Bad Tiles on the EPD
Executive Summary
W PP1 TT6 -> EQ2 Board 8 Channel 24 - Dead at SiPM/FEE (confirmed 6/7)
W PP2 TT10 -> EQ3 Board 4 Channel 14 - Dead at SiPM/FEE (confirmed 6/7)
E PP1 TT31 -> EQ4 Board 6 Channel 7 - Reconnected. Should be good again after 6/7 - TBC.
W PP2 TT27 -> EQ4 Board 14 Channel 11 -> QT Channel is bad, remapped. Result TBC.
Details
Maria has brought to our attention several bad tiles, see: drupal.star.bnl.gov/STAR/blog/mstefaniak/Monitoring-QA-EPD-plots
Maria's Mapping Test is at: drupal.star.bnl.gov/STAR/system/files/7patternTest_02june23_.pdf
W PP1 TT6 -> This tile is connected to EQ2 Board 8 Channel 24 in the db as of today 6/7.
This tile was connected to EQ4 Board 8 Channel 0 on May 12th. It looked bad there.
Mike noted this tile looked bad in his bias scan. It seems that we saw some dark current before, but we are no longer seeing that. (GUI Confirms no dark current.)
Since the tile looks bad plugged into multiple QT channels, this is likely a dead channel.
W PP2 TT10 ->This tile is connected to EQ3 Board 4 Channel 14 in the db as of today 6/7.
This tile was connected to EQ2 Board 14 Channel 14 on May 12th. It looked bad there.
This channel has no dark current and an abnormal RMON. (Gui confirms).
Since this tile looks bad plugged into multiple QT channels, this is likely a dead channel.
E PP1 TT31 -> EQ4 Board 6 Channel 7 in the db as of today 6/7 and seems to be the case since at least May 12th. This tile looks perfectly fine in the offline analysis (day 148) but not on day 157. The online plots (blue ones) show it as completely empty. This tile looked fine for run 24151011, there was a long access and then it did not look fine, starting in run 24151063.
It looks like nothing in the database has changed (any of the 3), but that the physical connection was somehow changed.
W PP2 TT27 -> Mismatch on 7 patterns test -> Connected to EQ4 Board 14 Channel 11 on May 12. Looks good in the offline plots. This looks like the bias is a little funny looking at day 157, but the db says that it hasn't changed. The pedestal on this channel is pegged at zero, so the question is whether there is an issue with the QT channel or with the FEE pulling the pedestal down. The mapping check shows that the pedestal is moving with changing vped, and the default vped setting seems reasonable.
Mike/Tristan Access Report
W PP01TT06 - we had suspected that this one was dead, and it is. Zero output of the diffRx board and zero dark current on the GUI. Either a dead FEE channel or (more likely) a dead SiPM.
W PP02TT10. Exact same story. We had expected to find this.
E PP01TT31 - This one had a reasonable dark current, and it was working just fine before the previous access, and then it completely vanished after the access. We suspected that it got pulled out or loose or something. And that was exactly correct. The cable was very loose at the QTB. Notably, this QTB is sitting right next to the newly-installed QTD. Clearly, Christian/Chris accidentally loosened this cable when they installed the QTD. We also looked at this signal on the scope and it looks very good. We see cosmic MIPs, and the DC level is spot-on, at ground (0V).
W PP02TT27 - on the bias scan, and also on the online monitoring plots, this one looked to me like it had a "below-zero pedestal." That would explain both the online plots and the fact that it failed most (but not all) of the Patterns. Also, the pedestal files itself corroborate a below-zero pedestal. That means it could be either the fault of the FEE (bad DC offset coming from FEE) or the QT itself. On the scope, the signal from the FEE looks very good; same description as E PP01TT31 that I just described. Therefore, I blame the QT itself, and we have remapped W PP02TT27 to EQ3 Slot 4 Channel 15.
in summary, two tiles are dead for the rest of this run. One was loose and should look fine again. And one was remapped.
- rjreed's blog
- Login or register to post comments