From alog46181 and alog46446 it almost seems like this Mephisto mode hop region was slowly creeping upon us since holiday (probably has been since the beginning since we ran away from 27.6 degC alog45519 which is now no longer a good region to operate). Today I finally found a setting where we can go from mode hop to mode hop and operate in the middle. We are how sitting ~0.1A away from mode hoping region. Using 2.4MHz/mA (I re-calibrated the current -> Hz with more data point) we now have +/- 240 MHz of wiggle room. Temperature doesn't appear to be affecting mode hoping as strongly as current does so I'm going to neglect that for now.

For a future reference, here's what I did:

I scanned the SHG to look at the multi mode peak (looking at transmission) and look at RF spectral analyzer to ensure that I had the beat note. I tried to go down in temperature but to compensate and keep the beat note at the same place I had to crank up the current, which I quickly ran into clamp current value (2.25A). Terry suggested I go up in temperature instead and hopefully the current wouldn't have to go so low to compensate while having a nice mode-hop free region to operate. I turned up the temperature by a degree, and sweep current left and right to see if there's any beat note. At 31.5 deg C region I was still limited by the clamp current and soon ran into another mode-hop region. I somehow skipped 32.5 deg C (or maybe I just couldn't find the beat note) and found the beat note at 33.6 deg C region (as plotted). I alternate between temperature and current to keep the beat note roughly at the same place (158 MHz +/-). I weren't patient enough to wait until the temperature really settle down before I write these frequencies down (so an uncertainty of +/- 1-2 MHz is possible).