User name Armstrong
Log entry time 21:20:40 on August 03, 2005
Entry number 150597
keyword=Target Fan Speed StudyTarget fan speed study
D. Armstrong, J.M. Finn. J. Katich, A. Kelleher (Part of the W&M/HAPPEX Mafia - Go Tribe!) To try to further reduce target density fluctuations, we try pushing the target to higher fan speeds.All data taken during this study should still be "production quality".
Cell Pressure = 212 psi
CHL 4K flow: 23.3 g/s
Beam current constant at 35 µA
Initial fan speed = 60 Hz
Fan speeds are as measured by tachometer.
Run Fan Htr_aft Htr_fore Cell_aft HP Htr
# (Hz) (K) (K) (K) (W)
---- ----- ------- -------- -------- ------
4090 65 6.79 6.70 6.97 109
4991 70 6.77 6.70 6.93 75
4092 72 6.76 6.69 6.93 67
The Target controls software would not let us go above
72 Hz. Probably this can be over-ridden. However, we notice
that the high-power heater exhibited largish (+-20 W) oscillations
of about 30 seconds in period at the 72 Hz setting (which were
also manifested in the same oscillations in the CHL flow).
Perhaps the PID parameters on the temperature regulation
might need tuning at this fan speed?
Note that to go much higher in fan speed would require either more CHL flow or raising the cell temperature, as we have not so much ballast left in the high-power heater power, and the fan dumps more power into the target at higher speeds. Note also the fact that the distribution of temperatures in the target loop gets more equilibrated at higher fan speeds, an evidence for more uniform cooling (and, I hope, smaller density fluctuations).
We decide to back down to 70 Hz for now, and take a longer run (Run 4093).
Results of the prompt analysis of these runs:
All values are the widths in ppm of the regressed distributions; BLUMI_SUM is the sum of the BLumis, DETLO is the sum of the Left and Right arm detectors, and DET1, DET3 are the individual Left and Right arm detectors.
Run Fan BLUMI_SUM DETLO DET1 DET3 # (Hz) ---- --- --------- ----- ---- ---- 4088 60 738 1168 1526 1536 4090 65 665 1131 1489 1492 4991 70 660 1126 1523 1474 4092 72 643 1134 1510 1492Conclusion:
A clear reduction in detector fluctuations is seen going from 60 to 65 Hz fan speed. A smaller decrease appears in going to 70 and to 72 Hz; perhaps buried in the statistics. A consistent improvemment is also seen in the detector widths.
It appears I have reduced the detector widths at this beam current by 3.4% (meaning a reduction in run time for the same statistical precision of about 7%). This may seem small - but 7% of the remaining 30 days in the scheduled run is 2 days of data taking.
Decide to stick at 70 Hz for the mean time. Perhaps we should see if we can change the software limit on the fan speed.
Alternatively, I am tempted to try running the cell at higher temperature, to see the effect on the fluctuations.
