Previous Day
| | Setup: DIB1/3/4 → 3.2 MHz filters → cables → DCRC55(1)/33(2)/50(3)
taking 0-bias data overnight, no flashing |
| 0905 | Tower/MC/Coldplate = 34.7/36.3/222 mK, Still = 1.23 K / 65 μbar, G1/G2 = 125/110 mbar, 1K Pot = 1.59 K / 6.5 mbar. MB = 34.5%
Boiloff rate: almost as slow as slow-mode |
| 0910 | Fill LN2 cold trap bath |
| 0910 | Tweak 1K Pot needle valve open a few more degrees. |
| 0930 | After LN2 fill and 1K Pot tweak, base is lower and pressures are higher: Tower = 33.6 mK, Still = 80 mbar, G1/G2 = 145/125 mbar, 1K Pot = 1.70 K / 30 mbar |
| 1040 | Start mini-IbIs of all active channels to get Rnormal and Rbiased for the TESs. |
| Channel | TES | QET bias [μA] | Rnormal [mΩ] | Rbiased [mΩ] | Rbiased/Rnormal |
| DIB1 pA | 4 | 200 | 89.9 | 80.3 | 89.2% |
| DIB1 pB | 3 | 210 | 91.7 | 86.7 | 94.5% |
| DIB1 pC | 2 | 200 | 67.8 | 66.5 | 98.1% |
| DIB1 pD | 9 | 220 | 95.3 | 92.6 | 97.2% |
| | |
| DIB3 pB | 11 | -200 | 95.9 | 99.6 | 103.9% |
| DIB3 pC | 5 | -194 | 99.8 | 90.1 | 90.2% |
| | |
| DIB4 pB | 8 | 225 | 101.7 | 96.1 | 94.4% |
| DIB4 pC | 1 | 190 | 96.5 | 87.7 | 90.9% |
| DIB4 pD | 0 | 199 | 80.4 | 39.6 | 49.3% |
DIB4 pD has been biased at a much lower point in the transition than the other channels. Yesterday this channel's QET bias was adjusted a bit lower than it had been, below the region of oscillations.
In last night's data, DIB4 pD's pulses were much higher than the others in all events. There are two possible reasons for this. This was the channel we were triggering on and the threshold seems to have been set too high, so that only pulses well above the noise were triggering. So the first possible reason is that maybe the data was biased to events right underneath this channel. The second possible reason is that it is set to a much more sensitive point in the transition via the QET bias.
To test the first possibility the data below was taken with a different channel as the trigger, and with a much lower threshold. This data seems to have much stronger pulses in the other channels but DIB4 pD is still generally the highest. So probably both of the reasons discussed above are happening to some extent.
| 1341 | Start data series of 0V data with Frontend Gain = 5/2/5, OG = 8/3/8, with 1000 beginning randoms, with no flashing in between series, and a 20ADC bin threshold trigger on DIB1 pB. Stop manually after about 20 minutes. |
| | QET bias [μA] |
| Channel | TES | Selected | Highest with pulses | Phase inversion | Highest with oscillations | Highest superconducting |
| DIB1 pA | 4 | 200 | 230 | 205 | 195 | 175 |
| DIB1 pB | 3 | 210 | 240 | 210 | 190 | 180 |
| DIB1 pC | 2 | 200 | 230 | 200 | 190 | 170 |
| DIB1 pD | 9 | 220 | | | | |
| | |
| DIB3 pB | 11 | -200 | -200 | | -180 | |
| DIB3 pC | 5 | -194 | -220 | | -180 | |
| | |
| DIB4 pB | 8 | 225 | 250 | 225 | 215 | 200 |
| DIB4 pC | 1 | 190 | 220 | 190 | 180 | 150 |
| DIB4 pD | 0 | 199 | | | 230 | 202 |
The table above shows where the biases have been chosen with respect to the transition region. Note that DIB4 pD has been biased much closer to the superconducting point than the others.
| 1452 | Commence LHe transfer (24.5%) |
| 1543 | End LHe transfer (90.9%). |
| 1556 | Release pressure from main bath. |
| 1624 | Release pressure from main bath. |
| | Seeking new QET bias values using the following procedure: get TES into transition, step down QET bias 1 μA at a time until it falls into superconducting. Then set the QET bias 5 μA higher than this. |
| 1701 | Finish finding the lowest possible QET biases for each channel. QET Biases listed below: |
| Channel | TES | QET Bias [μA] |
| DIB1 pA | 4 | 179 |
| DIB1 pB | 3 | 182 |
| DIB1 pC | 2 | 184 |
| DIB1 pD | 9 | 191 |
| | | |
| DIB3 pB | 11 | -176 |
| DIB3 pC | 5 | -175 |
| | | |
| DIB4 pB | 8 | 207 |
| DIB4 pC | 1 | 165 |
| DIB4 pD | 0 | 202 |
| 1705 | Driver Offsets of all channels, including disconnected channels, set so that the baselines are ~-0.5V. |
| 1707 | Start 15 one-hour data series of 0V data with Frontend Gain = 5/2/5, OG = 8/3/8, with 1000 beginning and ending randoms (no end randoms), with no flashing in between series, and a 20ADC bin threshold trigger on DIB1 pB. |
| 1710 | Fill LN2 cold trap full. |
| 1750 | Stop sequence because trigger threshold is too low. |
| 1751 | Increase DIB1 pB threshold to 50 ADC bins and start 2 minute test series |
| 1752 | Stop test series |
| 1755 | Data looks like all pulses, so decrease DIB1 pB threshold to 40 ADC bins and start 2 minute test series |
| 1757 | Stop test series |
| 1759 | Looks good. Start 15 one-hour data series of 0V data with Frontend Gain = 5/2/5, OG = 8/3/8, with 1000 beginning and ending randoms (no end randoms), with no flashing in between series, and a 40ADC bin threshold trigger on DIB1 pB. |
| | Unfortunately it looks like DIB4pD has been railed since the beginning of this data. It was OK during the 17:55 test series. Lame. |
Next Day