2017/01/25 (Wednesday)
| Status: Circulation began yesterday. | |
| 0916 | Tower/MC/Coldplate = 60.3/34.3/202 mK, Still = 1.3 K / 0.110 mbar, G1/G2 = 122/128 mbar, 1K Pot = 1.71 K / 34 mbar, Main bath = 31.3%. boiloff is a little faster than slow-mode |
| 0926 | Fill LN2 cold trap. |
Below are SQUID settings from the last run, on Dec 5, which was the last time we ran in full readout mode. We'll start with these.
| channel | SQUID bias | Lockpoint |
|---|---|---|
| DCRC1 pA | -76.5 | -3.1 |
| DCRC1 pB | -81.9 | -3.4 |
| DCRC1 pC | -70 | -5.7 |
| DCRC1 pD | -82.6 | -1.8 |
| DCRC2 pA | 50 | 1.67 |
| DCRC2 pB | 64.9 | 2 |
| DCRC2 pC | 60.3 | 3.07 |
| DCRC2 pD | 60.6 | 1.5 |
| DCRC3 pA | -84.8 | -4.68 |
| DCRC3 pB | -86.7 | -3.06 |
| DCRC3 pC | -48 | -0.9 |
| DCRC3 pD | -51.3 | -4.7 |
| 0950 | Start MIDAS web interface, no programs running except mhttpd, logger, lazy_0, mserver. Set ODB with above values, enable DCRC 1,2,3, set all QET biases to 300/-300, set up DCRC1 as timing master. |
|---|---|
| 0955 | Plug DIB cables into DCRCs (DIB 1/3/4 → DCRC1/2/3). Power DCRCs. Tower temperature spikes to 250 mK than falls again. Plug in timing cables with DCRC1 as master. |
| 1005 | Start DCRCdriver, TowerFE, TriggerFE |
| 1007 | Zap SQUIDs. |
| 1028 | Didn't see any temperature excursions on last Zap. Zap again to be sure. |
| Still nothing, but it looks like dcrc_driver01 and towerfe3_01 had crashed. | |
| 1049 | Restarting dcrc_driver01 and towerfe3_01 seems to work at first, but they crash again of their own accord after a few minutes. MIDAS gives the following error messages: |
<blockquote>10:48:06.516 2017/01/25 [mserver,INFO] Client 'dcrc_driver01' (PID 254094) on database 'ODB' removed by cm_watchdog (idle 60.3s,TO 60s)
10:48:03.511 2017/01/25 [mserver,INFO] Client 'towerfe3_01' (PID 254096) on database 'ODB' removed by cm_watchdog (idle 60.3s,TO 60s)</blockquote>
| 1102 | Restarted mhttpd etc. with kill_daq_umn.sh script. |
|---|---|
| 1105 | Start dcrc_driver01. Starting towerfe3_01 crashes dcrc_driver01 immediately and vice versa. |
| 1108 | Unpower DCRC4. Can start dcrc_driver01 and towerfe3_01, but starting triggerfe2 crashes dcrc_driver01. |
| 1111 | Disabling all DCRCs allows all 3 programs to run without crashing each other. |
| 1114 | Enabled the DCRCs one at a time, each time ensuring that all 3 programs can start. Now I all 3 programs can start with all 3 DCRCs enabled. 10mK tower temp rise seen. |
| 1115 | dcrc_driver01 and towerfe3_01 both crash with same error message as above. |
| 1118 | Can't connect to DCRC3(50) via telnet session. Power cycle it, telnet connectivity restored. |
| 1119 | Restart programs in the order dcrc_driver01, towerfe3_01, triggerfe2. That seems to have solved it! |
Lesson learned: if a DCRC is not connected or is not communicating, then MIDAS will crash if that DCRC is enabled in the ODB.
| 1123 | Zap Squids. |
|---|---|
| 1215 | First pulses seen. Decent-looking pulses on DCRC1pC at unusually low QET bias: 30 μA Tower = 66.1 mK, dropping slowly. |
| 1304 | Changing QET biases to get pulses on all channels. DCRC1 PB,PC (both side2) have strong pulses at ~50 μA. Any lower and they seem to fall into SC after a while. DCRC1PA has smaller pulses at 50μA. DCRC1 PD has strong 60Hz noise. Relocking does nothing. |
| 1420 | Start LHe transfer @ MB = 22.7% |
| 1440 | DCRC1pD looks fine now, 60 Hz noise just disappeared. |
| 1505 | Estimate event rate by counting pulses in 20 50-ms traces in DCRC1pD. Count 65 pulses in 1 second, so event rate is 65 Hz. |
| 1515 | The DCRC1 channels with 50 uA QET biases (pA, pB, pC) fell into superconducting. Setting all DCRC1 QET biases to 75 uA puts them in transition, all show pulses. |
| 1517 | Cease LHe transfer (92.6%) |
| 1530 | Preparing to measure Tc's on DCRC1 channels. Set all QET biases to 0 and apply 12 mV 400 Hz test signal to QET lines. |
| 1535 | Tower = 65.4 mK |
| 1600 | Applied MC heat to get tower up to 200 mK, but saw no transition. Did we actually have the channels in superconducting?? |
| While watching the triangle wave forms we noticed that the output signals were all about 14 mVpp, with driver gain = 2x. This implies TES resistances around 80 mΩ, i.e. normal resistance, not s/c. Huh. | |
| 1630 | While tower was warm, no pulses could be seen in any of the DCRC1 channels. As it cooled below 70 mK again, pulses could be seen at similar QET biases as before. As tower continues to cool, QET biases keep having to be raised to keep the channels out of s/c and showing pulses. |
| 1635 | Tower = 66.4 mK |
| During the transfer, the 1KPot fell from 1.71 K / 35 mbar to 1.65 K / 24 mbar. Should tweak it open more. | |
| 1640 | Tweak 1K Pot needle valve open a little more. 1K Pot → 1.70 K / 40 mbar. |
| 1645 | Fill LN2 cold trap bath |
| 1650 | Tower = 65.0 mK |