OD HV Relay Card Test Run 02/2000

HGB, 02/11/2000.

... under construction ...

Summary

During his 1st shift on Jan.30, 2000, Andy Stachyra swapped two of the original paddle cards with the new prototype relay cards. Location: hut 1, crate 1, cards 1 and 2. Due to some incompatibilities with the current crate mechanics, the new cards could only be partially inserted into the crate. Though functioning perfectly, I found from Andy's description and photos (see below) that the exposed cards could impose some hazard for visitors or shift persons going through the hut, especially since the HV traces of card 1 (outer right slot) were no longer protected within the crate's frame. I decided to have the boards removed again after a couple of days of collecting plenty of normal run data. So, after 8 successful days of normal runs without any problems, the old cards were swapped back to the original setup.

As mentioned in our upgrade proposal, we're planning to replace the subrack frames with new models anyway, which mechanically allow the new relay cards to be inserted completely.

All in all, the OD tube channels 0 through 23 were running through the new relay switches instead of jumpers from run 8322 (Jan.30, 21:18 JST) through run 8331 (Feb.7, 09:57 JST). Starting with run 8333 (Feb.7, 12:22 JST), the old configuration was used again.

Photos

Shown below are photos of the 2 prototype relay cards in the OD HV crate. Zoom in for more detail by clicking on the appropiate thumbnail below:

OD HV Relay Card Test Photos

OD PMT channel performance statistics before, during, and after the installation period

Here are samples of OD PMT pulse distribution data from 24-hour runs before, during and after the prototype relay cards were operating in hut 1.

run 8293: Jan.27/28, 2000; before relay cards, with original setup.
run 8328: Feb.03/04, 2000; mid time during relay card setup.
run 8333: Feb.07/08, 2000; after the original paddle cards were swapped back.

Direct comparison of the data samples shows no significant changes in the statistics, just typical fluctuations (within approx. +/- 2% relative between 24-hour "Normal" runs) as observed before making any changes on the OD HV setup. Especially, possible crosstalk between noisy channels (e.g. cable ID 20019 = channel 1.1.01.04) and neigboring "good" channels, and between "good" or noisy channels and "dead" channels (e.g. cable ID 20004 = channel 1.1.01.07) were of special concern, but the statistics show that the "healthy" channels have no different behavior than before - with some exceptions, but not directly related to the relay cards (see notes below).

In short, success! The new relay cards are working perfectly!

Here are the gory details:

PMT statistics table, listing and comparing the first 72 OD PMT channels (= first 6 paddle cards in hut1, crate 1).
Plots made from the table: (Click on thumbnail for zooming in)
Notes:
1. Noisy channel no. 1.1.01.04 (cable ID 20019) has fluctuating statistics, typically ranging from 50% through 80% "occupancy" (= 0.5 - 0.8 hit pulses per event trigger) as observed during the past years; therefore, the peaks in the comparison plots are expected.
  But the neighboring "healthy" channels show no effects.
2. Dead channel 1.1.02.04 (cable ID 20025) generated lots of hit pulses despite removed jumper. The fluctuations were pretty wild during the observation times so that relative statistical measurements between runs went far out of range (outside +/-100%, cut off in the comparison plots above).
  I suggest to disconnect the plastic connector of this channel soon from the rear of the paddle card! I suspect it has no contact to ground so that the signal wire acts like a long sensitive antenna now, and the statistics show that it's getting worse.
3. Channels 1.1.03.11 (cable ID 20035) and 1.1.1.05.02 (cable ID 20059) showed somewhat higher pulse rates after Jan.30, so after the relay card installation. But their signals were still handled by the original paddle cards, not the new relay cards. And the statistics didn't change after the original cards were swapped back in slots 01 and 02. So, this effect doesn't come from crosstalk!
  I suspect that the plastic connectors were sitting somewhat loose before Jan.30. And during the relay card swap the cables probably got wiggled slightly which then changed the electrical contacts in the plastic connectors, maybe.
4. Channel 1.1.02.11 (cable ID 20012) shows unexpectedly lower pulse statistics after Feb.07, so after the original paddle cards were replaced again. During the relay card operations, the statistics of this channel were the same as before Jan.30.
  Again, I suspect a loose contact in the plastic connector. Someone needs to check it at the next OD PMT "party".

back to Outer Detector High Voltage Paddle Card Upgrade page.