Make New Entry,
Make Followup Entry
User name R. Michaels
Log entry time 17:26:24 on March11,2009
Entry number 265609
keyword=HAPPEX DAQ deadtime checks
To test how high a rate we can run the HAPPEX DAQ (relevant to
the future) I studied ROC23 (counting room crate) which should
be our slowest crate and the "bottleneck".
ROC23 reads 8 old-style ADCs, 1 scaler, 1 FLEXIO, and the TIR.
This is the most of any crate.
The deadtime on a scope was 113 to 121 usec (fluctuating). The
scope showed some odd flat-lining. I may post a follow-up in
the HAPLOG explaining this, but it naively implied a deadtime.
So, I took this TIR output signal and fed it into the scaler
together with the input to the Trig. Super. -- the ratio gives
the deadtime quantitatively.
Next, I increased the rate by changing the oversampling. It's
a cheap way to do it but you'd rather increase the MPS.
I stepped : 30 Hz, 300 Hz, 600 Hz, 750 Hz (the highest I could
go was 750 Hz with this method) and from the scaler ratio the
deadtime was always zero. Also, the observed DAQ rates on the
scope were consistent except at 750 where it seemed to be 1.2 kHz,
but in CODA the graph showed consistent rates, so 750 Hz was
750 Hz. Scope may be flakey (?). Need to repeat this summer.
So far, it looks like we can run at least 750 Hz.