Since the test stand upgrade, the side OSEM sensor channels (X1:SUS-QUAD_M0_OSEMINF_SD_IN1 and X1:SUS-QUAD_R0_OSEMINF_SD_IN1) have been very noisy, to the point where enabling the SD DOF control loops would trip the watchdog because of HF signals amplified by the upward slope in the software output filter.

The noise looked vaguely digital in dataviewer but it was hard to tell due to the poor time resolution, and until today we hadn't probed in more detail. However doing a triggered time response measurement in DTT revealed that an otherwise plausible looking signal is being chopped by a rectangular window at 16 Hz with a 1/8 duty cycle (i.e., 1/128 s on out of 1/16 s). See the attached plot, which shows X1:SUS-QUAD_R0_OSEMINF_SD_IN1 (which exhibits the problem) against X1:SUS-QUAD_R0_OSEMINF_F1_IN1 (which doesn't).