2018/09/20 (Thursday)
| Setup: DIB1/3/4 → 3.2 MHz filters 2/3/1 → cable 3/1/2 → gender changer 3/?/2 → HVConn 1/3/2 → HV 1/2 → DCRC55(1)/DCRC50(3) 1.1kV → NaI B/C → Ortec 452(gain=5, shaping = 6 us, output=10V) / Ortec 575A(gain=10, shaping = 3us) → x10/x10 attenuator → DCRCX(4) DCRC33(2) → kepco 1kV power supply → HV input |
|
| status: | Realize only now that the NaI amplifier polarities are incorrect. They have both been set to pos when the NaI pulses are actually negative. It could be that the amplifiers are linear for some distance in the opposite polarity, but then saturate. Either way, that sucks since all our data was taken this way. Further, none of the test data taken yesterday managed to reach the same pulse amplitude as 2.6 MeV events. We need a fast pulse source that can produce output amplitudes of -20 V to -30 V to do this test. Still further, this morning the display on the TDS3034B oscilloscope is going haywire. There was nothing plugged into it when this started, and a reset doesn't help… |
| 0935 | Tower/MC/Coldplate = 29.8/30.2/185 mK, Still = 1.0 mW / 1.3±0.1 K / 0.080 mbar, G1/G2 = 127/122 mbar, 1K Pot = 1.60 K / 17 mbar, Turbo on IVC = 0.86 nbar, MB = 34.6% |
| 0938 | Fill LN2 cold trap bath. |
| 1445 | Commence LHe transfer, MB = 27.1% |
| 1500 | After a bit of archeology, found what seems to be a design document for the ebox Dsub-50 feed through bulkhead. Here is a photo of the document which was found in the “Final Designs” folder of the green filing cabinet in lab. It lists a company, “CeramTec” and part number “16804-01-W” for the dsub feed through. Here is the CeramTec catalog which lists that part having a 500 V DC rating. |
| 1519 | Start 40 min NaI amplifier bkg dataset with NaI detector: Agilent 81101A → Ortec 452(gain=5, shaping = 6 us, output=10V, Neg) / Ortec 575A(gain=10, shaping = 3us, Neg) → x10/x10 attenuator → DCRC4PB/PA, Driver Gains = 1/1, PhononWaveFormLength = 1024 (512 samples), PhononPrepulseSamples = 64 (64 samples) 1000 BOR randoms, trigger on DCRC4pA/B threshold = 5/5 |
| 1530 | Unplug HV from HV adapter board to calibrate HV output. Turn DCRC2Q[0] down to 0, this corresponds to max HV output. Turn up HV at Kepco power supply until HV = 250 V This requires adjusting the coarse and fine control knobs. Then, adjust DCRC2Q[0] and record the HV output. The values obtained (with slight modification, see below) are also saved in the HV calibration file /home/hep/cdmssoft/scdmsDAQ/MidasDAQ/online/scripts/pysequencer/HV_DCRC2_control_cals_250V_180920.txt This calibration file should be used to generate scripts with the current HV setup. |
| DCRC2QI (V) | HV (V) |
| 11.52 | -0.032 |
| 11.51 | 0.145 |
| 11.5 | 0.346 |
| 11.45 | 1.37 |
| 11.4 | 2.42 |
| 11.25 | 5.66 |
| 11 | 11.08 |
| 10.5 | 21.95 |
| 10 | 32.84 |
| 9.5 | 43.67 |
| 9 | 54.54 |
| 8.5 | 65.39 |
| 8 | 76.26 |
| 7.5 | 87.1 |
| 7 | 97.97 |
| 6.5 | 108.85 |
| 6 | 119.69 |
| 5.5 | 130.56 |
| 5 | 141.4 |
| 4.5 | 152.28 |
| 4 | 163.11 |
| 3.5 | 173.99 |
| 3 | 184.87 |
| 2.5 | 195.71 |
| 2 | 206.58 |
| 1.5 | 217.42 |
| 1 | 228.3 |
| 0.5 | 239.14 |
| 0 | 250.01 |
| 1600 | Plug HV back into HV board. HV output was at 0.1V, rises to 0.5V when plugged in. Adjust DCRC2Q[0] to 11.54 v to set HV=0.005 V. Seems like the calibration might be a little different with a load. |
|---|---|
| 1604 | Let's turn up the bias while watching DCRC1PA and see what happens! Also turn on the logger in case it's interesting later. Triggering on 5Hz randoms. Measuring leakage a few seconds after changing bias, reading Vpp off of SQUID Tuning GUI |
| DCRC2QI (V) | HV (V) | Vleak (mV) | DCRC1PA noise Vpp (V) | Notes |
|---|---|---|---|---|
| 11.54 | 0.004 | 0.006 | ||
| 11.25 | 5.8 | 0.6 | 0.006 | |
| 11 | 11.1 | 10 | 0.006 | |
| 10.5 | 21.95 | 25 | 0.006 | |
| 10.1 | 30.6 | 28 | 0.006 | |
| 9.6 | 41.5 | 31 | 0.007 | |
| 9.2 | 50.2 | 35 | 0.007 | |
| 8.7 | 61 | 37 | 0.007 | |
| 8.2 | 71.9 | 40 | 0.0075 | |
| 7.8 | 80.6 | 40 | 0.01* | spikey, still pulses |
| 7.3 | 91.4 | 39 | 0.03* | 1615 |
| 1617 | Set HV = 0 and switch polarity |
|---|
| DCRC2QI (V) | HV (V) | Vleak (mV) | DCRC1PA noise Vpp (V) | Notes |
|---|---|---|---|---|
| 11.54 | -0.28 | -55 | 0.005 | |
| 11.2 | -7 | -23 | 0.006 | |
| 11 | -11.2 | -22 | 0.006 | |
| 10.5 | -21.96 | -30 | 0.006 | |
| 10.1 | -30.64 | -35 | 0.006 | |
| 9.6 | -41.5 | -45 | 0.006 | |
| 9.2 | -50.2 | -48 | 0.007 | |
| 8.7 | -61.04 | -48 | 0.007 | |
| 8.2 | -71.9 | -47 | 0.006 | |
| 7.8 | -80.6 | -46 | 0.006 | |
| 7.3 | -91.4 | -50 | 0.006 | |
| 6.9 | -100.2 | -48 | 0.007 | |
| 6.4 | -111 | -42 | 0.007 | |
| 5.8 | -124 | -48 | 0.007 | |
| 5.5 | -130 | -53 | 0.006 | |
| 5 | -141 | -60 | 0.007 | |
| 4.6 | -150.1 | -65 | 0.007 | |
| 4.1 | -161 | -67 | 0.006 | |
| 3.6 | -171.8 | -65 | 0.007 | |
| 3.2 | -180.5 | -68 | 0.007 | |
| 2.7 | -191.4 | -69 | 0.007 | |
| 2.2 | -202.2 | -71 | 0.008 | |
| 1.8 | -210.9 | -70 | 0.008* | spikey, still pulses |
| 1.3 | 221.8 | -75 | 0.009* | |
| 0.9 | 230.4 | -75 | 0.008* | |
| 0.4 | 241.3 | -82 | 0.009* | |
| 0 | 250 | -83 | 0.01* | 1638 |
| 0 | 250 | -58 | 0.008* | 1642 - leakage flattening out, open LN2 purge further |
| 0 | 250 | -56 | 0.008* | 1650 |
| 1652 | Turn down HV bias using change_HV_250_to_0_DCRC2_250V.msl |
|---|---|
| 1659 | Fill LN2 cold trap full. |
| 1700 | Sequence finishes. |
| 1705 | DCRC2Q[0]=11.518, HV=-0.62V. Play with this a little bit manually to find a better zero in the negative polarity with the HV board connected. Update the 2 points around 0V in the HV calibration file /home/hep/cdmssoft/scdmsDAQ/MidasDAQ/online/scripts/pysequencer/HV_DCRC2_control_cals_250V_180920.txt A copy of the file is here. |
| 1711 | Flash for 30s |
| 1714 | Wrote a sequence to ramp and take a dataset using these new calibrations. It's called HVscan-240×1-4_5PreBias-DCRC2_250–30Flash-30wait-2hrData-10minSubRun.msl Will run this when the detector cools. |
| 1742 | Start 2 hr, -240 V bkg dataset with 4% prebias (-250V) for 5 min, split into 10 min subruns, with CDMS and NaI detectors running and DCRC2 as HV controller. Using sequence HVscan-240×1-4_5PreBias-DCRC2_250–30Flash-30wait-2hrData-10minSubRun.msl CDMS detector: Driver Gains = 2x, QET biases = (280/260/305)/(140/280/295) uA, Front-End Gains = (10/10/10)/(10/15/10) NaI detector: 1.1kV → NaI B/C → Ortec 452(gain=5, shaping = 6 us, output=10V, Neg) / Ortec 575A(gain=10, shaping = 3us, Neg) → x10/x10 attenuator → DCRC4PB/PA, Driver Gains = 1/1, PhononWaveFormLength = 1024 (512 samples), PhononPrepulseSamples = 64 (64 samples) 1000 BOR randoms, trigger on DCRC3pB/C/D threshold = 7/10/9 |
| 2020 | Data stopped showing up after first dump of 8th series. Looks like MIDAS webpage is down, also looks like K100 Status temperature/voltage page is not updating. |
| 2025 | MIDAS was completely down. Restart MIDAS. Start DAQ programs. Sequencer is restarting where it left off. |
| 2050 | Data is being recorded but the K100 Status page is still not updating. |
| 2110 | The last four series (after the MIDAS crash) produced empty files. So something was/is still wrong. (turns out there was a power failure…) |
| 2120 | There was a power failure somewhere around 1907. Got in to lab at ~2100, pumps had turned off, IGH had reset and closed all valves, tower was sitting at 500mK. G1/G2 were around 100mbar, P1 was off scale, opened V3, G1 decreased, so fridge pressures were still low. Somehow the turbo pump was running (again? still?) and the pressure was at 1 nbar. Restarted pump power, restarted 1K pot, circulation pump, and opened valves and opening V6 to start circulation through cold trap 2. |
| 2135 | Computers back up, restarted clean room. Tower = 170 mK and falling, P1 is now back down to < 1mbar, G1/G2=156/157 mbar, V6=16.5%. Open V6 fully. |
| 2143 | Tower = 85 mK, G1/G2 = 149/151 mbar, 1K Pot = 1.60 K / 18 mbar |
| 2147 | Power cycle DCRCs and restart MIDAS. Get a little temperature spike to 170 mK |
| 2157 | Played with MIDAS a little bit, DCRCs seem to be working fine, can even see pulses on some traces. SQUIDTuningGUI keeps crashing though. Whatever. |
| 2158 | Tower = 35.4 mK, G1/G2 = 128/131 mbar, 1K Pot = 1.60 K / 19 mbar |