2017/10/18 (Wednesday)
| Status: Initial LHe transfer yesterday | |
| 0815 | MC(Cernox)/Sorb/1K Pot = 8.33/8.2/6.34 K, Fridge(G1) = 564 mbar, MB = 75.2%. Turbo = 2.2 nbar pumping on IVC. IVC = 19 nbar. Stop pumping on IVC. Sorb & Pot temp & fridge pressure increased overnight, Cernox decreased. |
| 0825 | MC(H35) = 8.20 K |
| 0825 | Pump out line to 1K Pot, then isolate pump, then open 1K Pot speedivalve. 1K Pot = 0.25 bar. Temperature dips to 4.5 K then rebounds back to 5.9 K. |
| 0830 | Turn on 4He pump, open the two valves to connect its exhaust to recovery. Start pumping on 1K Pot, → 0.04 mbar. Dips to 5 K then rebounds to 6.1 K. |
| 0835 | Open 1K Pot Needle Valve to 10 degrees. 1K Pot → 3.5 K / 1.0 mbar Open Needle Valve a few more degrees. 1K Pot → 3.3 K / 2.4 mbar |
| 0840 | Open Needle Valve a few more degrees. 1K Pot → 3.1 K / 3.6 mbar Open Needle Valve a few more degrees. 1K Pot → 3.0 K / 5.0 mbar Open Needle Valve a few more degrees. 1K Pot → 2.5 K / 8.0 mbar |
| 0845 | Open Needle Valve a few more degrees. 1K Pot → 2.3 K / 22.0 mbar Open Needle Valve a few more degrees. 1K Pot → 1.7 K / 34.0 mbar Open Needle Valve a few more degrees. 1K Pot → 1.8 K / 55 mbar |
| 0850 | MC(H35)/Sorb/1K Pot = 5.5(falling)/<4.0(min)/1.71 K, Fridge(G1) = 443 mbar (falling) |
| 0855 | MC(H35)/Tower = 3.16(falling)/3.17 K , Fridge(G1) = 344 mbar (falling) |
| 0910 | Tower = 2.38(falling) K , Fridge(G1) = 92(falling) mbar. 1K Pot = 1.62 K / 24.0 mbar. |
| 0925 | 1K Pot surge for no reason. Pressure rose above 500 mbar, then fell again to about 50 mbar. |
| 0925 | Tower = 2.20(flat) K , Fridge(G1) = 62(falling slowly) mbar. 1K Pot = 1.69 K / 55.0 mbar (stabilizing). |
| Tower temperature was flat but after 1K Pot surge it began to fall again. | |
| 0945 | Tower = 2.07(flat) K , Fridge(G1) = 59(falling slowly) mbar. 1K Pot = 1.69 K / 55.0 mbar, pressure oscillating. |
| 0950 | Close V9 leaving 59.6 mbar in dump. Close V3. Turn on 3He pump, which will push a little more mix into condenser. Fridge pressures rise a little then fall again. |
| 0950 | MC(H35)/Coldplate/Still = 2.00/3.05/2.32 K |
| 0955 | Start circulation through Trap 1. V6 → 12.8%. Condenser peaks at 125 mbar. 1K Pot pressure spike to 350 mbar. Tower falls to 1.90 K. |
| 1000 | V6 → 13.2%. Condenser peaks at 127 mbar. Tower → 1.70 K. |
| 1000 | MC(H35)/Coldplate/Still = 1.58/2.47/1.86 K |
| 1005 | V6 → 13.5%. Condenser peaks at 129 mbar. Still → 400 mbar. Tower → 1.55 K. |
| 1010 | V6 → 13.8%. Condenser peaks at 130 mbar. Still → 10 mbar. Tower → 1.36 K. |
| 1020 | V6 → 14.2%. Condenser peaks at 134 mbar. Still → 5 mbar. Tower → 1.2 K. |
| 1022 | MC(H35)/Coldplate/Still = 1.04/1.29/1.22 K |
| 1025 | V6 → 15%. Condenser at 141 mbar and slowly climbing. Still → 1.7 mbar. Tower → 1.0 K. |
| 1033 | 1K Pot = 1.69K, 65 mbar. Pretty steady. |
| 1034 | Condenser flat at 138 mbar. Still → 1.6 mbar. Tower → 940 mK. |
| 1035 | MC(H35)/Coldplate/Still = 0.637/0.690/1.25 K |
| 1040 | V6 → 100%. Condenser flat at 137 mbar. Still → 0.22 mbar. Tower → 765 mK. |
| 1043 | MC(H35)/Coldplate/Still = 0.353/0.440/1.13 K |
| looks like the Tower/detector is taking longer to cool this time probably because we started pumping out exchange gas a little early, so it didn't cool to 4K. We'll watch the MC thermometer. | |
| 1110 | Condenser = 119 mbar, Still = 0.19 mbar, MC(H35)/Tower/Coldplate = 207/686/377 mK |
| 1145 | Condenser = 120 mbar, Still = 0.19 mbar, MC(H35)/Tower/Coldplate = 201/665/370 mK |
| 1150 | Open V14 to let the rest of the mixture from the trap into circulation. Condenser peaks at 150 mbar, Still peaks at 850 mbar, MC peaks at 232 mK then falls. |
| 1150 | 1K Pot = 1.68 K / 50.0 mbar |
| 1210 | Plug DCRC55(1) into DIB1, with appropriate rerouting board. Start MIDAS DAQ and pulseviewer. Trigger randoms 20 Hz. Both qI and qO traces are live, no noticeable pulses. |
| 1215 | Tried changing trace length in pulse viewer to a power-of-2 but gui crashed. So instead, changing trace lengths in ODB, then restarting the gui. Sticking to powers-of-2 trace lengths. |
| 1225 | With standard trace length try applying 5 V bias to external electrode. Trigger on charge channels with threshold 100 ADC - see strong inverted pulses at low-ish rate. Works with either qI or qO triggering or both. |
| 1230 | Condenser = 114(flat) mbar, Still = 0.45 mbar (V14 open), MC(H35)/Tower/Coldplate = 195/666/376 mK. Close V14. |
| 1240 | Turn off MIDAS programs and power down DCRC1. Switch DCRC1 to DIB3, look at qO. Trace looks fine. Apply 5 V bias to external electrode and try triggering on qO with 100 ADC threshold, see strong inverted pulses. Also try -5 V bias, see strong erect pulses. |
| 1250 | Turn off MIDAS programs and power down DCRC1. Switch DCRC1 to DIB4, look at qI and qO. Trace looks fine. Apply -5 V bias to external electrode and try triggering on qO and qI and both with 100 ADC threshold, see strong erect pulses. |
| 1340 | Condenser = 114(flat) mbar, Still = 0.18 mbar, MC(H35)/Tower/Coldplate = 186/625/354 mK. 1K Pot = 1.68 K / 50.0 mbar. |
| 1405 | Try DCRC6 on DIB4. Looks like the charge channels have much higher white noise. Also baselines were way up around 6 V. On DCRC_UI page, change Vlock of both channels to -0.1. This gets closer to the middle and seems to fix the noise, looks like the others now. See strong erect pulses with -5 V bias. But something's a little weird with triggering on qI. It seems like it's triggering randomly even at high thresholds, and turning off the qI trigger doesn't seem to make it stop triggering. qO looks like it's maybe fine. Actually it looks like both behave strangely - it's like with threshold triggering there's also a lot of random triggering. |
| 1435 | Try DCRC6 on DIB4 but use ODB settings from DCRC1 (just change its IP address). Similar trigger problems. Looks a little different now: turning trigger on/off starts/stops triggering, but no matter how high the threshold it looks like it's triggering randomly. |
| 1450 | Try reading the registers via telnet. Looks like the behavior is as expected there - if the charge trigger is set, rd 17 changes accordingly, and rt shows trigger pointers (more if threshold low, less if high). But even when all triggering is off, the pulse viewer still shows updating traces, even though the telnet session indicates that the dcrc has no trigger pointers. |
| 1500 | Trying DCRCs 1/2/3. |
| Setup: DIB 1/3/4 → 3.2 MHz filters 1/2/3 → cable 1/3/2 → gender changers → rerouting boards 1/3/4 → DCRC55(1)/33(2)/50(3) | |
| 1505 | Try turning on all 5 charge triggers each with 100 ADC threshold. Looks ok in pulse viewer. |
| 1511 | Start series of data with Vbias(external) = 10.05 V, -0.01 mV Vtest, with 1000 beginning randoms, and a 100 ADC bin threshold trigger all 5 charge channels. |
| 1518 | Looking with BatViewer, seeing good pulses but also a lot of noise traces. Maybe one channel's threshold is too low. Increase all thresholds from 100→150 ADC. |
| This doesn't improve situation much. Wondering whether the pulse viewer is causing it to trigger on test signal. Turn off pulse viewer, stop run, stop and restart MIDAS programs. | |
| 1523 | Start series of data with Vbias(external) = 10.05 V, -0.01 mV Vtest, with 1000 beginning randoms, and a 100 ADC bin threshold trigger all 5 charge channels. |
| 1525 | Condenser = 114 mbar, Still = 0.18 mbar, MC(H35)/Tower/Coldplate = 178/594/423 mK. 1K Pot = 1.68 K / 50.0 mbar. Coldplate getting warmer, that's a little strange. |
| 1530 | Look at new data with BatViewer - it looks much better! All events look like good pulses and all channels seem to be showing good pulses. |
| 1545 | Over 20 minutes, the dump sizes have fallen a bit, from about 25MB to about 14MB. Losing neutralization. |
| 1545 | Experiment with trigger windows, because the high frequency noise is very sinusoidal - so try to filter that out. The default seems to be ChargeTriggerBaselineLength = 16, ChargeTriggerSampleLength = 5 Try doubling these. |
| 1555 | Doubling the trigger windows cut the dump sizes down by around a factor of 2, so it certainly had an effect. Wonder if the ADC threshold is a sum of samples… if so should halve it. Try cutting all the thresholds from 150→75 ADC. |
| 1600 | Dump sizes have risen again, probably close to what they were before the window change (a little lower but following the trend of loss of neutralization). Looking at events they all look like strong pulses so we aren't triggering in the noise much yet. |
| 1603 | Try reducing all thresholds from 75→50 ADC. |
| 1610 | Data still looks good, not triggering into noise, strong pulses. |
| 1611 | Try reducing all thresholds from 50→40 ADC. |
| 1620 | Dumps are huge now, ~ 200 MB. But when you do see a pulse it's in the right place, so this may be a good threshold for the lowest threshold. |
| 1621 | Set DCRC1 thresholds 40→50 ADC |
| 1625 | Dumps are still large, and looking at BatViewer it seems as if DCRC3qI is the one that's triggering on noise (and maybe has the highest peak-to-peak noise). So try adjusting its threshold up |
| 1626 | Set all thresholds to 40 ADC except DCRC3qI→50 ADC |
| 1630 | Dump sizes back to normal, 10 MB. So these are probably good thresholds: 40 ADC except DCRC3qI = 50 ADC, with Trigger Baseline Length = 32 and Sample Length = 10 |
| 1640 | Stop the run, ground detector, to investigate LED functionality. |
| 1641 | We should have DIB LEDs on DIBs 3,4. DIB3 also has connections for the electrode LEDs. Switching amongst these requires flipping a few sets of switches on the DIB3 rerouting board. Try each LED individually and watch tower temp to determine if it is operational. |
| 1650 | Fill LN2 cold trap bath |
| 1702 | All DIB LEDs work. The electrode LEDs do not seem to respond to the DCRC. |
| 1710 | Use the multimeter diode test to measure all the LEDs on DIB3. Measure them from the DCRC side of the bias board with the appropriate set of switches enabled for each LED type. Only the UV LEDs show a measurable voltage on the multimeter, but all 3 types (DIB, electrode IR, electrode UV) heat the fridge when the diode test is performed. This seems to indicate all the LEDs are functional, but the DCRC is failing to drive the electrode LEDs. |
| 1716 | Fill LN2 cold trap full. |
| 1724 | Use flashandflash tool to start 12 hour bake using IR DIB LEDs on DIBS3,4. Tower temp looks like it will stay around 860 mK. |
| 1728 | Condenser = 119 mbar, Still = 0.18 mbar, MC(H35)/Tower/Coldplate = 216/883/418 mK |
| 2220 | Tower temp = 860 mK |