Earlier this week I went up to Bates to measure the foils that were stretched and framed.  These were foils 2,7, and 8 from the '1st Glass' set.

The power supply and the ammeter were the ones used for the tests here at MIT, ported up to Bates for the purpose.  The gas was routed from a fume hood nitrogen tap.  For the first test (foils 2 and part of 8), a single line was run, and the pressure was adjusted manually (with very small twists of the valve on the fume hood) until the hiss was not audible (I pressed a finger against the outlet and could feel a slight puff when I removed it).

The foils were being stored in a dry box, but when we routed the hose to the test rig, we had to unplug it from the box.

The frames for the foils don't quite agree with the mount points on the test rig.  The mount points do seem to agree with the holes pre-cut in the foil.  In particular, when placed in the rig the hole at the bottom center misses the frame hole by ~0.5cm, but is aligned with the hole in the foil.  This was true for all three foils.

I put Foil #2 into the test rig and waited only a few minutes before beginning my measurements.  Waiting ~5 minutes for the currents to settle between measurements, I was able to get the foil up to comparable voltages as when it was tested here at MIT.  The only new feature I noted was that sectors 1 and 4 had strange behavior starting at 300V -- instead of relaxing to a constant leakage current, the current would drop to zero, then spike back up to ~1 or 2 nA, then drop to zero again.  This behavior was not seen during the measurements here at MIT.

During the first attempt on Foil #8, I already saw currents comparable (~0.2nA higher) to the MIT measurements, though again I only waited a few minutes between placing the foil and beginning to ramp the voltage.  Sector one began to draw current at ~400V, and continued to draw it even when I backed down from there, and sector 7 showed the 'bouncing' behavior from above at 200 and 300V (though not above).

On May 19 I returned and finished measuring Foil #8 as well as #7. At this point the gas line had been replaced with a T connection so that nitrogen was being supplied to both the dry box and the test rig.  The flow rate was much higher than at MIT, with a loud, audible hiss.  It did not seem to deflect the foil.  The rig was also now securely grounded both to the table and the electrical outlet.  

I was able to ramp Foil #8 to 700V and measure a current on most sectors.  Sector 3 drew large current even at 100V (which it did not during the previous test).  Sector 9 began to draw current at 700V after a single current spike, and continued to draw current even when I ramped it back down.  In general , the agreement between measurements there and at MIT was good.  Sector 2 had been problematic (slow to settle) at MIT, but was fine at Bates.

During the test the gas flow also began to fluctuate in time with some clicking from the dry box.  I believe a valve there was closing once the box reached the desired humidity, but because of the T connector, this was shunting more flow through the box.  These fluctuations caused current spikes at higher voltages.  To prevent this, I periodically opened the door to the dry box so that its humidity was ~15 instead of the 10.0 that seemed to trigger the valve).

Foil #7 was also stable up to 700V on the lower sectors.  Sector 2 drew ~75nA at 700V, but did not spike.  This value was much lower than the typical current spike or other problem currents I've seen.  Sector 6 did not seem to draw any dark current at any voltage, so I assume something is wrong there.  Sector 7 was only stable up to 600, and sector 8 began to draw a small current at 500V (9.5nA).  Sector 9 was also stable only to 600V.  The values were consistently lower than the measurements here at MIT.

Conclusions:

Overall, all three foils seem to be functioning after being mounted.

The various current draws seem to be transient effects (new ones appeared, old ones disappeared).  I suspect they're related to debris on the foils.

Some 'bouncing' behavior was seen when testing at Bates.  This hasn't been seen before, and may also be due to debris.

Foil #7 has what seems to be a dead sector (no leakage current), though I could see no mechanical break.  I didn't have a chance to troubleshoot very much, but did verify that the setup worked fine on another sector immediately before and after trying sector 6, meaning the problem must be in the foil or its connection to the test rig.

Higher gas flow definitely seems to enable higher voltages.  We should modify the procedure here at MIT to do this.

There seems to be a mis-drilled hole in the frames (or in both the rig and foil together), and possibly two.