2017/10/30 (Monday)
| Setup: DIB 1/3/4 → 3.2 MHz filters 1/2/3 → cable 1/3/2 → gender changers → rerouting boards 1/3/4 → DCRC33(2)/DCRC55(1)/DCRC50(3) | |
| 0855 | Tower/MC/Coldplate = 82.8/59.3/364 mK, Still = 6 mW / 1.30 K / 0.20 mbar, G1/G2 = 299/289 mbar, 1K Pot = 1.652 K / 38 mbar. MB = 70.2% |
| 0900 | Fill LN2 cold trap bath. |
| 0915 | Try to flash for 60 seconds using the DIB LEDs on DIB3&4. Used the flash-and-flash tool, which applies the standard 0.6 mA. But for some reason there is no rise in temperature - it doesn't work. DCRCdriver, TowerFE and TriggerFE seem to be crashing after flash attempts. Try restarting DAQ entirely. Still not working. Enabling LEDs on DCRC3 seems to be the specific thing that crashes the programs. But turning them on either DCRC1 or 3 doesn't produce any temperature change. |
| 0935 | Power cycle the DCRCs, see if that helps. Yes, now it seems to work. |
| 0935 | Flash with DIB LEDs for 60 seconds using flashandflash tool. Tower reaches a little over 700 mK. |
| 0951 | Apply 10.01 V bias to detector |
| 0952 | Start series of data with Vbias(external) = 10.01 V, with 1000 beginning randoms, and a 30/25/25/40/40 ADC bin threshold trigger on DCRC1qO/DCRC2qI/qO/DCRC3qI/qO, with trigger baseline = 32 samples and trigger length = 10 samples (these are 2x the usual values). |
| 0958 | Looks like it's mostly triggering on noise. Increase all trigger thresholds by 10: 30/25/25/40/40 → 40/35/35/50/50 |
| 1001 | Triggering looks better now |
| 1010 | Dump sizes are falling. The 8th dump, after adjusting triggers, it was 88MB, then it fell until it was below 40MB and then fluctuated a bit. |
| 1010 | Stop run, dial Vbias down to 0 V, ground detector, another 60s flash with DIB LEDs. |
| 1037 | Apply 10.00 V bias to external electrode |
| 1037 | Start series of data with Vbias(external) = 10.01 V, with 1000 beginning randoms, and a 40/35/35/50/50 ADC bin threshold trigger on DCRC1qO/DCRC2qI/qO/DCRC3qI/qO, with trigger baseline = 32 samples and trigger length = 10 samples (these are 2x the usual values). |
| 1055 | 17 dumps so far but the dump size hasn't fallen. I think it's triggering too much on noise. |
| 1056 | Disable all triggers except qE (which is DIB4qO = DCRC3qO). Leave threshold at 50 ADC. |
| 1100 | Seems to be triggering constantly. Dumb. |
| 1105 | Try disabling all triggers |
| 1108 | ok I guess that worked, next dump is less than 1 MB |
| 1110 | reenable triggering on qE (which is DIB4qO = DCRC3qO) with threshold 80 ADC |
| 1115 | it's triggering constantly again |
| 1117 | Try enabling only qC = DIB1qO = DCRC2qO. ADC = 35. |
| 1120 | OK, this has decreased the dump size to 6.6 MB, previously was around 180 MB. BatViewer shows all events have pulses in DCRC2qO. |
| 1124 | Stop run, dial Vbias down to 0 V, ground detector, another 60s flash with DIB LEDs. |
| 1152 | Apply 10.00 V bias to external electrode |
| 1152 | Start series of data with Vbias(external) = 10.01 V, with 1000 beginning randoms, and a 35 ADC bin threshold trigger on DCRC2qO, with trigger baseline = 32 samples and trigger length = 10 samples (these are 2x the usual values). |
| 1155 | First few dumps are 40-44 MB. |
| 1215 | Dump sizes started falling at Dump 8. By Dump 13 they reached a minimum of 6-7 MB. |
| 1220 | Stop run, dial Vbias down to 0 V, ground detector, perform 6s flash with DIB LEDs. Tower peaks at about 360 mK. |
| 1233 | Apply 10.00 V bias to external electrode |
| 1233 | Start series of data with Vbias(external) = 10.01 V, with 1000 beginning randoms, and a 35 ADC bin threshold trigger on DCRC2qO, with trigger baseline = 32 samples and trigger length = 10 samples (these are 2x the usual values). |
| 1245 | Dumps started out around 42-44 MB. Start falling at Dump 7, reach min of 6-7 MB at Dump 13. This doesn't seem significantly different, implying that a 6-second flash is as good as a 60-second flash. |
| 1254 | Stop run, dial Vbias down to 0 V, ground detector, perform 1s flash with DIB LEDs. Tower peaks at about 255 mK. |
| 1325 | Apply 10.00 V bias to external electrode |
| 1325 | Start series of data with Vbias(external) = 10.01 V, with 1000 beginning randoms, and a 35 ADC bin threshold trigger on DCRC2qO, with trigger baseline = 32 samples and trigger length = 10 samples (these are 2x the usual values). |
| 1440 | Dumps started out around 40-43 MB. Start falling at Dump 7, reach min of 5-7 MB at Dump 13. This is also not much different. |
| 1445 | Stop run, dial Vbias down to 0 V, ground detector. |
| 1450 | Not sure we can think of an easy way to flash for less than 1s. So instead, try changing LEDPulseWidth on ODB from 100 us to 10 us. Then we'll try another 1s flash. |
| 1455 | We'll connect LED voltage to Phonon ADC and take data during the flash to see the LED signal. So, turn off DCRC2 trigger, enable ConnectLEDVoltagesToPhononADCs on DCRCs1&3, turn on DCRC1 pA trigger with threshold = 100 ADC, turn off initial randoms. |
| 1456 | Start series of data with detector grounded, with 0 beginning randoms, with LED voltages connected to pA and pB on DCRC1&3, and a 100 ADC bin threshold trigger on DCRC1pA. |
| 1456 | Use flashandflash tool to do 1s flash with DIB LEDs with LEDPulseWidth reduced from 100 to 10 us. Tower peaks at 240 mK. Update: this was actually 100 us pulses because flashandflash hard-codes this. |
| 1458 | Stop data. Look at the data during flashing with BatViewer. Don't see anything in the phonon channels except DCRC1pA, and it's not a small enough pulse to fit in the 128 us window. But we see pulses in all the connected charge channels. It looks like the pulse is 100 us wide in the charge channels, not 10 us. There were 403 events in the dump. LEDRepRate was at its standard value of 5000 us. So there should have been 1s/5000us = 200 LED pulses. So it seems like it was flashing for maybe 2 seconds. Could be that the sequencer script has an overhead and setting 1s of flashing actually gives you 2s. |
| 1510 | Restore ODB settings for normal data taking: set 1000 initial randoms, turn off DCRC1pA triggering, turn on DCRC2qO triggering. (Leave the LEDPulseWidths at 10 us and leave the LED voltages connected to phonon ADCs on DCRC1&3.) |
| 1532 | Apply 10.00 V bias to external electrode |
| 1532 | Start series of data with Vbias(external) = 10.00 V, with 1000 beginning randoms, and a 35 ADC bin threshold trigger on DCRC2qO, with trigger baseline = 32 samples and trigger length = 10 samples (these are 2x the usual values). |
| 1545 | Dumps started out around 39-42 MB. Start falling at Dump 7, reach min of 5-7 MB at Dump 12. This is also not much different. |
| 1558 | Stop run, dial Vbias down to 0 V, ground detector. |
| 1600 | As before we'll take data during flash: turn off DCRC2 trigger, leave enabled ConnectLEDVoltagesToPhononADCs on DCRCs1&3, turn on DCRC1 pA trigger with threshold = 100 ADC, turn off initial randoms. Also increase DCRC1 Phonon trace length from 128 to 512 (51 us → 205 us). |
| 1605 | Noticed that the LEDPulseWidths are back at 100 us on the ODB. Must be that the flashandflash script sets this. So don't use flashandflash. Instead set test.msl so that it simply turns on LEDs, waits 1s, then turns them off. |
| 1610 | Set LEDPulseWidths to 10 us on DCRC1&3. Leave LEDRepRate at standard 5000 us. |
| 1611 | Start series of data with detector grounded, with 0 beginning randoms, with LED voltages connected to pA and pB on DCRC1&3, and a 100 ADC bin threshold trigger on DCRC1pA. |
| 1611 | Use sequencer to do 1s flash with DIB LEDs with LEDPulseWidth reduced from 100 to 10 us. Tower peaks at 135 mK. |
| 1612 | Stop data. Look at the data during flashing with BatViewer. Unfortunately I failed to turn off initial randoms. But I do see pulses in the live charge channels and they do seem to be 10 us square waves, followed by an after pulse. |
| 1615 | Restore ODB settings for normal data taking: set 1000 initial randoms, turn off DCRC1pA triggering, turn on DCRC2qO triggering. |
| 1636 | Apply 10.00 V bias to external electrode |
| 1636 | Start series of data with Vbias(external) = 10.00 V, with 1000 beginning randoms, and a 35 ADC bin threshold trigger on DCRC2qO, with trigger baseline = 32 samples and trigger length = 10 samples (these are 2x the usual values). |
| 1645 | Dumps started out around 39-44 MB. Start falling at Dump 7, reach min of 5-7 MB at Dump 12. This is also not much different. |
| 1655 | Stop run, dial Vbias down to 0 V, ground detector. This time I won't flash at all - let's see whether just grounding the detector then resetting the bias is enough. |
| 1701 | Apply 10.00 V bias to external electrode |
| 1701 | Start series of data with Vbias(external) = 10.00 V, with 1000 beginning randoms, and a 35 ADC bin threshold trigger on DCRC2qO, with trigger baseline = 32 samples and trigger length = 10 samples (these are 2x the usual values). |
| 1705 | Dumps started out around 37-38 MB, suggesting at least partial neutralization. Start falling at Dump 4, reach min of 5-7 MB at Dump 9. So with no flashing at all, hold time decreases from usual 7-8 minutes down to 4 minutes. But clearly, simply grounding the detector then reapplying bias is enough to partially neutralize. (And note that it was not left grounded very long. If it were left for 20 minutes, typical wait time after flashing, would it be even better?) |
| 1718 | Stop run, dial Vbias down to 0 V, ground detector. |
| 1720 | Fill LN2 cold trap bath |