06/02/15
A look at the data rate at which Pause/Resumes start to cause missing data.
Pulser settings:
Slow comparitor thresholds were altered using the ASIC4 page to artificially increase the data rate in the ASICs.
Default settings for all modules, apart from the changing slow comp threshold.
nnaida13+14 both functioning well. nnaida11+12 both have one noisy channel contributing to high data rate.
Echoing Alfredo's previous post, using the AIDA Good Events and AIDA Data Rate from the stats page as metrics, I found there to be no Pause/Resume items present in the data up to a Good Events rate of ~250-300 kHz and Data Rate of ~1-1.2 MHz.
Around this point the data shifts very quickly from containing no Pause/Resume items to having a rate of ~18 Hz, almost as a step function. At this point the Good Events and AIDA data rate sharply reach a maximum rate of 300 kHz and 1.2 MHz respectively. The rate of AIDA SYNC data items drops sharply at this point and decreases slowly with decreasing threshold.
Included are some plots showing this for each of the modules and some summary plots.
Further to the data rate at which the appearance of Pause/Resume items causes significant data loss, I looked at what happens when you break this threshold and then come back down, i.e. do the Pause/Resume items persist?
In short, no. Starting from a threshold of 64 and working down to 1, then taking the slow comparitor threshold back to a level at which no Pause/Resumes were present, the Good Events, AIDA SYNC and AIDA Data rates all went back to their original values within a couple of refreshes of the stats screen (and stayed as such thereafter). |