previous day
| | Overnight status: DCRC50/55 → no filter/10MHz filter → HV slot 1/2, zero bias, baking detector with the two available LEDs in HV setup. Tower = 210 K |
| 0915 | MC/Tower/Cold plate = 93/210/166 mK, Still = 1.11 K / 0.13 mbar, G1/G2 = 180 mbar, 1K Pot = 1.72 K / 22 mbar, Main bath = 6.7% |
| note: | boiloff rate unusually high since the transfer two days ago, about 1.9%/hour (usually it's less than 1.7%/hour) |
| 0932 | Begin LHe transfer at main bath = 6.1% |
| 1010 | Cease LHe transfer (93.9%). |
| 1100 | Stop baking detector. After 40 minutes tower cools below 33 mK. |
| 1230 | Take IbIs data in current setup. |
| note: | the IbIs curve for HV2pB (the problem channel) implies that the QET bias was set much too high on Friday. |
| IbIs results |
| QET biases chosen for R_bias of 200 mΩ |
| Channel | Electronics | QET bias (μA) | R_normal (Ω) |
| HV1 pA | DCRC50 | 93.1 | 0.613 |
| HV1 pB | DCRC50 | 114.1 | 0.646 |
| HV1 pC | DCRC50 | 88.5 | 0.710 |
| HV2 pB | 10MHz → DCRC55 | 68.8 | 0.743 |
| HV2 pC | 10MHz → DCRC55 | 110.9* | 1.043 |
| HV2 pD | 10MHz → DCRC55 | 88.5 | 0.858 |
| *: this channel had at be set to R_bias of 300 mΩ |
| 1345 | Strange very regular pattern of spikes in the charge channels, one positive followed by one negative, ~80 Hz full cycle, +/- 130/30 μV in Qi/Qo, not pulse-like, more like Gaussians with about 50 μS width. Temperature elevated about 2 mK from earlier base, maybe a Bad State. Power cycle both DCRCs - pattern disappears. |
| 1455 | 6 kHz noise seen in HV2pB. Got a bit better after turning on the HV power supply. |
| 1510 | Try taking HV to near maximum: -15 V, then increasing Side1 bias. At Side1 = +10 V, phonon pulses (on HV2) disappear but come back - rate is low so it is probably only the largest pulses. At Side1 = +14 V phonon pulses disappear but can be brought back by reducing the QET bias. We also found that the HV (crystal bias) can be increased a small amount over the maximum if the QET biases are correspondingly lowered. |
| 1530 | Measure current from HV by opening HV box top and measuring V across 1 MΩ test resistor with multimeter: |
| Current draw from HV supply |
| HV (V) | Current (nA) |
| -1.1 | -1 |
| -1.8 | -2 |
| -3.9 | -5 |
| -6.9 | -10 |
| -9.9 | -15 |
| -12.9 | -20 |
| -16.9 | -25 |
| -19 | -30 |
| -21.2 | -35 |
| -23.9 | -40 |
| -26.8 | -45 |
If was assume an offset of -0.4 V in the HV supply gauge, this is consistent with a 600-700 MΩ total resistance seen by the HV. Since the bias resistor is 400 MΩ this implies a remaining 200-300 MΩ through which current is leaking. This could be in the HV board (although we measure more than a GΩ when it is disconnected), or the cables, or the detector (though the detector should be nearly insulating). The disappearance of phonon pulses implies some heating effect at the detector. Also at higher HV settings a tower temperature increase is seen:
| 1600 | Monitor temperature while increasing HV: |
| Tower temperature versus HV |
| HV (V) | Temperature (mK) |
| -17 and smaller | 32.3 |
| -20 | 32.7 |
| -22 | 33.1 |
| -24 | 33.4 |
| -26 | 33.7 |
| -28 | 34.5 |
| -30 | 34.9 |
Using the cooling power curve from Run 53 and the current base we can estimate the added power as a function of tower temperature: P[μW] = (T[mK]/12.4)^2-6.79. Knowing the current and the heating power we can calculate the implied detector resistance to be around 200-400 MΩ (although it is not constant).
| 1640 | Testing leakage current with Dsub cables disconnected from ebox. |
Set bias voltage based on power supply meter. Measured voltage across 1MOhm resistor on HV board with multimeter to find leakage current.
| Leakage current vs. time 10V bias |
| time | I_leakage (nA) |
| 0 | 7 |
| 15sec | 3 |
| 30sec | 2.7 |
| 1min | 2.4 |
| 3min | 2.1 |
| 5min | 1.5 |
| 10min | 1 |
| Leakage current vs. time 20V bias |
| time | I_leakage (nA) |
| 0 | 20 |
| 30sec | 5 |
| 1min | 3 |
| 1min 30 sec | 3.5(?) |
| 2min | 3 |
| 5min | 2.7 |
| 1748 | Main bath = 84.1% |
| note: | this is a normal boiloff rate for this level, 1.28%/hour |
| 1753 | Filled LN2 cold trap full |
next day