User name R. Michaels
Log entry time 15:57:23 on October21,2001
Entry number 74905
keyword=BPM readout modification
Today I modified the readout of the Struck 7510 "burst mode" ADCs
for BPM/raster. Purpose was to reduce the phase jumps observed
between different ADCs. (About which e00102 probably doesn't care.)
I increased the frequency by a factor of 2.5, going from 4 microsec
per clock cycle to 1.6 microsec. Then I skip every 2nd buffer location
and read 4, thus sampling ~1/5 of the rastered orbit. (Yes, e00102
does not raster for most of its data.) Previously we read 6 samples, so
it should be checked that software like ESPACE does not complain.
Should be ok since the number of samples is written as part of a header
to be decoded. The deadtime was reduced (75 instead of 90 microsec).
I looked at rastered current, and saw typically 1/5 of the sine wave.
I cannot easily verify that we solved the phase jumps until we get beam.