2021/03/02 (Tuesday)
- Lab in-person: Derek
- Virtual partner: Matt
| Status | Run 74 in progress. |
|---|---|
| Setup: DIB1/2+acc → SL 41/43 → 3.2 MHz filters → cables 2/3 → nothing/nothing → DCRC50(3)/55(1). No lead/poly shield. | |
| 1100 | Tower = 92.4 mK |
| 1135 | MB = 65.2% |
| Yesterday we disconnected the accessory boards from DCRC1 and plugged in the cable directly. We tried some data but it failed to process - it rejected all randoms from DCRC1, not sure why. But it may be related to very high noise in the channels. This is I think a new problem. | |
| 1150 | Measure noise with Pulse Viewer in the four channels, looking at 10kHz: DCRC3qI/qO/DCRC1qI/qO = 50/4/150/10 e-6. These are, in order, channels 3/1/4/2. So there is a clear pattern: noise is very bad in the center channel, q4, and gets better as you go out toward the outer rim. |
| Try zapping SQUIDs to see if that affects the noise. No improvement. Probably the best bet is to powercycle DCRC1. | |
| 1200 | Apply +5V bias. Noise does not get worse - perhaps slightly better in the two worst channels, q4 and q3. Return bias to 0V. |
| I didn't do the same measurements before we powercycled DCRC1, but looking at noise traces in BatViewer suggests the q4 noise is 10X what it was before, and q3 is about 80% bigger, and the other two aren't that much worse. So it's suggestive that we have a big noise source in q4 (DCRC1qI) and it's bleeding over into q3 a bit. | |
| 1645 | Tower/MC/Coldplate = 92.7/31.1/185 mK, Still = 1.42 K / 0.130 mbar / 0 mW, G1/G2= 211/203 mbar, 1K Pot = 1.67 K / 36±2 mbar. |
| 1645 | MB = 57.3% |
| 1646 | Fill LN2 cold trap bath. |
| 1700 | While Derek was in the lab, we were planning to troubleshoot the DCRC noise. But as soon as we looked at it, the noise was much lower than before - it just fixed itself! Looking at 10kHz on the Pulse Viewer again, DCRC3qI/qO/DCRC1qI/qO = 1.3/1.0/3.2/1.1 e-6. This is I think very close to what we saw before. Note that again noise is highest in q4 and gets smaller as you go out - pretty sure we were seeing this before. |
| OK let's try some data. | |
| 1704 | Start 31min series of +7V data (bias applied after 1 min), with 1000 beginning randoms, trigger threshold 80/80/80/80 on DCRC3qI/qO/DCRC1qI/qO, no in-run randoms, with 1.6384 ms pulses (8192 length) and 400 us prepulse (500 length). Use sequence r74_takeandflash_dcrcbias.msl |
| Triggering looks fine. | |
| Turns out this data looks nearly identical to the data with the diverter boards present. No 60keV line visible, muon blob in the same place. |