Beams-doc-1737-v2


The accompanying postscript and pdf files show 
the proton position and sum signals from the BPMS 
HA32 and HE32 for the shot which went in at 14:23 
on March 15, 2005.

  The information comes from the IQ data in the
lumberjack data logger.  The computed position
does not include the survey offsets.
There is approximately one second between points.

   One can see the first proton bunch
go in at point 12 and the second at point 43.
The proton position is set to zero when the
sum is below threshold (50 counts).

   The odd thing is that the HE32 data is exactly
fixed from points 12 to 43.   The HA32
data shows the normal fluctuations.
Every BPM in the E3 house shows this fixed
behaviour.  All of the BPMS in A3 and D3
show the expected varitions, similar
to HA32.  The BPMS in B3 and C3 were not datalogged
at this time so it is not possible to check them.

  At the previous injection, Monday March 14
at 8:56 AM, all 3 BPMS in all three houses,
A3, D3, E3 had the expected variations..


  In version 2 of this document I have added several
new gif files.  The all show similar plots to the
above - details of sum and position signals for a period
of a minute or two at the start of an injection.

   All of the plots have 5 traces:
  - T:BEAM      - Total beam intensity
  - T:HPE32     - Position at HE32
  - T:HPE32[1]  - Sum signal at HE32
  - T:HE32IQ[1] - I from Proton A
  - T:HE32IQ[2] - Q from Proton A


e3_ok_mar14.gif
 - From a shot on March 14. All looks good.
 - HA32 and HD32 also look OK at this time.
 - Note that before injection the sum is zero and the position is
   off scale.

e3_bad_mar15.gif
 - This is the same data as in the postscript and pdf files
   mentioned above.
 - After the first bunch went in and before the second
   bunch went it, the proton sum signal has a sensible
   value but it has an exactly fixed value.
 - During the same time period, the proton position is fixed at a silly
   value.
 - At this time the I's and Q's are also fixed - I checked the
   printed values by hand - they are exactly the same.  They do
   not have a variation which is too small to see on this scale.
 - At this time HA32 and HD32 looked OK.

e3_bad_mar17.gif
- This is a new problem.
- Before injection, the proton sum signal has a non-zero value and the
  position is fixed at zero, not off-scale.
- After injection the data seems to make sense.
- The plots for HA32 and HD32 look the same.

e3_bad_mar17_extended.gif
- Same plot as the previous one but extended to about 20 minutes
  earlier in time.  Note that the system became confused in the
  middle of one of the tuneup stores.
- Mike Martens suggests that maybe there was a failed injection 
  of a second bunch at the time that the system became confused.
  He can check this.
- HA32 and HD32 look the same.