Monitoring glitches in the frequency divider N. Blaskovic N. Blaskovic FONT Meeting
Contents Monitoring the glitches of the 714-to-357 MHz frequency divider: set-up and results The effect of the use of Colin’s filter to reduce glitching N. Blaskovic FONT Meeting
SET-UP 1 22/04/12 00:56 (onwards) KEY N. Blaskovic FONT Meeting Doug’s original set-up SET-UP 1 22/04/12 00:56 (onwards) BNC 18 “kicker pre trigger” fanout scope CH1 (trigger) BNC 16 LO box (In Out Raw) scope CH2 20 16 Φ/2 (2008) scope CH3 Φ/2 (2011) scope CH4 KEY Power splitter 1.5 Hz Amplifier (dBm) Φ/2 Frequency divider (year) 714 MHz Attenuator (dBm) Colin’s filter 357 MHz N. Blaskovic FONT Meeting
SET-UP 1 N. Blaskovic FONT Meeting Data collection (Doug’s scope DAQ): only the value of CH2, CH3 & CH4 at the sampling index (shown by the circles) are logged for each acquisition (approx. every second) Voltage of CH2, CH3 & CH4 vs. time for an acquisition: N. Blaskovic FONT Meeting
SET-UP 1 N. Blaskovic FONT Meeting 20 glitches in CH4 (no filter) over the first 17500 acquisitions 0 glitches in CH3 (with filter) over the first 17500 acquisitions ‘Slipping’ of CH2, CH3 & CH4 relative to CH1 trigger evident for acquisitions 17500-43000 Plot of sampled CH2, CH3 & CH4 voltages vs. acquisition number: 6 h (0.8 acq/s) N. Blaskovic FONT Meeting
SET-UP 2 22/04/12 17:52 (onwards) KEY Freq. dividers interchanged BNC 18 “kicker pre trigger” fanout scope CH1 (trigger) BNC 16 LO box (In Out Raw) scope CH2 20 16 Φ/2 (2011) scope CH3 Freq. dividers interchanged Φ/2 (2008) scope CH4 KEY Power splitter 1.5 Hz Amplifier (dBm) Φ/2 Frequency divider (year) 714 MHz Attenuator (dBm) Colin’s filter 357 MHz N. Blaskovic FONT Meeting
SET-UP 2 N. Blaskovic FONT Meeting No glitching recorded in either CH3 or CH4 (over around 4 hours) ‘Slipping’ of CH2, CH3 & CH4 relative to CH1 trigger evident Voltage of CH2, CH3 & CH4 vs. time for an acquisition at 18:17 (FONT PC time): Sampling close to CH2, CH3 & CH4 peak N. Blaskovic FONT Meeting
SET-UP 2 N. Blaskovic FONT Meeting No glitching recorded in either CH3 or CH4 (over around 4 hours) ‘Slipping’ of CH2, CH3 & CH4 relative to CH1 trigger evident Voltage of CH2, CH3 & CH4 vs. time for an acquisition at 18:51 (FONT PC time): 34 minutes later, the sampling point is no longer sampling CH2, CH3 or CH4 peaks! N. Blaskovic FONT Meeting
SET-UP 2 N. Blaskovic FONT Meeting Possible sources of ‘slipping’ of CH2-4 relative to CH1 trigger: Phase drift of the 1.5 Hz trigger relative to the 714 MHz Varying shape of the falling edge of the 1.5 Hz pulse wobbly CH1 (trigger) N. Blaskovic FONT Meeting
SET-UP 3 23/04/12 01:17 (onwards) KEY CH3 used as trigger N. Blaskovic BNC 16 LO box (In Out Raw) scope CH2 20 16 Φ/2 (2011) scope CH3 (trigger) CH3 used as trigger Φ/2 (2008) scope CH4 KEY Power splitter Amplifier (dBm) Φ/2 Frequency divider (year) 714 MHz Attenuator (dBm) Colin’s filter 357 MHz N. Blaskovic FONT Meeting
SET-UP 3 N. Blaskovic FONT Meeting This set-up allows a clear count of glitches. Left running overnight, the number of acquisitions collected was 230412. Number of glitches (in CH4): 2022 Plot of sampled CH2, CH3 & CH4 voltages vs. acquisition number: 13 h (5 acq/s) N. Blaskovic FONT Meeting
SET-UP 3 N. Blaskovic FONT Meeting Zoom in... Plot of sampled CH2, CH3 & CH4 voltages vs. acquisition number: N. Blaskovic FONT Meeting
SET-UP 3 N. Blaskovic FONT Meeting Zoom in further... Glitching on occasions occurs between consecutive acquisitions, with an acquisition rate of 5 acq/s Plot of sampled CH2, CH3 & CH4 voltages vs. acquisition number: N. Blaskovic FONT Meeting
SET-UP 3 N. Blaskovic FONT Meeting But glitches in CH3 (with filter) are indistinguishable from glitches in CH4 (no filter): Glitch in CH3 Given that CH3 is set as trigger, CH2 (714 MHz) is shifted through a whole period CH3 (357 MHz, filter) does not glitch, as it is acting as the trigger CH4 (357 MHz, no filter) apparently glitches, due to the glitch in CH3 (trigger) N. Blaskovic FONT Meeting
SET-UP 3 N. Blaskovic FONT Meeting But glitches in CH3 (with filter) are indistinguishable from glitches in CH4 (no filter): Glitch in CH4 Given that CH3 is set as trigger, CH2 (714 MHz) undergoes no change CH3 (357 MHz, filter) undergoes no change CH4 (357 MHz, no filter) glitches N. Blaskovic FONT Meeting
SET-UP 4 23/04/12 15:50 (onwards) KEY Initial set-up; no filters BNC 18 “kicker pre trigger” fanout scope CH1 (trigger) BNC 16 LO box (In Out Raw) scope CH2 20 16 Φ/2 (2008) scope CH3 Initial set-up; no filters Φ/2 (2011) scope CH4 KEY Power splitter 1.5 Hz Amplifier (dBm) Φ/2 Frequency divider (year) 714 MHz Attenuator (dBm) Colin’s filter 357 MHz N. Blaskovic FONT Meeting
SET-UP 4 N. Blaskovic FONT Meeting 1 glitch in CH4 (2011 frequency divider, no filter) 0 glitches in CH3 (2008 frequency divider, no filter) Plot of sampled CH2, CH3 & CH4 voltages vs. acquisition number: 2 h 20 min (0.8 acq/s) N. Blaskovic FONT Meeting
Summary The 357 MHz signal coming out of the frequency divider does glitch Glitching occurs as often as every 0.2 s Rate of glitching varies throughout the day Colin’s filter reduces the occurrence of glitches: not a single glitch was recorded when using the filter N. Blaskovic FONT Meeting
Epilogue: a single slide on kicker gain matrices N. Blaskovic FONT Meeting
Comparison of kicker gain matrices: December 2011 – April 2012 P2 P3 GainDec* = K1 -3478 476 K2 889 -2341 Gain18AprCal2** = -8243 759 2034 -4741 Gain18AprCal2 ./ GainDec = 2.37 1.59 2.29 2.03 * final kicker calibration, as used for fbRun6,7,8_Board1_coupled_141211 ** kicker calibration 2 on 190412 N. Blaskovic FONT Meeting