Make New Entry, Make Followup Entry
User name R. Michaels
Log entry time 09:15:04 on November 9,2010
Entry number 339257
keyword=info about helicity problemsHere is a look at the percentage of "bad" data since Friday 4 pm
when we claimed to have fixed the multi-pulsing of the LNE that
affects the scaler that reads the helicity at the full rate.
There is evidently still some multi-pulsing, though it is less
than before Friday.
I'll need to take a 2-hour access to replace some modules and
safeguard against this.
Meanwhile, here's what it's doing to the data. It affects ring-buffer
of helicity info. See fig 1. Here are 101 runs of duration exceeding
2 min since Fri 4 pm. Runs 8385 up to now. Most runs lose less than
about 1% of data. There could be an offset due to imperfect parameter
for the fraction of gate-off, so it might be different than 1%.
Then there are a few runs with more missing. The worst case is
run 8461 which lost 10 min (15%) of data. Nothing special about
that run. It is scary because we have no control over when the
electronics decides to fail, assuming that's what is happening.
The other option for decoding the helicity is to use the TIR data
and the clock. So far, it seems like the TIR data are clean.
However, since the helicity is flipping at 960 Hz while the data
rate is typically 200 Hz or less, many helicity transitions are
missed and it will be a challenge to reconstruct the helicity
robustly using the TIR alone.
FIGURE 1
