Lee came into the lab today and discovered that the TEC controller's range on the integrator is very small, which is why I was having trouble driving the laser frequency. If you center on this range (~10.2 kOhm) you can then drive a signal into the back of the TEC so that there are constant fringes on the Mach-Zehnder read out. I did this via the moku (15 mV @ .1 Hz, corresponds to roughly 1 fringe). "2.png" is this first attempt, red is the readout PD, blue is the driving signal. "3.png" is then hooking up a PD farther upstream to use as a power reference instead of the driving signal (see also cryolab_frequency_noise_measurement_pdf.pdf, the lighter color laser path in that is used). "4.png" is then using the moku's math channels to do (readout PD)/(power reference) to give the fringes at a constant offset. At this point its very easy to align to maximize these fringes. "7.png" shows the improved visibility. "5.png" shows the moku set up used. There are some shenanigans with the input ranges when using an oscilloscope in multi-instrument mode where you either have it on 400mV setting and its saturated, or its on 4V setting and it auto attenuates the signal by 20dB. So the digital filter box is undoing that 20dB attenuation. For unequal powers, I believe the formula is V =(2*sqrt(34*99)/(99+34))=87%, so not perfect but a vast improvement.