Displaying report 1-1 of 1.
Reports until 18:33, Friday 27 October 2017
LHO VE
kyle.ryan@LIGO.ORG - posted 18:33, Friday 27 October 2017 - last comment - 11:13, Saturday 28 October 2017(39193)
Y2-8 ion pump off for unknown reason
I noticed a pressure "bump" at Y-end recently and have now confirmed that the ion pump mounted at BT port Y2-8 has been off for the past few days.  Today's troubleshooting including re-enabling the High Voltage output.  This resulted in power limited (700V @ 0.5 amp) energization (no ion pumping while inadequately energized).  Next, the HV cable was moved to he other HV channel with a similar result.  The controller-end of the HV cable was disconnected and a VOM used to measure the cable+pump resistance (2.5 Mohm, insufficient output voltage of the meter makes this test "almost" useless).  The HV cable was reconnected to the controller and the pump-end disconnected. In this configuration, the controller wouldn't output HV as the cable connector shield wasn't grounded and, it would seem, the cable isn't shorted.  

I noticed that the pump-end of the HV cable was noticeably "warm" after applying 700V at 0.5 amps.  This observation combined with the interpretation that the cable or pump-end connector aren't shorted indicates that the low resistance is internal to the pump itself.  Ahhh!  The fabled "whisker" theory.  

Monday, we will adapt a custom HV cable and use the "Big Gun" 10,000V 600 watt power supply to, hopefully, vaporize any troublesome whiskers that have formed between the electrodes.  



Comments related to this report
john.worden@LIGO.ORG - 06:36, Saturday 28 October 2017 (39194)

The pump is probably in a condensing environment at times so there may be conductive buildup on the pump feedthrough or the interior of the cable connector. 

I would be surprised at whisker buildup as that pump has not pumped significant gas loads. 

Also, I would fake out the interlock at the pump and make sure you can get high volts on the cable alone.

kyle.ryan@LIGO.ORG - 11:13, Saturday 28 October 2017 (39195)
Yes, I'll disconnect the pump-end connector and then ground the shield.  This should satisfy the controller and get it to energize the HV output.  As John W. points out, the pump-end connector shorting would explain why it was warm and is much more likely than the pump being internally shorted when considering the pumps torr*L exposure history.  
Displaying report 1-1 of 1.