2015/07/01 (Wednesday)
| Overnight status: baking detector, Cf source on fridge at detector level | |
| 0813 | MC/Tower/Cold plate = 109/297/169 mK, Still = 2 mW / 1.13 K / 0.13 mbar, G1/G2 = 240/225 mbar, 1K Pot = 1.69K / 15 mbar, Main bath = 27.3% |
| 0815 | Fill LN2 cold trap Dewar |
| 0820 | Remove Cf source from fridge |
| 0825 | Stop flashing. Used the vi to do this, then checked rd 30 on the PuTTY sessions: all three read 0800 as they should. |
| 0921 | Investigating DIB3pC (DIB3/DCRC50 with the 3.2 MHz pi filter). As yesterday it shows no pulses except very large saturated pulses. But increasing the QET bias from 55 μA to 80 μA seems to restore it to normal behavior. |
| 0926 | Same thing with DIB3pD: increasing QET bias from 89 to 99 μA restores it to normal behavior. |
| 0939 | DIB3pA has ~50 kHz noise, requiring relatively high threshold, 70, to avoid noise triggers. |
| 0957 | Since QET behavior seems to have changed, run IbIs on DIB3→3.2 MHz filter→DCRC50 for pC and pD. Set Frontend Gain/Output Gain/Driver Offset to 1/0/0 on each. |
| 1108 | Analyzed IbIs curves to get optimal QET biases (R = Rnormal/4) for DIB3pC and DIB3pD: |
| Optimal QET biases for DIB3→3.2 MHz pi filter→DCRC50 | ||
|---|---|---|
| pC | pD | |
| DIB3 | 80 μA | 106 μA |
| Current Frontend Gain / Output Gain / Driver Offset / QET Bias | ||||
|---|---|---|---|---|
| pA | pB | pC | pD | |
| DIB3 | 7/1/-0.056/73 | 7/1/-0.092/85 | 7/1/-0.030/80 | 7/1/-0.072/106 |
| 1122 | Adjust driver offsets and reset thresholds to lowest values avoiding noise triggers: |
|---|
| Adjusted driver offsets (V) | ||
|---|---|---|
| pB | pD | |
| DIB1 | -0.082 | -0.090 |
| Current phonon thresholds | ||||
|---|---|---|---|---|
| pA | pB | pC | pD | |
| DIB1 | 25 | 25 | 35 | 20 |
| DIB2 | 25 | 20 | 25 | 30 |
| DIB3 | 80 | 25 | 20 | 30 |
| 1134 | Oscillation on DIB2 phonon channels, ~150 Hz 10 mVpp. Seen on DIB1 and DIB3 also with smaller amplitude, ~7.5 mVpp. |
|---|---|
| 1154 | Flash 30 seconds |
| 1300 | Set -4/+4 V, start bkg data |
| 1317 | Stop data, set 0/0 V, flash 30 seconds |
| 1332 | DCRCs enter a Bad State: temperature rises from 32 mK base to about 36 mK, 250 Hz oscillation on DIB3 phonons, strongest on pA at 30 mVpp, also seen on DIB2 at 15 mVpp, DIB1 noise around 25 mV pp. |
| 1354 | Unplug in order DCRC50, DCRC55, DCRC33. Noise still present after the first two disconnects. Repower all boards. |
| 1405 | Reset DCRCs and zap SQUIDs. Noise still present. Bad State persists. |
| 1441 | Unplug all three DCRCs and wait. Initial temperature spike to 47 mK, then 5 minutes later another increase, peaking at 58 mK at the 11 minute mark. Clearly it takes some time after unplugging to escape the Bad State. Took about 30 minutes for the temperature oscillations to stop and the tower to reach its 31 mK base. |
| 1548 | Repower DCRCs. Zap SQUIDs. (Note: initially had some communication problems - from some past experience I think it is important to unplug the timing jumpers before repowering the DCRCs. |
| 1505 | Commence LHe transfer (14.7%). |
| 1600 | End LHe transfer (92.9%). |
| 1646 | Take IbIs data for DIB3 → 3.2MHz pi filter → DCRC50 with Frontend Gain = 1, Output Gain = 0(setting), Driver Offset = 0 |
| 1707 | Flash overnight. Put Cf source in shield at detector level. |
Lesson from today:
- When DCRCs are in a Bad State, it takes time after unplugging them for things to settle down (as evidenced by temperature fluctuations), as long as 30 minutes.