Last week, Keith posted the results of a study of folded magnetometer channel data (alog 29166) aimed at understanding the results of recent changes to the timing system (primarily LED reprogramming and power supply switching). This is a follow-up, looking at the spectra of the same channels, and tracking the behavior of the two combs which the timing system interventions were intended to mitigate.
Detailed plots
Overview table (daily spectra, selected dates)
Full data set (daily spectra)
Full data set (cumulative spectra since Jul 1 2016, covering date ranges where Fscan SFTs were available)
These plots were generated from Fscans + spec_avg_long + my own plotting tools.
Timeline
July 14-21 comparison: before and after initial updates to timing slave card firmware (blinking LEDs turned off in many places, but not on timing fanouts)
July 21-Aug 6 comparison: firmware updated for EX fanout; CPS timing fanout power supply changed
Aug 6 - Aug 18 comparison: firmware updated for CER, MSR, EY fanouts
Notable features
"2 Hz" comb here means the 1.999951 Hz comb which is visible at approximate odd-integers. "1 Hz" comb here means the comb with 1 Hz spacing and 0.5 Hz offset.
H1:PEM-EX_MAG_EBAY_SUSRACK_X_DQ
1 Hz comb improved July 14-21 at low frequencies (<15 Hz)
2 Hz comb unchanged
H1:PEM-EX_MAG_EBAY_SUSRACK_Y_DQ:
1 Hz comb improved July 14-21 at low frequencies (<15 Hz)
2 Hz comb mostly unchanged, slightly improved above 50 Hz from 7/21 to 8/6
H1:PEM-EX_MAG_EBAY_SUSRACK_Z_DQ
Both combs mostly unchanged
H1:PEM-EY_MAG_EBAY_SUSRACK_X_DQ
1 Hz comb improved Jul 14-21 [Attached: example 1], then gets slightly worse on subsequent days
2 Hz comb mostly unchanged, slightly improved in the 80-100 Hz range Aug 6-18
H1:PEM-EY_MAG_EBAY_SUSRACK_Y_DQ
1 Hz comb improved Jul 14-21, mostly unchanged otherwise
2 Hz comb improved Jul 14-21 [Attached: example 2], slightly improved in the 80-100 Hz range Aug 6-18. [Attached: example 3]
H1:PEM-EY_MAG_EBAY_SUSRACK_Z_DQ
1 Hz comb improved Jul 14-21, improved again Aug 6-18
2 Hz comb appears at first glance to get worse Jul 14-21. However, this doesn't look like our near-2Hz comb. It looks like part of a 1Hz comb viewed at low harmonics.
