From: Frank Geurts (geurts_at_rice.edu)
Date: Sun Mar 16 2003 - 13:38:06 EST
fred bieser wrote:
> it isn't clear to me if the quoted symptoms are all from a single run or
> several with consistent results (??) I agree with Bill that if (in several
> runs) things look good for a while and then go bad it is probably not the
> cabling; HOWEVER it is not impossible: if half of a differential pair is
> missing, signals might actually get thru for some time and then become bad
> thru pumping of the floating line(s). As a hardware guy, I am always ready to
> blame the software but in this case I know there has been activity in the
> area and I woiuld guess the software has not changed (?)
>
Software hasn't changed in a month or so and my analysis of the daq log
files (see startrg-hn) shows no evidence for hardware problems on the
TOF logic and readout side: everything runs smoothly up and until some
random moment during the run. What we see happen at that point is that
although the Bira generates LAMs all over the place it doesn't find a
word at its input register. This is most likely *not* a Bira failure
since the next run it'll operate again and performs exactly like we
expect. All the raw data we take looks perfect, so -again- I don't see
any evidence of hardware failures on TOF's side.
We know for a fact that the rear connectors of the TCD can come loose
(it wouldn't be the first time) and we also know that last Wednesday a
lot of activity took place at the TCD rack. The rear connector carries
the word that goes to the Bira input register and that is were we see
the (random) erratic behaviour. I really think we should check whether
this connector is properly plugged in.
TOF is dead in the water now for three days ... a quick access of 10min
is all the trigger folks would need to re-plug the back connector.
-frank
This archive was generated by hypermail 2.1.4 : Thu Jul 24 2003 - 00:39:38 EDT