[Dsg-rich] [EXTERNAL] Re: [New Comment] RICH HV issue

[Marco Mirazita]

Hi all,
the reasons why I doubt the voltage spikes are real are that, as Valery 
said, the data don't show any spike in the current, and also that we 
have set in the HV crate a limit on the supplied voltage to 1100V (the 
maximum recommended by Hamamatsu).
Best regards,
Marco


Il 2020-01-22 22:00 baltzell at jlab.org ha scritto:
> RICH HV ISSUE [1]
> 
> 	* Edit [2]
> 	* Delete [3]
> 
>  Lognumber 3768485 [1]. Submitted by mirazita [4] on Wed, 01/22/2020 -
> 11:59 [5].
> 
>  Last updated on Wed, 01/22/2020 - 12:17
> 
> There is 1 comment...
> 
>  THE UNV/OVV STATUS AND [6]
> 
>  _new_
> 
> by baltzell [7] on Wed, 01/22/2020 - 15:57
> 
> The UNV/OVV status and measured voltage excursions in the archive are
> both isolated to a single HV board and time-coincident. So pretty
> obviously not a "problem in the readout of the epics data".
> 
> 		LOGBOOKS:
> 		RICH [8]
> 
> 		TAGS:
> 		RICH [9]
> 
> 		ENTRY MAKERS:
> 		Marco Mirazita
> 
> 		BACKLINKS:
> 		Follow-up Re: RICH HV issue [10]
> 
> During the last days, the shift takers noticed random alarms signaling
> spikes in the RICH HV system.
>  I downloaded the epics data of the monitored voltage Vmon and current
> Imon of the HV crate for each PMT tile.
>  The plots are in this file (too big to attach to this logbook entry):
> https://clasweb.jlab.org/wiki/images/a/a6/PlotHvMyaData.pdf [11]
> 
> Each plot is one tile. The red line is the Vmon (in V) and the black
> line is Imon (in muA).
>  The horizontal axis in the time in minutes, the zero corresponds to
> the 00:00 of Jan. 20, the end of the axis to the 00:00 of Jan. 22.
>  The data are sampled every 2 seconds.
> 
> We see:
> 
> 1) According the epics data, at time=2500 (Jan 21 around 5:40pm) most
> of the HV have been turned off for about 100 min.
>  However, the following tiles don't show this down time at all in
> epics: 59,60,80,81,97->94,102->106,111->114. The following tiles show
> instead a much shorter down time: 64->67, 71->74, 78,79,86->89,115.
>  The time corresponds to when a communication error happened, that
> required a reboot of the HV crate (see elog
> https://logbooks.jlab.org/entry/3768188 [12]).
> 
> 2) The following tiles show some random spikes over time: 46-47,
> 51->53, 57-58.
>  The Vmon spikes are sometimes quite big, reaching up to 1200 V.
>  There is no correlation between the Imon and Vmon, as I would expect
> if a hardware problem occured (with perhaps one exception in tile 47).
>  However, these tiles are located close each other, on the left side
> of the panel in row 12,13,14 (see:
> https://clasweb.jlab.org/wiki/images/1/14/RICH_HV_map.pdf [13]) and
> are connected to the patch panel B.
> 
> My conclusions:
>  The fact that the tiles showing random spikes are located close each
> other is suspicious. However the pattern of these spikes doesn't
> clearly indicate a problem in the HV crate. it looks to me more like a
> problem in the readout of the epics data.
> 
> 
> Links:
> ------
> [1] https://logbooks.jlab.org/entry/3768485
> [2] https://logbooks.jlab.org/entry/3768485/edit?destination=email/send
> [3] 
> https://logbooks.jlab.org/entry/3768485/delete?destination=email/send
> [4] https://logbooks.jlab.org/user/mirazita
> [5]
> https://logbooks.jlab.org/entries?start_date=1579708754&end_date=1579715954&book=RICH
> [6] https://logbooks.jlab.org/comment/26354#comment-26354
> [7] https://logbooks.jlab.org/user/baltzell
> [8] https://logbooks.jlab.org/book/rich
> [9] https://logbooks.jlab.org/tag/rich
> [10] https://logbooks.jlab.org/entry/3768517
> [11] https://clasweb.jlab.org/wiki/images/a/a6/PlotHvMyaData.pdf
> [12] https://logbooks.jlab.org/entry/3768188
> [13] https://clasweb.jlab.org/wiki/images/1/14/RICH_HV_map.pdf