2019/11/11 (Monday)

Previous Day

Setup	DIB1/2/4 → 3.2 MHz filters → cable 1/2/3 → gender changer 1/2/3 → movable source controller/board with external Feedback resistor/nothing → DCRC55(1)/DCRC33(2)/DCRC50(3)
0910	Tower/MC/Coldplate = 38.4/30.8/193 mK, Still = 1.51 K/ 0.130 mbar / 0 mW, G1/G2= 162/134 mbar, 1K Pot = 1.67 K / 32±2 mbar. MB= 30.8%
0915	Fill LN2 cold trap bath.
1215	Begin LHe transfer @ MB = 26.2%
1300	End LHe transfer @ MB = 92.1%
1350	Trying to understand fridge heating during bias yesterday. If the pattern holds, we are currently in a slightly elevated temp state following a -4/+4 V, 0V dataset. Set bias to +4V/0 V in the ODB. Tower falls from 38.2 → 37.0 mK. Set bias to +4V/-4V in the ODB, tower temp stays flat.
	the pattern here of breakdown is pretty mysterious. It seems that +4/-4 V is safe, something bad happens with -4/+4 V, although not immediately.

Nick has found that the overall phonon collection efficiency at 60 keV has gone down every time we move the source - it is now down by a factor of 3 compared to before we started moving the source! Hypothesis is that each heating/cooling cycle from the motor is trapping magnetic flux in the Al fins. If so we should be able to heat then cool the detector (without moving the source) and return it to its full PCE. Can think of two ways to do this:

Stop fridge circulation and 1K Pot, using the procedure followed in Run57(? I think) that was used when testing the effect of removing the mu-metal shield.
Try increasing duty cycle of LEDs and performing a “hot bake”, maybe can get it hot enough that way. That would be preferable because it can be controlled by the sequencer. (Or should we call it a “hot flash” - we don't need to do it a long time, just get it above Al Tc=1.2 K. And “hot flash” is a funnier name.) (Or “overflash”, or “high T flash”, or “aluminum flash”.)

1622	Default LED settings are PulseWidth/RepRate/Current = 100 us / 5000 us / 0.6 mA. When baking like this, the tower reaches 115 mK. Let's try cutting the RepRate (really repetition period) down. Set LEDRepRate = 500 us on all 3 DCRCs and enable them, also enable 15V power on DCRC3 and ground detector.
1626	Tower reaches ~300 mK
1630	Turning off DCRC3 LED drops the temp a bit, but turning off DCRC2 LEDs does not. Try setting DCRC2 LED Currents to -0.6V. Now the tower rises to 570 mK. Looks like we had the polarity wrong on the RevC. Should fix existing sequences to account for this! (Later: Done for all existing sequences and modified flashandflash.py which generates the sequences for DCRC_UI. )
1635	Increase LED PulseWidth to 200 us (40% duty cycle), tower rises to 820 mK.
1638	Increase LED PulseWidth to 400 us (80% duty cycle), tower rises to 1.15 K.
1638	Increase LED PulseWidth to 500 us (100% duty cycle), tower temp starts to fall, this might be a bad setting. Return LED PulseWidth to 400 us (80% duty cycle), tower rises to 1.15 K.
1644	Set LED current = 0.8/-0.8/0.8 uA on DIB 1/2/3. Tower temp rises to 1.6±0.1 K and is somewhat unstable. MC = 1.0 K
1649	Turn off LEDs. Restore settings to DCRC 1/2/3: PulseWidth/RepRate/Current = (100/100/100) us / (5000/5000/5000) us / (0.6/-0.6/0.6) mA and disable DCRC 3 15V power.
1652	Tower = 840 mK and falling.
1655	G1/G2/P1 peak at 327/286/0.3 mbar. They had been rising steadily from 182/152/0.13 mbar when the LED high T baking first began.
1657	Tower = 326 mK
1659	Fill LN2 cold trap full.
1700	Tower = 164 mK
1710	Tower = 59 mK
1717	Tower reaches 45 mK and flattens out abruptly.
1725	MC = 33.9 mK
1727	Tower = 43.7 mK and cooling again but at a slower rate.
1818	Tower = 37.7 mK, phonon channels all showing pulses and baselines are about right.
1823	Start 6x S1/S2=+4/-4 V loop of 30 min Bkg data series with edited phonon settings, Driver gain = 1, with 1000 initial randoms, with 30s flashes and 30 min cooldowns trigger on DCRC2pA/B/C/D = 8/8/8/8 ADC using sequence take_and_flash_loop.msl (with negative DCRC2 LED bias)

Next Day