There are two calibration runs for you to check out: these are runs
11117 (LEDTDC in SM3/4) and 11119 (LEDTDC in SM1/2).
Let me point out that the progress -if any- that we made yesterday
was after I asked Alice regarding the last time LEDTDC runs were
successful. These were some *March-April 1995* runs that I copied from
tapes and checked them on an event-by-event basis in order to see
what they have different with respect to the ones we run in September.
Well, there were obvious differences: while the March-April runs
had both ends of the ERP ADCs saturated, the latest ones did not.
Moreover, I noticed that whenever side "0" of the ERP was saturated,
both the fake box and LED bit in the SPU was present while this
did not happen when side "1" was saturated. This suggested to me
that the problem might be buit-in in the calibration software. After
poking around the current (v9 / 9-JUNE-1995) and previous calibration
driving software, I created a new version where obvious changes were
made. Try to have a look at VXMACB::$1$dia0:[SCINTSOFT.CALIB.LED.DBG]
The two runs I list on the top of this email have the FAKE
box and the LED bit in the SPU always present (for a total of ~1300
times, which is -I think- the number of generated LED events). The fake
box seems to be providing the stop for all the events. The only thing
I don't quite understand is the huge difference between the fake
box adc values in SM3.
--Erik