sPHENIX discussion of the superconducting solenoid

[Kin Yip <kinyip AT bnl.gov>]

Agenda and powerpoint :  https://indico.bnl.gov/conferenceDisplay.py?confId=4182

1.   Carl Schultheiss presented a description and analysis of Monday's two 500 A ramps and their fast and slow discharges.  
  
     The main thing can be seen on p.13 (and p.15) that he showed the power supply voltage oscillated ~280 Hz and the difference between inner and outer was > the threshold of 50 mV voltage.
     After 36 times over the threshold, the quench detector system triggered the fast discharge.

     Carl attributed this to "imperfect" compensation which we'll have to work on tomorrow (the next test).  Carl later mentioned to me that the "trigger" was not due to "overcurrent" setting
 because the trigger ("quench signal") ---purple on p.13 --- came down (from high/good to low/bad) first before the PS_volt signal --- green --- went down.

1.1 Correction/clarification on Carl's 2nd slide (Carl agrees),

  (i) for the 1st ramp, it's 0.5 A/s from 0 to 300 A but it's 1/A s from 300 A to 500 A.
 (ii) the 3rd line from the bottom should say "Fast Discharge occurred at 496 Amps", not "469 Amps" as it's obvious from the slide 11.


2.  At the time of writing, there is still a mystery of a difference of almost 300 A between the Magnet Current and Power-Supply Current.   It's not a simple stupid scale factor as we have checked with
Winston Pekrul.   This is something that we need to sort out tomorrow, in addition to the compensation.

3.  Carl also pointed to a unsynchronous delay between Logger Boards A and B (which may be seen on p. 24).   This doesn't affect anything yet though he might cure it using the un-used
voltage tap channel (for VT11 as it tended to open at any given time).

4.  Jim Mills that we should write down our details.   I think that the Elog is a good place for all these records :

http://elog.pbn.bnl.gov:8080/elogs/entryList.jsp?ELOG=sPHENIX-Magnet-High-Field-Test_2018


5. On my slide, I showed that there was a temperature jump of 0.54 K for the MT01 sensor (the only one for coil left) which was due to our fast discharge at ~496 A.   Assuming 7.28 K would be
the temperature that the coil would go non-superconducting (which is true for 4600A according to Pasquale), I  did a quick but naive calculation assuming heat proportional current^2, it'd reach 7.28 K
if we did fast discharge at > 1123 A.

6.  I also mentioned quickly that strain gauge logger stopped around 2:30 pm on Monday.  I noticed it yesterday and we restarted the Manager yesterday.  But most, except 4, gauges
showed ~0.48... which meant that the voltage/resistance read ~0 (like a short ?!).   Last night, after 6:30 pm, I went to power off and unplug the power cord for 20 seconds and then re-plugged and switched on the KeySight module.   I restarted the Manager and saw the same ~0.48 in the Pet page.  I felt defeated and went home.  But this morning, Winston told me that he found that the logger was off since 6:30 pm, and after he restarted the Manager again, all the readings went back to normal !!


7.  We discussed whether we should run the test tomorrow on Thursday.  Because Paul informed us that the "tank" probably only got to ~50% tonight when he needs to divert cooling to the Magnet
while trying not to draw too much LHe, and at this point, the re-generation stops.   So, we might not have as many hours as on Monday.    The hope was that Monday's consumption of 8% per hour
hopefully won't happen again and we'll hopefully go back to ~6.5% per hour.

Since the works that we want to do need to use the real load (real Magnet at 4.5 K), we decide to run tomorrow anyway, starting 9 am.