2022/02/02 (Wednesday)
- Lab in-personnel: Matt
| Status | Run 75 in progress |
|---|---|
| Setup | Tower Faces 5/3/2 → Cables 2/1/3 → nothing/diverter_board+extra_board/nothing → HVboard CONN1/2/3 → HV1/2 → DCRC55(1)/50(3) Lead shield in place. Am Source y = 0.03 mm |
| 1350 | Goldman/Tower/MC(H35)/Coldplate = 40.8/71.7/33.8/183 mK, Still = 1.40 K / 0.110 mbar / 0 mW, G1/G2= 265/228 mbar, 1K Pot = 1.59 K / 11±1 mbar, MB = 67.2%. |
| 1355 | Fill LN2 cold trap bath. |
| 1530 | Working on remote-control HV setup, referring to remote_hv instructions. |
- Original setup: Kepco 1KV → variac → cheater cord → orange wall outlet
- Connect HV output to HP meter, disconnect from HV board
- As per instructions, change to Kepco 1KV → variac → relayed power strip → relay board → cheater → orange wall outlet
- Powered and enabled DCRC2. Tried turning on relay by changing Charge Bias [1] to 2. No effect - relay didn't work
- Switched back to non-relayed power
- Dial Kepco output to 300V
- Step up DCRC2 Charge Bias [0] in steps of 0.5 V. HV output goes down as it should. Hits minimum of around 50 mV with QI = 11.64 V
- but not entirely stable - has shifted suddenly up to 1 V a couple times, for several seconds
| 1540 | Plug HV box power into HV supply. Watching DCRC1pA - no increase in noise. |
|---|---|
| 1600 | Use Root Sequencer to step DCRC1qI from 11.64 down to 7.76 over a total time of 20 seconds. Takes the bias up to +98 V. Some increase in noise seen. |
| 1605 | Goldman = 41.2 mK. Set QET biases to target dP/dI=1.0 uV and adjust driver offsets so baselines ~ -0.9 V |
| 1609 | Start +98.15 V data series, 25 min long, with 1000 initial randoms, no in-run randoms, trigger on DCRC3pC threshold 30. Start series with the Start Button. |
| 1630 | Goldman = 41.0 mK. Adjust QET biases to target dP/dI=1.0 uV |
| 1635 | After ending data sequence, use Root sequencer again to adjust HV down to ~0V over 20 seconds. Takes a little time to relax to the minimum - drops from 150 mV to 60 mV over about 2 minutes. |
| 1640 | Disconnect HV box from HV supply. Let's try simulating a DAQ crash - turn off the DAQ programs, then turn them back on, see what happens to the HV output. |
| 1645 | Turn off DAQ programs including DCRC Driver. No change in HV output. Turn off DAQ programs. Then adjust ODB setting for DCRC2qI from 11.64 to 11.0. Then turn on DAQ programs. When DCRC Driver is turned on, HV changes to 14.8 V. Reset to 11.64, HV goes back to 55 mV. |
| So: it is probably safe to leave HV connected even without the relay, which is designed to automatically shut off the HV if the DCRC output dies somehow. But a DAQ crash will leave the DCRC in whatever state it was in. And restarting the DAQ should be OK too - but this is more dangerous because sometimes ODB settings change when you don't expect them to (like for instance the DCRC version of DCRC2 will change from 3.0 to 3.1 sometimes which totally screws things up). So if the DAQ does crash it may make sense to wait till you're in the lab to disconnect the HV box first. Or, if it crashes while the detector is still biased, then it may make sense to physically dial down the HV and start over. Also - there were certainly times in the past where an error resulted in a sudden HV change, and it doesn't automatically destroy anything. But we try to avoid it. |
Analysis of 98V data:
- It's a bit hard to see the Am lines with partitions. PA/PT actually worked best. Lines look distorted - pD is anomalously low.
- Empirical Luke gain = 19.1, expected 26.8 (maybe some loss from pD problems…). Pretty sure I identified 14 keV correctly.
- noise width 19.2, with no attempts to clean things up. Slightly better than 20V.