Make New Entry, Make Followup Entry
User name R. Michaels
Log entry time 17:08:58 on November 5,2010
Entry number 338691
keyword=double pulsing on LNE solvedI think I solved the multi-pulsing on the LNE which caused all the
scaler / ring-buff problems. Looking on inf. persistence at the
LNE signal which loads the scaler, I could see occassional multiple
pulses on the 10 usec scale. This was traced to a certain module
which was loaded down with outputs and seems not to like driving
too many outputs (e.g. removing one output helped). So, I fanned
out the signal to another module and also put the LNE through a
gate generator to ensure that the width is 35 usec. This seems
to have cured the problem, but we'll need to see data to be sure.
The multi-pulsing together with the hypothesis that the control bits
are out-of-phase by 1 data word wrt to the data bits explains *ALL*
the bad results. In addition, that recovery procedure I suggested
earlier today should work.
At the moment, the helicity is being flipped at 960 Hz (I don't know
why). The observed rate agrees. Previously this plot showed a
lot of jitter and was often a factor of two higher than the correct
rate.
When we were at 120 Hz, the observed rates were
helicity = 37.5 Hz (makes sense for Quad pattern)
QRT = 30 Hz (good)
LNE = 120 Hz
These are in the last 3 channels of the normalization scaler.
The scaler.map will be updated to show the names.
FIGURE 1
