Home built projector V2

[Theoneinformant]

Ok, so I have worked out most of the major issues with the scanners now, so I thought I would upload some pictures, see what you all think.








Breadboard was an epic ebay find, 75$!!, for inch thick 30" x 10" board. all optics are mounted on this board. Upper level holds the power supplies and the drivers. Using a 12v Meanwell PS for fans and diode drivers, Lovell Diode drivers, PT-30K scanners, assorted ebay find kinematic mounts, Newport and Thor labs, flex mounts from optlasers, diodes from DTR.

Rectangular dichroics are from OPT, round dichroics are Stanwax. aluminum bits I machined, aside from the 20mm diode module holders, which are whatever was on amazon prime.

I left plenty of room between 1st and 2nd dichroics for correction optics and room between dichroic sets and scanners for beam expander/spatial filtering prior to the scanner block.

I accidentally killed the red diode last weekend, but have a new one incoming. Scanners are mostly tuned. I still cannot get the circle completely round, I am going to run through tuning again here tomorrow. Images are pretty solid though.


Blanking/color goes funny occasionally though on the ilda test pattern, I have no idea why. Its like timing get off between the scanners and drivers. Haven't had it happen on anything other than the ilda test pattern though. I can also see the blanking move while the pattern is displaying, not sure what to do about it.

So, ideas, questions, comments? Input is appreciated!

Casey

[Bradfo69]

I understand the desire for a double bounce from your first mirror to the round dichros, as that makes your near field alignment much easier. It almost appears as though that you have a dichro on the first brass flex mounts but maybe that's just an aberration in the photo. I don't necessarily get the Z bounce after the dichros and before the scanners - unless that allowing more room for future correction optics and... taking advantage of the pre drilled breadboard holes. My initial thought on looking at this though is that you're probably getting a lot of losses through so many bounces. Have you measured how much you are losing?

[Theoneinformant]

Double dichroics in the first stage are for alignment. The z bounce is just a place holder for more optics, and to allow for easier alignment into the scanner block. I do not have a LPM yet so not sure of power out. The first dichroics are on flex mounts. It was a leftover from the first projector design, and because they are at the proper optical height for the diodes I kept them to allow for alignment.

[kecked]

Solder the diodes in place. Those connectors for diodes are a dice roll. I lost plenty....

[SaltyRobot]

I love watching home built projects come together. Keep up the good work. If you need any advise hit me up.

[dkumpula]

Thanks for posting, Casey. All those bounce mirrors will certainly make nearfield alignment a piece of cake, but I agree with Brad regarding the losses. My guess is you are losing 3-5% per bounce mirror when they are clean and significantly more when they aren't. Like yours, my optics decks aren't sealed, so I find dust and fog particles mount up over a year even with low use. If you have plenty of power, no big deal I guess.

What software are you using? That info will help us chime in regarding your optimization questions.

Keep up the good work!

-David

[Theoneinformant]

Ok, so many more problems with this thing... Blue 2nd dichro broke when I was re-aligning the projector, it was just a hair too thin for the thor labs mount and the metal on the set screw contacted the glass first so after pulling the optics deck apart and re-configuring it so that blue is a direct shot through the remaining dichros and aligning the other lasers to blue we finally had it working. Full color for the first time, for about 2 hours... and then the red diode just went LED in the middle of playing an ild file. Nothing was being touched, moved or anything. Just went dead . This is the second Oclaro 700mW 638nm diode that has died, and this time I had turned down the driver even further 950ma, so I don't know.

So now I have 2 Mitsubishi 500mW 638nm diodes incoming, another lovell driver, pbs cube and waveplate. Stanwax also had lasorbs in stock so I have one per diode on the way. Figure I will run the mitsu's even lower, closer to spec sheet mA, and run the lasorbs - hopefully no more dead red.

Alright, to answer dkumpula, the initial configuration was more of a proof of concept and ease of alignment. I have a few other projects in the works, so when I work out all of the kinks on this projector, it will be simplified, condensed and sealed up. For software, I am running LFI for .ild and using fubar2000 for playback of some .wav's I have. Shortly I will have Digisynth to play with as well.

Blue diode is a m140 @ 1A, green is a n520 @ 1A (pretty awesome beam specs for a multimode diode), and red was the oclaro @ 950mA, about to be pbsed mitsu's. at 7-750mA each in a tec'ed block.

[dkumpula]

Hey Casey, Sorry for the challenges, but just chalk them up as learning experiences. Everyone who's built has had their share.

Your red diode is quite resilient as I recall, so based on your description, I'd be inclined to think you had a bad/intermittent connection between the driver and the diode. This will kill your diode faster than you can make a "yo' mama" joke. Your leads between the drivers and the diodes strike me as long-ish, but not likely enough to introduce any significant level of capacitance.

-David

[Theoneinformant]

Leads are 4-5", no longer. Should not cause issues. I was led to believe that the oclaro's are finicky anyway, and the mitsu's were more robust

[Theoneinformant]

I suppose I will have to solder the diode leads onto the driver boards. I really don't think it was the problem. But cheap insurance, right?

On another note, I am getting some "jelly beaning" on images that the scanner is at the upper end of its speed, very pronounced, seperated dots. Almost like I added chop in from an analog console. My research says, diode temp issues, driver boards too slow, or signal issue. Any ideas?


test pattern is at max angle, so circle is off... Looks fine at 8*