2021/04/07 (Wednesday)
- Lab in-person: Matt
- Virtual partner: Derek, Zach
| Status | Run 74 in progress. Warmed the fridge to fix short and thermal touch. LN2 transfer yesterday. The DIB2qI short reappeared at around 130 K, so we need to warm up again. |
|---|
More on the DIB2qI Feedback-to-Chassis short:
- This short is not present at room temperature but shows up at LN2 temperature. Early in this run it was a weak short but it became a strong short later. Bumping the fridge caused the short to momentarily disappear.
- I had guessed that it was a problem at the SQUET. This SQUET has the interposer, which could cause a short at the flyover cable. At room temperature I could not reproduce the short by moving things physically. I tried improving the insulation - this obviously didn't work.
- I now suspect the Side Coax, ST-3-10. This was used in Run 71, on DIB2 of the iZIP G106a, so it was used with charge channels. And similar shorts were seen appearing at 77K. We saw these shorts after two separate cooldowns. (And a different SQUET was used with that DIB than in this run.)
 | The short is presumed to be a touch from the FET gate line to chassis. This would explain the 25 nS seen between Feedback and Chassis - it's read through the 40 MΩ feedback resistor. Drain and Source showed no shorts - but this will depend on polarity. We should measure Chassis→Drain/Source too, this will probably show a short larger than 20 nS. |
- A touch from a wirebond from either the coupling capacitor or the feedback resistor, on the gate wire side, could explain the short. Some side coax photos here from Run 67. Photos from the Run 71 installation and the current run insulation show copper foil covers on the ends of the side coaxes - could be that after thermal contraction this touches a wirebond.
- But: in Run 67 in the first installation we saw a similar short using ST-3-10, and at that time there was no foil cover. The short showed up again at 77K after a second installation. In the final installation we used a different side coax, ST-3-8.
- For the current run I chose ST-3-10 over ST-3-8, because ST-3-8 showed an open bias resistor. Probably should try to fix this and maybe use ST-3-8 in this run.
| 1330 | Matt enters lab |
|---|---|
| 1330 | MC(Cernox) = 77.2 K, Sorb = 134.2 K, G2/G1 (fridge+trap1+dump) = 569/608 mbar, 3He pump off |
| 1330 | Turbo on Cryostat Vacuum Jacket = 3 nbar. With Turbo valved out for 1 minute, Cryostat Vacuum Jacket = 54 nbar. |
| 1340 | Check the DIB2qI FB → Chassis short : 0.22 nS. It disappeared! Probably should push forward then and do LHe today… |
| 1345 | Stop pumping out Cryostat Vacuum Jacket, switch pump manifold back to IVC port and pump out line. |
| 1350 | Check IVC pressure: 1.2 mbar. Start pumping out IVC with Scroll Pump. At <0.1 mbar switch to Turbo Pump. |
| 1350 | Turn on 3He pump and open V6 to 12.7%, reestablishing circulation like yesterday. G2/G1/P1 → 694/743/0.60 mbar and settle. |
| 1355 | Close V12a speedivalve for 1 minute to assess throughput. G1 falls at 70.9 mbar/min, good. Reopen V12a speedivalve. |
| 1400 | Pump mixture from fridge and Trap 1 to Dump |
| Throughput test procedure, Cold Trap 1: |
- All mixture in dump, LN2 trap 1 cold, condenser & still valves open (3He pump on - optional)
- Open V9, V13a, V12a+speedivalve, V3. Wait for G1 to be close to G2.
- Close V12a+speedivalve and V3, open V1
- monitor G1
| 1405 | Perform mixture throughput test at MC = 77 K using above procedure with 3He pump on. Let G1 rise to 555 mbar before Step 3, then it falls to 70 mbar, then rises. Throughput = 47.3 mbar/min. A little higher than last time, 2021 Feb 17 |
|---|---|
| 1405 | Open mixture to both sides of the fridge and cold LN2 Trap 1. |
| 1425 | Do electrical checkout on DIB2+accessories, more thoroughly checking the charge stuff to chassis, and both polarities for Drain and Source. Everything is below around 0.3 nS (feedback, drain, source; bias is around 0.5 nS.) qI values for Feedback→Chassis and for Chassis→Drain&Source are slightly higher than the qO cases, but it's only like 0.05 nS higher. Everything looks good. |
| 1435 | Do electrical checkout on DIB1. Feedback/Drain/Source to Chassis and back are a little lower than DIB2, all less than 0.2 nS. |
| 1435 | Try Leak Detector on IVC, with Turbo still pumping in parallel. (We've only been pumping out the helium for 45 minutes.) LD = 530 pbarL/s. With Turbo valved out, LD → 3000 pbarL/s. Open Turbo valve to continue pumping IVC. |
| 1440 | DIB2qI short shows up again!! Feedback→Chassis fluctuating around 10 nS. Chassis→Source around 25 nS. Chassis→Drain around 50 nS. They are all fluctuating a lot. |
| OK screw it then - this connection is too unstable. We'll warm up and swap out that side coax. [image not archived: doubt2.svg] | |
| 1455 | Begin blowing out LN2. Main bath (G3) → 680 mbar+1atm. |
| 1500 | Currently DIB2qI Chassis→Drain is reading around 12 MΩ |
| 1506 | Main bath (G3) = 300 mbar+1atm and falling, looks like blowout is finished. Closing V1a causes G3 to rise, so yeah it's finished. |
| 1510 | Stop LN2 blowout |
| 1515 | Fill LN2 cold trap bath |
| 1515 | remove hose from MB LN2 fill pipe and put stopper in pipe, for MB pumpout |
| 1515 | Start pumping out main bath with scroll pump. Takes 10 minutes to fall to 1 mbar. |
| 1525 | Pressurize MB with helium to 123 mbar above 1 atm. |
| 1530 | Check leak signal on IVC for the heck of it: 20,000 pbarL/s. Maybe the IVC seal was bad after all, hard to say though. |
| 1535 | Let 2.5 mbar He exchange gas into IVC for the warmup. (For standard cooldown it should be 63 μbar) MC(Cernox) temperature starts to rise a little |
| 1540 | Isolate IVC then switch pumping line over to Cryostat Vacuum Jacket port. Pump out line with Scroll Pump, switch to Turbo @ <0.1 mbar. |
| 1545 | Let 14.0 mbar N2 gas into vacuum manifold. Then open cryostat vacuum jacket valve to let this into the vacuum jacket. Vacuum Jacket → 540 ubar. |
| 1550 | Isolate fridge then start pumping LN2 Trap 1 contents to dump (so we won't have to keep filling the LN2 Trap bath) |
| 1555 | Currently DIB2qI Chassis→Drain/Source is very stable at only 1.088 MΩ |
| 1600 | MC(Cernox) = 79.1 K and rising at 4.6 K/hour. |
| 1615 | Stop pumping out LN2 Trap 1. Isolate trap, then connect fridge (both sides) to dump and 3He pump. Turn off 3He pump. |
| 1630 | MC(Cernox) = 83.5 K and rising at 10.3 K/hour. |
| 1650 | Examine ST-3-8 in the clean room and take some photos. Note that the copper foil cover plate has Kapton tape on the inner surface to prevent shorts. Verify that qOuter bias resistor reads open. Couldn't tell from the photos which of the 5 wirebonds connecting the four 10 MOhm resistors in series is bad. |
| 1655 | Matt exits lab |
| 1700 | MC(Cernox) / Sorb = 89.0/188 K |
| 1930 | MC(Cernox) = 115.4 K and rising at 9.6 K/hour. |