26 close doors (partly) to the cavity-box, fiber-servo ON, intensity-servo ON
run24 fiber-stab ON (357 MHz from ARTIQ to mixer) and intensity-servo OFF
run23 fiber-stab OFF (178.5MHz from ARTIQ to AOM) and intensity-servo OFF
run22 fiber-stab OFF (178.5MHz from ARTIQ) and intensity-servo ON
run21 fiber-stab OFF and intensity-servo ON (free-running VCO - not the way to do this measurement!)
run20 fiber-stab ON and intensity-servo OFF
run19 fiber-stab and intensity-servo on
Shifts of +/-200 Hz due to touching cables going to the cavity -> cavity-system tilts??
ID5814 rabi-flop 75.9633 MHz, 500 shots, 100 pts, 5.3s/shot = 44min
run18 / ID5813, 1kHz scan, 25Hz resolution, 2ms probe
ID5812 rabi-flop 75.96325 MHz, A-beam. 0.1-20 ms, 50 pts. 100shots. 2.6s/shots =
run16 / ID5809, C-beam +C1 at 76.9633 MHz. ?/100 counts max
run15 / ID5807, B-beam +C1 at 76.9633 MHz. 0/100 counts max
run14 / ID5805, A-beam +C1 at 76.9633 MHz. 37/100 counts max
run13 / ID5802 C-beam
only 1 jump?? maybe repumper was overridden, always on??
run12 / ID5799: Try the B-beam, it has different power-level for both 422nm and 674nm
Only 2/100 jumps
Look for motional side-bands
run11/ ID5796 +C1 red side-band, 75.9633 - 0.457 = 75.5066
30 kHz range, 100Hz step, 301 pts, 4.6s/pt
+C1 Carrier 75.9633, 1 kHz scan 40Hz step, 25 pts. ID5790 / run6. MaxP=35/100
blue: 75.9633 + 0.9424 = 76.9057 ID5791 / run7 - no dark counts - (but this is the wrong range!? AOM is double pass!)
blue: 75.9633 + 0.9424/2 = 76.4345 ID5792 / run8 - single dark counts
blue: 75.9633 + 0.9424/2 = 76.4345 ID5793 / run9 - 3kHz range, 100Hz step, 5 dark-counts at 76.4333
run10 scan 76.42, 30kHz range, 100Hz step. This should capture both radial-lo and radial-hi modes? ID5794. 301 pts * 4.6s/pt = 23 minutes
Lunch-break clock-spectrum, run5, ID5789
36 kHz span, 40 Hz resolution, 900 pts, 4.6s/pt = 69 minutes
rabi-flop shows tau_pi = 2.5 ms - so use that!
ID5788 rabiflop again, 75.9632 MHz, 18 dB probe att. 0.1 - 10 ms probe, 50 pts.
500 shots, 2.4s/shot
clearout 10ms, cool 30ms, repumper 35ms, detect 10ms. bright/dark 120/10
Check peak ID5786, 5787 (20Hz AOMres)
ID5784 rabi flop at 75.96325 MHz. probe-att 18 dB - 0.1-5 ms probe, more points and 300 shots.
ca 2.3 seconds/run (50 points).
But here we might have lost the fiber-stabilization servo - because DDS-attenuation goes wrong..
ID5783 rabi flop at 75.96325 MHz. probe-att 18 dB
+C1 peak at 75.9632 MHz, 2 kHz span, 50 Hz resolution
41 points, 4.6 s/point
Good excitation-prob 37/100!
Fourier limit should be 0.8/3ms = 267 Hz
Clock-spectrum with A-beam
Bright/dark levels 120/10 in 10ms
Optimized clock-cycle with 30ms cooling, 35ms repumper.
3 ms probe, 75.9594 cecnter, 0.018 span, 100 Hz resolution. 18 dB probe attenuation
179 points, 4.6s/point = 14 minutes
Look for seculars with tickler
0.92 to 9.96 MHz scan, 4dB att, shows dip at 0.9426 (upper radial?), 2nd radial not visible with A-beam
B-beam scan 0.9 to 1.0 MHz: lower at 0.8964
5dB att / 20ms detection: 0.89565 and 0.9424 MHz
Vbias from last time (-5.5, 0.2563, 28.007, 44.332)
uMotion 300kphot: A 2.7% (29s), B 4.2% (54s), C 3.0% (54s)
Set detuning to 210 MHz. Fluorescence A strongest, B weakest, C middle.
07:27:30Z load with 1.55A (1.34 W, slow rise to 1.36W) - 07:38:00Z ion signal! (10min30s)
Camera exposure time 2 s, zoom out to see full image. Image has moving horizontal-stripes?
Detuning 185 MHz
A-beam power levels
- 405: 6mW max - adjust to ~2.5 mW for loading
- 461: 200 uW
- 422: 0.9 uW
Beat 23.1612 MHz
Driftcomp 79.762800 MHz
Predicted line-center at 79.9594 MHz
Intensity servo set-point 1.15V
RF 70mVpp 16.709MHz, monitor 1.57Vpp
Check 461nm. R=4.645kOhm. I=199.32mA (check with dispenser 5A, 2min)
Linear model for micromotion
S = C *[ V - V0]
where V0 are the optimal voltages
2020-04-16 C-matrix (V0 = [ 0.25631175, 28.00738916, 44.33204023])
[[ 0.6155015 , 0.25619819, 0.20220264],
[ 1.68776957, -0.2430226 , 0.16205872],
[ 0.66540325, 0.13719351, -0.19730336]]
Compare to 2019-12-19 C-matrix with V0 = [ 0.24380421, 26.67929547, 43.62223335]
[[ 0.6688455 0.22836817 0.14897131]
[ 1.36885973 -0.19586379 0.12287398]
[ 0.66160432 0.12331246 -0.15400281]]
The inverse of Cfit gives voltage-directions (as columns) that change only one contrast-value
These directions have changed by 2..7 degrees between the 2019-12-19 and 2020-04-16 datasets.
Moving in the voltage-direction VA changes only the A contrast (not B or C). In other words VA lies both in the B-optimal and the C-optimal planes.
Directions are (2019-12-19 and (2020-04-16) (V12dif, V34cm, V34dif)
NOTE: changing voltages by this amount gives a contrast-change of +1 (100%) - not recommended experimentally, try only 10% or 1% of this vector
- VA12 0.1239 2.4101 2.4619 / VA04 0.1235 2.1165 1.8881 (angle-change 4 deg)
- VB12 0.4418 -1.6631 0.5661 / VB04 0.3759 -1.2290 0.4130 (angle-change 2 deg)
- VC12 0.4723 1.0045 -3.6602 / VC04 0.4353 1.1596 -2.7941 (angle-change 7 deg)
Note1: VB and VC lie in the A-optimal plane (moving in VB or VC directions does not change A-contrast)
Note2: moving along VA moves us 'maximally' away from the A-optimal plane
All three planes and vectors:
Zeeman AOM attenuation vs. 674nm power, measurements from 2019-01-31
ca 14:30Z lost ion (why?) - shut down lasers
Not-so-optimized clock-cycle in use:
Estimated on/off for SRS shutter is 5ms, and for ASE 4ms-on, 8ms-off
Distribution of PMT counts from the run4 dataset
Poisson distributions fit by eye. 101 frequencies * 100 shots/freq = 10100 datapoints in histogram
Fits roughly 285 counts/20ms bright state, and maybe 12 counts/20ms dark state
Distribution of dark-counts inconclusive - maybe ambient light-conditions affect this
2 kHz range, 20Hz step, 101 points
40ms probe, 24 dB DDS attenuation
rabi-flop tests 5747, 5748
18 dB DDS attenuation gave max P at ca 2.5 ms
21 dB attenuation -> ID5749
24 dB attenuation -> ID5750
Inconclusive results - probability rises from 0 but then levels off (+noise). No clear oscillations
focus on the +C1 peak at AOM 75.9532 MHz
range 2 kHz, resolution 50 Hz, 41 points
search again for clock transition 2020-04-16-run2
same parameters as run1, but now we have the 2nd shield
duration 180 points * 9 s/pt = 27 min
Install 2nd magnetic shield. (no collar on the 1st shield)
Contrasts 300kphot: A 2% (27s), B 0.3% (43s), C 3.3% (52s)
Clock-transition search. 2020-04-16-run1 (ID 5729)
Bright 300, dark 45, in 20ms detection window
75.9595 MHz center, 18 kHz span, 100Hz resolution, 180 points.
20 ms probe length
Clock-laser cavity intensity servo. Set-point at 1.4V. Amplifier at 370mA gives measurement also around 1.4V
Amplifier current-limit is 450mA - servo may drive it beyond limit -> unlock
Fiber-stabilization servo, reboot ARTIQ-DDS producing 357 MHz reference signal (careful when setting DDS attenuation?)
Apply the optimal voltages (0.2563, 28.007, 44.332)
Now get: 300khpot A 5% (26seconds), B 5.5% (48seconds), C 3% (43s)
To get more fluorescence from B/C, set detuning to 210 MHz
(ID5717) Try NPL algorithm around the optimum. 3x3x3 = 27 points. ranges 0.2...0.3 27.5...28.5
A 9s, B 16s, C 14s = 40s/point * 27 points = 18 minutes
Vopt [0.2563, 28.0074, 44.332]
From visual fit, relative intensities A/B/C = 1.0 / 0.2 / 0.4
Fluorescence collection efficiency now 1/1150 vs. 1/500 in 2019-12-05 data.
NOTE: A-beam has enough power to produce a wider line! Lineshape is narrower with the weaker B and C beams.
Detuning scans, from 205MHz -20...+14 MHz, 51 points.
ID-5704, A-beam, 1000ms gate
ID-5705 is PMT counts/20ms with ASE on/off (at 205MHz)
ID-5707, B-beam, 1000ms gate, detuning scan
ID-5708, PMT counts/20ms, B-beam, ASE on/off, 205MHz
ID-5710, C-beam, 1000ms gate, detuning scan
ID-5711, C-beam PMT counts/20ms, ASE on/off
Now at (-5.5, 0.24, 28, 44.25)
Try 300kphot, to confirm optimal point:
A: 3.6 %, 3.3%, 3.9% (ca 32seconds to collect 300kphot)
B: 4.5 % (73 seconds)
C: 2.6 % (63 seconds)
Try moving in A-optimal direction (+0.3, -0.45) -> (28, 44.25) gives A 3.6%, B 2%, C 3%
Back to (0.24, 27.7, 44.7) -> A 3% (11s), B 12% (23s), C 11% (19s)
09:53Z result is Vopt ( 0.458, 29.067, 44.961)
This Vopt gives 100kphot contrasts A 48% (10seconds), B 18% (21seconds), C 20% (18seconds)
08:50Z Try NPLv3 program. ID5660, 4x4x4 matrix = 64 points
scan range 0.2..0.3 26.7..28.7 43.2..46.2
measurement time A 10s, B 26s, C 19s = 55s, so duration ca 60min (?)
repeat contrasts after B-focus adjustments
A 3%, B 13%, C 13%
Try to focus B-beam tighter, to get more fluorescence
lens 1/8-turn closer gave 80-100 counts/20ms.
1/16-turn closer gave ~90 counts
try 1/8-turn outwards: 60-80 counts
(A beam gives >200counts, all beams ca 250 counts)
scan V34dif to find A-optimum at (-5.5, 0.24, 27.7, 44.7)
100kphot A 2.9%, B (very weak) 13%, C 12%
try change of (+1, -1.5) = (27.7, 44.85)
A 4%, B 14%, C 14%
Try (25.7, 47.85) change is (-1, +1.5)
A 3%, B 67%, C 70%
A-optimum line slope is -1.5 V34diff / V34cm (result from 2019-12-04)
Look for A-optimum V34dif-scan, optimum at (-5.5, 0.24, 26.7, 46.35)
A 5%, B 57%, C 55%
Try voltages from 2019-12-19 (0.24, 26.7, 43.6)
A 48%, B (very weak, align?) 21%, C 7%
Vbias from yesterday (-5.5, 0.2668, 22.705, 45.994)
Contrasts (100kphot) A 74%, B 82%, C 75%
Blink test with 200uw to 500uw of 405nm. Looks OKish with 2 levels 400/40 bright/dark in 20ms
all beams ON give 150 counts/20ms at 185MHz detuning.
205 MHz detuning gives 380 coutns/20ms
(2nd try) more 405-power 2.56 mW
1.55A 7:13:30Z start. (1.33W) 07:22:07Z signal! (~9min 30s)
(1st try) 1.55A (1.32W .. 1.34W)
07:01Z start. stop 07:11Z (10min) - no signal (405-power drifted down to 0.9mW)
RF 70mVpp, monitor 1.61Vpp. No change in frequency 16.709MHz
A-beam through chamber:
422nm 1uW, detuning 185MHz for loading
405nm 1.2 mW (manually reduced fiber-coupling efficiency
Try to load with lower 405nm power. With the old laser we had around 1 mW of 405nm
Conclusion(?): new 405-laser significantly changed trap loading efficiency, in 4 tries we always got 2 or more ions (sometimes dark ions)
matlab figure from 2019-12-04 with A-opt and B-opt lines
modify this (offset lines to new values, same directions as before) and look for crossing of lines -> (0.2668, 22.705, 45.994)
Contrasts 100kphot, A 4%, B 7%, C 10%
strange blinking - two ions?
Vbias (0.2668, 24.7, 43)
A 5%, B 72%, C 64%%
Vbias (0.2668, 24, 43)
A 22%, B 64%, C 64%
(4th load) load with 1.55A (1.3W) - 5min13s gave signal
at 185MHz detuning blink test gives 100/30 bright/dark levels (in 20ms)
at 205 MHz more fluorescence 250/30 bright/dark in 20ms
(3rd load) load with 1.6A - weak signal - 5min 40s gave first a cloud, then suddenly 2 (?) cold ions?
blink test (500uW of 405nm) shows 3 levels ca 0, 175, and 350 counts in 20ms
emptied trap with 3Vpp of DC-modulation (26.8Vpp probe)
Test at what DC-modulation we loose the ion. RF is at 70mVpp, monitor 1.59Vpp. back to V34 = (24.8, 43)
1Vpp Dc-modulation (probe 9.2Vpp), ion moves sideways. sometimes large-jump, so we have a dark ion also?
405 restored fluorescence again
1.5Vpp (probe 13.5Vpp) - still see ion, moving sideways
2Vpp (probe 18.5Vpp) - still see ion
3Vpp - lost ion
strangely ion goes dark, blinks, and then is recovered by 405/461 light. This repeats a few times.
After optimizing beams, S-contrasts A=37%, B=77%, C=71%
Detuning now 205 MHz
Try DC-modulation at 1Vpp , probe 9.2Vpp. Not much movement of ion
(2nd load) Try to load with 1.6A
4min30s to signal - now maybe 1 ion?
Blink test looks OK (2 levels) with 400uW of 405nm, levels are bright/dark 60/20 counts in 20 ms
empty trap by lowering RF, glitch is between 37mVpp and 36mVpp
go back to 2019-12-19 bias-voltages of (-5.5, 0.2668, 26, 43)
Optimize beams, A and B-beams give fluorescence
DC-modulation 1Vpp 500mHz, HV-amp (nominal gain 10x) probe sees 6.9Vpp
2-ions move side-to-side 'a lot'
old bias voltages from last time (-5.5V, 0.2668, 10, 17)
(1st load) Try to load ion with 1.7A dispenser. 2s (!) exposure time on camera (no intensifier)
loaded 2 (?) ions in 2min50s
Blink-test seems to indicate 2 ions (blinking visible with lower 1mW 405-power)
With PMT and still lower 405-power of 500uW see three-level PMT-trace (0, 1, 2 ions)
ID5377 blink-test, 250uW 405-power. Mostly in 2-ion-fluorescence-state now. 200counts/20ms = 2 ions. ca 100counts/20ms = 1 ion. Dark is 20counts/20ms
A-beam power levels, through the chamber
- 422nm: 0.8 uW
- 461nm: 215 uW
- 405nm (after warm-up 10+ min): 5.6 mW
add DC-modulator circuit to helica-resonator. Lower resonance at 16.709 MHz. 12VDC-supply for DC-modulator HV-amp exchanged.
1.55 Vpp monitor voltage with 70mVpp from Keysight generator
422nm Rb-signal 4 DIVs at 500mV/DIV. AOM to 185 MHz
461nm check. R=4.645 kOhm, I=199.3 mA. check dispenser 5A for 2 min -> I=199.32mA
20:58Z ion lost. storage time roughly 3.5 days (stream: https://www.youtube.com/watch?v=zy_vQ-RwKdc)
16Z leave live-stream on for the weekend
Look for secular modes, with the tickler. Removed 2nd shield to check tickler-connection - but it was connected OK to artiq-DDS
Configuration: RF 70mVpp, monitor 1.71 Vpp, Vbias = (-5.5, 0.2668, 10, 17)
All beams ABC on
(see 2019-12-18) DDS att 4dB for the radial modes at 1 MHz, 10 dB for axial at 2 MHz
Axial mode attenuation can be increased, signal is very clear on PMT. 16..17dB attenuation works
With too much tickler, the ion heats up and takes a long time to recover into a cooled state
typical axial scan range 1.9 to 2.1 MHz, 201 points, 40 ms detection time
typical radial scan range 1.0 to 1.2 MHz, 401 points, 40ms gate, 4dB att
When no unit is given, assume MHz in the tables below
How does V12cm change z/r?
|70mVpp, -5.5 V12cm
|70mVpp, -4.5 V12cm
|70mVpp, -3.5 V12cm
How does RF-drive change seculars?
|65mVpp, -5.5 V12cm
||1.8408 (broad peak with 18 dB, narrower with 19dB)
|70mVpp, -5.5 V12cm (repeat of first row above)
|75mVpp, -5.5 V12cm (monitor 1.83 Vpp)
||2.1325 (nothing with 20dB, clear peak with 18dB
Lower radial mode seems harder to detect? The fluorescence dip is shallower, requires more tickler
12:01Z ion is still trapped.
Note: all measurements y-day were with B-beam, it gave the best fluorescence. Beam alignment is very sensitive...
run8 (5307): longer 60ms probe, 2 Hz resolution, 400 Hz span, 200pts a 12 s per point
This run was interrupted because the clock-laser unlocked from the cavity
Visual fit to run7 data gives ca 60 Hz 1sigma AOM. Roughly 240 Hz optical FWHM.
run7 (5308): focus on the +C2(?) peak, to get a linewidth
30ms probe, 10 Hz resolution, 1 kHz span, 101 pts
randomized scan! ca 9 seconds to probe 100 pulses
smaller peak at 75.9641 (as before) - this might be +C1
stronger peak at 75.96721 - this might be +C2
calculations indicate line-center at 75.9608 MHz
Try longer scan to see zeeman peaks
run6: 75.96413 center, 7 kHz span, 20 Hz step. 20ms probe.
No B-field-coil currents applied (I1,I2,I3 all zero)
Install 2nd B-shield (NOTE: no collar-metglas around 1st shield, and inner temp-sensor is loose, no door-caps on 2nd shield)
Try to scan peak with long 50ms probe and 5Hz AOM step
center 75.96413, span 500 Hz.
Try Rabi-flopping program
highest peak was at 75.96413 MHz, so probe there.
ID 5300, up to 50ms probe, max was 26 jumps at 27ms probe
ID 5301, shorter probes 1ms-20ms
no oscillations seen - inconclusive?
Search for the clock transition
From PMT-tests set bright_count=160, dark_count=20 (20ms detection window)
First try center 75.96 MHz, span 100 kHz, step 1 kHz, 101 points. 10ms probe pulse.
seem to find a peak at 75.964, measure that with 100Hz resolution
2020-04-02-run3 20ms probe, center 75.96412 MHz, 10Hz step, 1kHz range, data ID 5298
2020-04-02-run2 10ms probe, center 75.9462 MHz, 10Hz step, 1 kHz range, data ID 5297
clock-cycle program from 2019-12 "sweep clock cycle in kernel v2". Some adjustments to DDS frequency-setting were required (due to artiq-5 upgrade?)
The clock cycle has parameters dt_clearout=10, dt_cool=50, dt_repumper=60, dt_detect=20 (in milliseconds)
clock_probe=10000 (in microseconds
N_rep=100 (number of pulses, AOM_att=18dB (zeeman AOM)
(0.2668, 20, 17) gives A 7%, B 29%, C 11%
Try manual searching/scanning for optimum bias-voltages
Scan V34diff towards lower values, V34diff=17V seems to improve A-contrast to around 13%
Contrasts manually, at old optimum point V1 = -5.7668, V2 = -5.2332, V3 = -16.79, V4 = 70.21600000000001
cm/diff format: (-5.5, 0.2668, 26.713, 43.503)
A 31%, B 53%, C 32%
Try uMotion NPL-optimizer (v3 program), with old ranges (see 6x6x6 result from 2019-12-19). Now to -5.5 V12cm (old setting)
Algorithm gives strange optimum voltages [13.05783089 -7.38635953 18.00593577]
Vbias now at (-5, 0.2668, 26.713, 43.503) =(12cm, 12diff, 34cm, 34diff)
manual umotion contrasts A 30%, B 56%, C 38%
422-detuning was left at 205MHz yesterday and not touched since
remove adjustable-iris and try 400um pinhole. Visual inspection indicates it is slightly smaller than the adjustable iris.
initially gives ca. around 200/40 bright/dark
In order to move the imaging-objective freely the 1st B-shield needs to be 'loose' i.e. not attached with the plastic M6 screws (but there is a chance of blocking beams if we move shield too much!)
0713Z PMT SNR - could the fixed pinholes improve this? (we have 50/100/200/400um pinholes to try)
Start of day bright/dark 300/80 in 20ms gate - with the variable iris set to maybe 1mm (?) diameter - lever-position marked with pencil.
Tried pixelfly without camera-objective, with sensor placed where the intensifier was previously. But this did not give any better image
reinstall camera-objective to where it was.
Clock-laser was unlocked from cavity at around UTC midnight
Was left on ca 18 local Wed evening
2020-04-01 April Fools ion
Tried uMotion histograms, ca 10-20% contrast with all beams
70/30 splitter for PMT reinstalled. Align by placing fiber-end(red-helper-laser) at PMT position, and look for the fiber-spot on the wide-field camera (IR window). 422-bandpass removed for this
PMT SNR maybe around 5, i.e. bright/dark in 20ms gate-time approx 300/60. Previously SNR was maybe 10, i.e. 500/50 photons in 20ms gate time
Image intesifier suddenly broke??
Intensifier removed, without moving pixelfly+camera-obj the ion can be seen, but image is not so great
11:15Z single ion loaded with 1.7A dispenser in 2 min 30s
Optimization of beams with ion-cloud (dispenser at 2.0A continuously during optimization)
Higher trapping voltage (110mVpp from siggen) may help locate the center of the cloud
Initially A and C beams cool the ion, after alignment also B-beam
V12cm = -5 V was set (similar to 2019-12 setting)
Moving around ASE-beam further increased fluorescence
Power levels at A-beam output (after chamber)
461: 230 uW
422: 0.9 uW
405: 7 mW
RF drive 16.773 MHz, 70mVpp gives 1.79 Vpp monitor as usual. 2nd monitor 930mVpp
Check 461 against Dispenser (5 A for 2 min). 4.645kOhm / 199.3mA
After optmizations Rb error-signal 4 DIVs peak-peak, about 2x larger than before
beat note 23.162593(40) MHz
Cavity -489.980 MHz SRS
Trapping and optimization efforts.
The PMT/Camera 70/30 splitter was removed.
Ca 2 minutes at 2.0A consitently resulted in a diffuse cloud
Image is of beams, rather than trap. Cloud fluorescence lifetime ~1-2 minutes.
Both A-beam and ASE-beam were focused to their original (small spot) settings
The Ando OSA wavelength accuracy was checked against Rb-locked 422 laser
With 185 MHz AOM detuning the reading is 421.3 nm - Sr+ cooling is 412.67 nm (paper on lab wall)
Try also enlarging the ASE-beam, by defocusing 1/4-turn.
Power levels at A-beam output
461: 215 uW
422: 0.3 uW (ca 4uW before A-fiber)
aligning fiberlaunch from 422 laser-breadboard helps, to 9uW before A-fiber
After alignment 422 at exit: 0.8 uW
405: >5 mW after warmup
Previously optimized trap frequency: 16.773 MHz
70 mVpp run, 1.79 Vpp monitor (probe1), 920 mVpp (probe2)
Coolinglaser. Detuning 185 MHz as usual for loading. Rb error-signal ca 2 DIV-peak-peak (quite small?)
Usual settings of I_ld 69.6 mA, and I_amp = -1800 mA
Defocusing A-beam objective by 1/4-turn enlarges the beam to (over)fill the inner-electrode to outer-electrode gap - spot diameter maybe 2x larger than previously.
461-settings to old values. R_tec=4.645 kOhm, I_ld around 200mA - check against dispenser (run for 2min at 5A): 199.3 mA
no luck with trap loading...
re-align camera (B-shield removal requires objective-removal), focus +750/2 um in from electrode focus.
Measured power-levels, A-beam, through chamber
422nm: 0.6 uW (Rb error-signal 500mV/DIV on scope, signal is ca 3 divs peak-peak)
461nm: 250 uW (Rtec=4.603kOhm, Ilaser=198.3 mA)
405nm: >5 mW (requires adjustment as laser warms up?)
New 405-pointer warm-up test with OSA. Spectrum does not shift much during 10 minutes.
Maximum at 405.1 nm - should match Sr ionization transition well?
Both 1st and 2nd B-shields now removed - to align beams onto ion
Wide-field camera installed - allows seeing ASE-spot (and blue spots) on electrode surfaces, for aligning
New 405-pointer, measured 404.5 nm
Fbeat 23.162962 MHz
Fcavity -490.006128 Hz
Clock-laser, old seed laser
Beat 23.162984 MHz
Fcavity -490.006425 MHz SRS
waiting for 405nm replacement...
clock-laser + beat was ok for almost 16 hours. Then laser unlocked (why?)
applied drift-correction +20.9 mHz/s is OK.
shutting down laser for now.
0720Z: Clock laser was locked and beat-note OK over night. Tracker unlocked (why?) in the moring ca 9 local
Beat note monitor (Siglent SA) shows <-25dBm peak - maybe too low for tracker?
New applied drift-correction of +20.9 mHz/s is roughly correct.
clocklaser TEC=17.176 kOhm, I=160.1mA. 1348-amp I=355.1mA, TEC=12.943kOhm
optimize the beat note with the comb waveplates, maybe 5dB improvement to -21 dBm @ 1kHz RBW now.
clock-laser TEC 17.175 kOhm. Leave locked
Could not load ion. Tried up to 2.2 A dispenser. Removed 2nd B-shield to better align A-beam.
RF drive 70 mVpp as usual, monitor is 1.77Vpp (alternative monitor 920mVpp). Resonance max is at 16.772 MHz
NOTE: 10 MHz from comb-divider-distributor was in bad shape... needs fixing!
Camera re-aligned to electrodes + 750um/2 focus.
Check 422 Rb-lines OK, check 461nm helper-dispenser (I=198.46 mA now after check)
First loading attempt (before lunch) did not work. Idisp=1.60A for 12 min, but no signal.
Cooling laser: I=69.6mA, Amp -1850 mA.
0820Z re-lock laser to cavity mode 1348.052 nm (-9.2 dBm at wavemeter). LD 160.1 mA, TEC 17.176 kOhm
1348nm Amp TEC=12.943 kOhm, I=355.1 mA
1525Z clock-laser has now been locked to cavity, and comb beat-note OK, for 24+ hours. Good.
Observed cavity drift +20.9 mHz/s
Applied correction is too large +27.5 mHz/s, by about +6.6 mHz/s
Apply a new (differential) correction of 2*(-3.3) mHz, to get a new total applied correction of +20.9 mHz/s
active from around 1525Z
Factors of 2 here are due to the double-pass AOM. Shift from DDS is applied twice to 674nm entering cavity.
All AOMs effectively shift up, from the cavity at -490 MHz SRS, to reach line-center at ca 0 Hz SRS at the ion.
0850Z comb f0-lock had unlocked overnight. re-align waveplates (30dB f0 peak) and re-lock.
optimize beat-note, but not much improvement, re-lock tracker.
laser had apparently stayed locked since friday afternoon. leave running.
seed TEC now at 17.178 kOhm (this is the one that eventually difts to a limit?).
1520Z leave over weekend
1350Z Switch back to old seed-laser at 1348nm + fiber pre-stabilizer
lock to the right cavity-mode, fbeat 23.163 MHz
The Toptica stays locked to the cavity for short periods. It is not clear how to use the "unlimited slow integrator" analog output from the servo as an input to the digital laser controller. The input can be used for controlling laser current, piezo, or temperature, with adjustable sensitivity. Additionally the Toptica seems sensitive to acoustic vibrations such as the lab-doors, and doors outside the lab.
1355Z try lower gain of +1mA/V for slow servo?
unlock - why??
13:40Z FineIn1 has drifted to about -0.29 V
12:02Z try larger range for ULI, now +/-5V (spec), +/-4V maybe in real life (was on +/-0.5V setting)
With larger range, need to slow down integrator speed
Fine-in1 controls current, +5 mA/V as before
Stays locked for about 2hours. ULI goes to limit of +/-0.4V (this FALC output is connected to "Fine in1" on toptica)
0839Z re-lock, try ARC 5mA/V.
0830Z laser has jumped to TEM00 at 1348.053nm
stayed locked for 2-4 hours. ARC was 3mA/V
17:07Z leave locked overnight... now also logging siglent SA spectra
Morning: Laser again in TEM02 mode, 1348.056nm
comb seems ok
2020-02-02, palindromic Sunday
beat-note OK for <1 hour...
12:48Z re-lock clock-laser to TEM00 1348.052nm, beat 23.13MHz
re-lock tracer at ca 23.1606 MHz
12:37Z clock-laser now locked to TEM02-mode (1348.045nm - far away!)? (left it Friday locked to TEM00).
also comb rep-rate lock unlocked?/strange? -> re-lock comb rep-rate
Measurement of Zeeman-AOM attenuation (dB) vs. optical power (W) on the breadboard, and estimated power delivered to ion
*TODO results to be inserted here*
405nm laser seems broken :( only 7 uW output!??
Beat-note now 23.1602 MHz, up from 23.150 on 2019-12-17 and 23.145 on 2019-11-29
NOTE comb measures drift-corrected light, not cavity! Residual drift is now:
10 kHz (4.5e-11) per month = 300Hz (1.5e-12) per day = 4 mHz/s (IR) = 8 mHz/s (RED) roughly?
Cavity at around -490.0944 MHz SRS. Short-term drift 24 mHz/s
Drift-corrected light at -330.4160 MHz SRS, residual drfit -2 mHz/s (RED).
After service-break, all systems seem operational, with imrovements/changes
- pt100 sensors added
- new PDH detector, S5973 phototiode, OPA657 opamp, 8k2 RF and +11V/V postgain
- fiber noise-cancellation (2*178.5MHz reference freq from ARTIQ)
- removed DC-mod circuit from hel.res. Now using N-short. This increases f_res.
- upgraded ARTIQ version
- moved Zeeman AOM-amplifier (heat source!) up from optical table to top left rack.
- Toptica as clock-laser
For 2019 results see index_2019