Its seems like the issue with finding the beat note has to do with us having a poor idea of where the ideal operating regimes of the two lasers are relative to each other. The laser crystal offsets were set to temperatures that seemed to be a good operating point at the time but now are not so great for some reason. Also the slow offset voltages are somewhat arbitrary, being ~0.69 V for south and 3.8 V for north. It would be better if these were set at the mid point of the tuning range (±10 V) for better range when operating only from EPICS channels controls.
We would like a north temperature vs south temperature map, similar to what Koji did in 40m:3759, to find the best operating points and to remove the stab-in-the-dark guess work that has been the method till now.
The slow frequency offset EPICS channels were set back to zero and the north/south temperatures were walked back to a point where we could see the beat note again.
We see a slope of -668 MHz/Celcius in the North laser temp Vs the precavity beatnote frequency.
Edit (awade: Sun Aug 13 19:02:29 2017): note that the slow laser frequency tuning is actually positive, it was likely that we were looking at the beat note reflected about DC so the above number has the opposite slope.
From the manual (found here on the ATF wiki) the manufacture says that the temperature tuning is negative as a function of frequency and the front panel BNC is positive.
Edit (awade Mon Aug 14 16:16:35 2017): Must have been tired. Yes the gradient in the graph is negative, so plot makes sense. See my reply post to this one for details on the laser operation