Andrew got hopping mad about laser mode hopping, so we decided to sweep the laser control voltage and look for stable regions.  I wrote a script to automatically step through the laser slow control voltage while measuring the DC power on the precavity transimpedence photodetector.  It's on acromag2 under /home/controls/CTNWS/computing/scripts/LaserDCPowerMeasurement.py.  First block one laser's path, say the South, to the transimpedence photodetector, and then run

python LaserDCPowerMeasurement.py North 10.0

to slowly vary the North slow laser control voltage from whatever it currently is to 10.0 volts

The South path has a fairly linear dependence of slow control voltage and output laser power.  (The more negative the transimpedence photodetector voltage, the more laser power was output.  Also, increasing the slow control voltage decreases the laser temperature.)  Basically this plot is telling us that at the highest laser temperature we get the most power output.  We also see some junk at 0 volts, and nonlinearities at -7.5 volts, 3.5 volts, and 7.5 volts.

The North path is also linear, with weird junk around 0.0 volts.  There's also a strange leveling-off at 2.5 volts before a sharp decline.  The North Laser DC power plot seems noisier, but this is just an artifact of the ADC. 

The range of the North path laser power output is 0.035 volts, while the range of the South path is 0.35 volts, 10 times greater.  This could be related to the laser power caps: the South laser can output up to 2 watts, while the North is limited to 0.5 watts. 

The next step is to choose a slow control voltage region from these plots, find the precavity beatnote, lock each cavity, find the transmission beatnote, demodulate it with a PLL, and take a preliminary cavity length noise measurement. 

Attached is the data from shown in these plots.