jeffrey.kissel@LIGO.ORG - posted 09:36, Friday 02 October 2026 - last comment - 10:07, Friday 02 October 2026(92163)
SPI L IFO Glitches/Buzzes/Has Dincontinuities when IMC Loses Lock
J. Kissel More operational checks: What happens to the SPI signal when the IMC loses lock? The MEAS IFO glitches and buzzes; the REF IFO remains constant. Maybe scattered light? See three attached examples of the SPI IFO MEAS and REF A PD displacement and raw phase signals versus the IMC lock state and power at MC_TRANS.
Images attached to this report
Comments related to this report
During Example 1's IMC lock loss:
- The PSL's refcav frequency stabilization servo (FSS) also glitched (according to TRANS and REFL PDs H1:PSL-FSS_TPD_DC_OUT_DQ and H1:PSL-FSS_RFPD_DC_OUT16)
- The PMC remained locked, but its transmitted power took an impulse, and only slowly recovered. The impulse response dropped the power almost 1 [W] from 105.1 [w] to 104.4 [W] (using H1:PSL-PWR_PMC_TRANS_OUT16)
- The ALS / SQZ / SPI pick-off path's input power also took an impulse. The impulse dropped the power from (39.55 [mW] or 38.9 [mW]) to (38.7 [mW] or 38.1 [mW]) in the ALS path monitor PDs (On-table Free Space PD H1:ASC-C_FIBER_EXTERNAL_DC_POWERMON or In-rack in-fiber distribution chassis PD H1:ALS-C_FIBR_INTERNAL_DC_POWERMON)
So, an IMC lock loss glitches the whole PSL seed light source in power and frequency.
It doesn't necessarily cause the SPI L IFO's phase unwrapping algorithm to jump to a different n*2*pi location on the phase unit circle, but it definitely causes intolerable amounts of inaccurate reports of table motion.
It seems like the operational windows to use the SPI is dwindling -- it at least needs the IMC locked and stable.
Images attached to this comment
One more example for good measure. In this example, the IMC took a while to re-lock. You see much more correlation of SPI Ls IFO Phase and Displacement signals gltiching with the FSS trying to reacquire. Again the PMC stayed locked the entire time, with one impulse but not contantly glitching like the FSS (so I removed the PMC channels from the trend.)
Images attached to this comment
