Beams-doc-1757-v2 (Revised) March 29, 2005 Rob Kutschke, CD/EXP In this revised version I have added the plots for March 29. All of the problems discussed in the previous version are now gone. The text of the previous version is below. The one real change is that Marv Olson inspected the timing on BPMs for which the previous studies suggested a timing problem. He indeed found timing problems and fixed them. He also adjusted delays so that the the D0 house has the beam appear at turn 10 (index 9) on all BPMs. The timing problem in the B3 house on March 27 is not yet understood but it has gone away. Some instances of 6 stable phases, instead of 5, remain. But the data for these BPMs is OK. The original text follow: The postscipt and pdf files included in this document show summaries of the tbt data from the shots on March 25, March 26 and March 27, 2005. With a few exceptions, the data looks very good. Without exception, all BPMs have discrete, not continuous phases. The remaining problems are discussed below. For the March 25 data, the plots show all BPMS from the houses, A3, B3, C3, D3, E3, B0. For the March 26 and March 27 data, the BPMS from D0 are included too. ( The D0 front end was rebooted soon after the shot on March 25 and I did not get the data on time ). For each BPM these files have 2 pages. The first page has three plots, 1) Position in mm vs turn number 2) Sum signal, in Echotek units, vs turn number. 3) Phase of the proton A signal, vs turn number. The second page has 6 plots. The left hand column shows details of the Fourier Transform of the position data. The right hand column shows details of the Fourier Transforms of the sum data. The top row shows details, at low frequencies where one can see synchrotron oscillations and quadrupole oscillations. If HV coupling is present, those lines would also appear in this frequency band; but I do not see any evidence for HV coupling in these data. The middle row of plots shows the magnitude of the Fourier Transforms over the rest of the frequency range. The bottom row shows a detail in the neighbourhood of the tune lines; it also includes the 1113/5 artifacts near 19kHz. One interesting BPM is VB35, page 23 and 24. At this bpm, the low frequency noise is very low and some interesting lines are present. I am not sure what they are. The other BPMs have too much noise at these frequencies to resolve the lines. The remaining problems are: 1) On a few BPMS, the magnitude of the sum signal depends the phase of the A signal: - See, VC37, page 43, in both files. - See HPD0U, page 93, in the March 26 file. The position information looks good even when this problem occurs. 2) On two BPMs both the sum and position depend on phase. - See VPD0U, page 95, and HPD0D, page 97, 3) On the March 27 shot, the timing for the B3 house is off. For all BPMs the first data appears at turn 6236 ( or index 6235, counting from 0). 4) On many BPMs, there are 6 discrete phases, not 5. In some cases, the 6th phase is regular but in other cases it is intermittent. In all other respects these data look fine. I believe that this problem is already understood as a race condition that can be corrected by adjusting delays. I wonder if problems 1) and 2) indicate that these BPMs are not yet properly timed in?