Hot tower and preshower channels in BEMC run 7

This is an update of channels I saw that may be worth looking at in the BEMC and BPRS if people are looking at the hardware now STAR is offline

BEMC Tower dynamic masking

This analysis comes from the log files of the StUpsilon trigger Run7 dAu, with 2285 Runs:

The StUpsilon masks hot towers (up to 25 maximum) per run, in addition to the hot towers listed on the database. It starts with an empty mask, then takes the average hits over all channels. Any that go above average are marked hot, if they recover during a run then they are removed from the mask.

Channels that recover during a run are most interesting as this shows a tube is fluctuating over a run

Plot of tower softId that recovered during a run, (see below for sorted list)

Plot of tower softId that went bad

This plot shows all the channels that went bad, which may show a hardware problem, or may just be high gain.The redline is the ones that recovered superimposed, so you can see how rare that was

Plots Made Into a Sorted list

These plots on there own are not easy to read, so I made a text file from them, with the counts sorted by softId vs. the frequecy they recover. The list is sorted from highest frequency to lowest so the first few entries are probably the best to look at

For 2285 runs in d-Au, the worst offender for recovering was channel 4130, for 49 runs (2%) and going bad on 227 runs (10%)

List of softId that recover, sorted by frequency

And again for the bad channels, the first entries are probably high gain

List of soft Id that went bad,sorted by frequency


The preshower has problems as well, as we are losing whole crates, a first thing to look at would be to see if we can get them back online (this kills about 20% odd)

The best paper reference to print out is Will's tube layout table, he made a lot of notes on the BPRS FEE here:

Will's table

This has Will's notes on bad tubes that he found dead FEE etc but lets list the highlights;

Main points that need looking at

This is a the problems that messes up most of the preshower

Bad PMT box FEE or power problems that I mask: 44 and 7 in Run 6 and parts of 7 in Run 7

PMT box 44 is totaly dead and kills 80 odd tubes, box 7 can't seem to make up its mind, but is not healthy

The Preshower has a 4x4 array structure (see Will's), the list of bad arrays I currently mask are:

This is small potatos to the crates but anyway:

West masking


14, 1

12, 2

10, 5

5, 1

4, 5

2, 4

2, 3

26, 4

24, 4

16, 3

16, 2

East masking

39, 2

32, 3

There is a list of individual channels, but that may likley be fiber optics and hard to fix