With a little bit of jiggling of the IPS sensor wires I got the sensor to completely fail, that is output to go to -10v (-32000 counts.) By power cycling the Kaman satelite box (not the HEPI Pier Pod) the sensor recovered. I repeated the process, jiggled the cable at the sensor end--no fault; jiggle at the Kaman box end--fault.
The Kaman satellite box is an off the shelf housing that is mounted inside the HEPI Pier Pod, the sensor cables go through a cutout in the Pier Pod to plug in. Maybe the shield of the lemo-like cable end is shorting on the Pier Pod box wall, or, hopefully not, it is a worse problem.
It is a little head scratching but I think the DC level of the IPS is problematic, This is regards to the alignment position. As the IPS was running yesterday after I unlocked, it was in a partially failed state during which the IPS readout was -7500cts; now with a coarse observed spectra looking pretty good (waiting to record a better one) it is now running about -5000. That 2500count difference is about 97000nm. Do I need to figure out how this will affect the Z, RX & RY Target Positioins...? It now should think it is higher than it was and will attempt to pull it down too low. Problem is most of the trends are on stops, before calibration, with a bad V2, etc.
Okay, assuming the current unlocked IPS values are what they would be if the V2 channel had not gone bad and my -7500 counts is correct for the unlocked bad state on V2. I calculated simple Z RX & RY values from the vertical IPS for the two states of the V2 sensor. I then applied the same percentage change to the Cart Bias Target positons for those dofs--close enough I think.
The loops are closed now, with no boost still, that problem remains.
The first plot attached here is the IPS & L4C caibrated spectra with the V2 IPS fixed and the HEPI Loops open. Compare this to the ones I put in last night, alhough most of those I'm pretty sure were with the HEPI Loops closed. V2 is certainly better and now V3 looks a bit worrisome but not as bad. However, if you look at the first plot from last night, where I show HAM4 from April, many of the IPS look similar to this HAM6 V3 channel. I'm pretty sure this HAM4 data from April are with the HEPI released from stops but with the control loops open.
Finally, attached is the HEPI Sensors with the loops closed. I'm not sure about the details but overall it looks like this fixed the IPS. I would suggest we look closer here to see if simple connector shell shorting to the Pier Pod case could be a systemic problem for all these sensors. If anyone wants to put their finger on this issue and decide on bug report, integration issue, or what, please do. I'll take assignments to further trouble shoot. To begin, I'll suggest using our HAM1 (still on stops) for a test unit.
Meanwhile--I still don't understand the Isolation controller magnitudes. I think this is why the boosts won't work. I'd like to collect new TFs with the ISI Damped and our final payload in place, although maybe HEPI isn't that sensitive to that change. Still new TFs and then go through all the commissioning steps to see if that changes the controllers and gets the boosts on.
Also remember, many of the IPS have limited range with the Bellows Shields position and I haven't run full static test diagnostics.
