Following up on the possible cause of the 2.6 peak (71205, 71092, 71005), note that CHARD_Y and DHARD_Y have large peaks at 2.6 Hz, that contribute to most of the loop RMS. Looking at the control filters, there are dips at 2.6 Hz, probably due a mild plant inversion in the design. Both CHARD_Y and DHARD_Y plant models have broad peaks at about 2.6 Hz. So it is possible that the 2.6 Hz peak in CHARD and DHARD is due to a poorly controlled plant resonance. A new OLG measurement of the two loops at 60 W would be useful.

We did some tests by changing the loop gains by a few dbs up and down and did not see a significant change in the 2.6 Hz peak. This is consistent with the origin being a plant resonance.

Additionally, CHARD_Y has a large peak at 1.06 Hz. looking at the model of the OLG, it looks like thsi could be a point of marginal stabiility where the phase approaches 180 degrees. It's possible that we are seeing some large gain peaking there. Again, a measurement of the OLG would be helpful. A similar peak is at 1.35 Hz for DHARD_Y