Ok, so it seems like way back in the mists of time, I built a dual blue 445nm using one of the kits from Dave and some of the first batch of A140 diodes that Briggs did the GB for. Also around that time, I bought a braodband PBS in the GB that P1tbull did, and that had been sitting around for a while.
SO, fast forward a few years and I'd never been quite happy with the knife edge arrangement - partly due to a slightly sloppy build by me, so with the missus away for a week I decided to upgrade the module to a PBS arrangement as one of my laser related jobs to do this week.
A 1/2 waveplate was duly ordered from Rob, and this evening I did the work.
I had to relocate the original knife edge mirror mount and turn the part the mirror fixes to 180 degrees to give me a flat face to mount the cube on. I also had to drill a small hole in the top, which is used to align with, by sticking a small hex driver in and tweaking back and forth. Simple but effective.
To get the cube square with the mount I put them on a glass plate (back of an iPhone actually...) so the bottom faces were flush, and then dabbed epoxy on the top of the join after cleaning it with alcohol.
I then mounted it, masked the transmission diode, and played with the waveplate to get optimum angle - marked witha thin Sharpy, and then took it off again, aligned, 2 dabs of epoxy and left face up in the vice to set.
Then back in the module, fired across the room to a mirror on the wall stuck with BluTack, and back again (about 5m totla distance) and did the final alignment and power test, which was pleasantly surprising. I'd read that there were quite a few losses with cube and waveplate combos, due to multiple faces, and less than perfect transmission characteristics. Power with the knife edge had been set at just over 1.9W, with the cube it was actually more, coming out at just under 2.1W
I tested the power using an FB3 projecting the Laser Media test pattern (because it has the laser on all the time) and from prior testing this gives about 95% of the power that a 5V signal does. So it's doing a bit more in the real world, but whats 100mW or so between friends!
And as its also a bit tighter, it should hopefully look brighter too.
Then remounted the blue dichro and re-aligned the whole projector. Job done! About 4 or 5 hours all together at a leisurely pace.
Next job - realign a CNI MGL-H 532 head which has developed 2 spots...
Anyway, here are some pics...