The out-of-loop PD for the PSL intensity stabilization (H1:PSL-ISS_PDA_OUT_DQ) was used to produce the fake strain channel during engineering run 3 in February. It was very clean overall, but there were a lot of low-frequency glitches. The glitches are all short and below about 55 Hz, and there seem to be more in the 10-25 Hz range. The glitches are every few seconds, not regular, but the rate was pretty much constant during the day. These glitches are still there in recent data. Robert Schofield told me that this frequency range isn't environmental but is likely to be electronic. The glitches are also seen in QPD_DX and QPD_DY.
If we can help diagnose this, don't hesitate to contact detchar@ligo.org and andrew.lundgren@ligo.org.
The attached plots are:
1. A typical day of ER3 as seen by Omicron (the replacement for the OmegaOnline web reports).
2. A similar plot for 5 minutes of recent (last weekend) data.
3. As above, but QPD_DX.
4. As above, but QPD_DY.
5. A selection of 2-second spectra, showing the instability.