Null method for finding best polarisation angle?

[The_Doctor]

Just thinking ahead... Looking for nulls is easier than looking for peaks, if not using expensive test gear... (It's also safer if I have to look at a laser spot for a while to make adjustments). So would it work well if I turned a 445 nm diode test mount on its side to project into a PBS cube for best passing of light, then turned a halfwave plate to reduce the light as much as possible, and then fix the plate and rely on this result for best output later when the diode and mount are oriented upright as intended?

[m0f]

I can't see why it wouldn't work. However I normally just tune for peak power, however do so with the power set just over threshold, which works well as the polarization doesn't change with increasing power levels. I find it much easier that way, and just turn the power up after getting it all dialed in.

[The_Doctor]

Thanks. For most adjustments I use low levels too, and it could work so long as eyes aren't saturated. Even then, watching for the angle each side of optimum at which some distinct change is seen, clean adjustment may be had by bisecting that larger angle. Even so, when testing the detune code in a synth I'm coding (when not working toward laser building, anyway..), I found that looking for the drop in sound at beat cancel points was extremely easy in comparison to detecting the strongest point of beat reinforcement. I'll try all these things once I get half-wave plates though. I'll lay odds on the null-thing, because the relative change in null-watching is much larger compared to the absolute change than it is when watching peaks, regardless of fixed power. What I'm not sure of, is whether for some weird reason the angle of max pass via the plate is not accurately perpendicular to the angle for lowest pass. I assume it would be, though, if incident light is 'normal'. But if there is any deviation on incident angle, then I'm not so sure about this.

[danielbriggs]

I can't see why it wouldn't work. However I normally just tune for peak power, however do so with the power set just over threshold, which works well as the polarization doesn't change with increasing power levels. I find it much easier that way, and just turn the power up after getting it all dialed in.

I am just being incredibly pernickity‎, but if working with very wavelength sensitive optics (as I was when using a razor edge bandpass filter [5-95% transition in 8nm], tune at full current, as 445nm diodes are especially sensitive to wavelength shift with injection current. Low injection currents = shorter wavelengths. Just something to bear in mind.

Doc... if combining 2x 445nm diodes for example, you could angle tune the 1/2WP's by directly monitoring the waste, without rotating the diodes. You will however have to have the cube mounted so that all 4 faces are accessibly, e.g. you've not adhered the cube to a brass mount for example.

e.g.

[The_Doctor]

Persnickety. (Oh how I love to be persnickety, at times.)

You're right though, monitoring the waste through the other path is the way to go. Solves everything. I can do that with everything else fixed in place.

Re current, point taken. Also, with knife-edging the current affects the total originating beam width, so I'd prefer to adjust with full current in that situation. (Especially if combing that with a PBS as I may do sometime).

[Eidetic]

It's a cosine function I think, so the sensitivity to angle adjustments at the top and bottom of the curve are low. I only tune for a minimum when super-accuracy is needed, for example when setting the polarization for a display when contrast is really important.