SLC Beam Phase Measurement Test Experimental Results
Measurement were taken of the SLC beam and klystron phase using the beam phase measurement test setup. The system was installed in sector 30, looking at the signals on the beam phase cable from station #8. The following results were obtained.
System Noise: Measured using the calibration path signal. For a signal amplitude of approximately 1/2 full scale, the measured phase noise (pulse to pulse) was 0.31 degrees RMS. (S-band).Note: measurement was 0.14 degrees with the measurement feedback set for 5 averages.
System Linearity: Measurements of the calibration signal phase relative to the reference phase were taken for various settings of the step attenuator. Based on the self consistency of these measurements a system phase non-linearity of 0.27 degrees was found for a 20dB signal level range, and 1.1 degrees for a 40dB signal level range (limited by noise in the measurement - signals were averaged).
Phase Measurement Drift: The measured drift of the calibration channel relative to the phase reference channel was 1.45 degrees (peak to peak) S-Band over approximately 2 hours. Note that this measurement includes drifts in the calibration amplifier and attenuator chain (see schematic). Drift in these devices would not contribute to measurement drift for real measurements. This should be taken as an upper limit.
Measured Klystron Vs Beam phase drift: This measurement was performed with the klystron always on standby time, and the apparatus alternately measuring the structure signal at beam time and at standby time. This was done to avoid interfering with beam operation. The measured drift is shown below.
Other multi-hour runs gave klystron Vs beam phase drifts of a few degrees to 10 degrees. It should be noted that the accelerator was being operated for the SLC collider during these runs. Some of the variation may be intentional beam tuning. The measured phase variation was 1.35 degrees RMS., and is not believed to have been limited by instrumentation.
Early results from processing the time dependant data within a single RF pulse indicate that it should be possible to directly measure structure de-tuning by looking at the beam induced fields. This system is also being used to evaluate klystron phase noise.
Additional Note on Measurement Noise: Later experiments with the NLC timing distribution system which uses the same digitial phase detection system have indicated a problem with the digitizer used for these experiments. The digitizer, a LeCroy 2262 uses two 40MHz digitizer, interleved to provide 80MHz output. Due to the logic design in the digitizer, even when single channel 40MHz mode is selected there are two possible digitizer sample times for a fixed trigger time. If the digitizer were perfectly linear, this would not affect the measurement results, however real non-linearities appear as noise on the output. This noise may have limited the results of this experiment.
Page by Josef Frisch frisch@slac.stanford.edu 04/22/2002
Links to: the NLC Home Page, the SLAC and KEK Institutional Home Pages.