Kyle and I independently leak checked IP4 gate valve on the beam tube side and found a very delayed He signal (many seconds to minutes). I bagged the flange joint with party store balloon mylar and found minutes delay from a vertex background signal of 4.8e-9 Torr-L/s to a leak rate of 8.5e-9 Torr-L/s, using UL1000 leak checker. It's still unclear if this flange joint is leaking or if He is migrating to the actual leak. Need N2 to flush out the He. After the leak rate crept up to 8.5e-9 Torr-L/s very slowly, I finally removed the bag and then it crept down even slower while I blew on the flanges.
Note that the flanges were somewhat gappy after installation (top and bottom of valve) a few months back. Gerardo torqued them metal to metal and found that the pressure improved on the IP4 pump side (its GV is closed), from April 13th.
IP4 was leak checked after it was installed last year with a small leak noted from 8" CFF: https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=40339
The background of the vertex volume is 4.8e-9 Torr-L/s of He, with UL1000 leak checker backing the main turbo and the vertex pressure at 1.9e-6 Torr (after recent vent). When leak checker+turbo are isolated from main volume, the background drops to 1.2e-9 Torr-L/s and when isolated from turbo drops to >>1e-9 Torr-L/s.
IP4 is 20-30 ft from main turbo+leak checker.
Bag the bonnet?
I can try that. I sprayed it and didn't see an immediate response.