Laser Projector (Red, Green, Purple) "Laser 3D Party Light" from Spencer's

[DogP]

I have one question how is this even visible at 4.9mw so that would be 2mw lets say 635nm 1 mw 532nm 2mw405nm

or is a large company like Spenser gifts selling IIIb projectors and saying screw the regs

The sticker on the front says MAXIMUM OUTPUT <5mW, CLASS IIIa LASER PRODUCT. I'm only projecting it about 10 ft. in front of me on my bedroom wall... and yes, it looks much better in the dark (though it's plenty visible with the lights on).

DogP

[james]

Weird. The 405nM usually shows up on a CCD way better than we humans can see it. That one looks weak. Can you see the violet?

Use a roll of white paper with a bleaching agent in it as a screen and you'll get a nice fluorescence in the blue region of the spectrum. It works way better!

James.

[absolom7691]

If you want to cheat and make that violet more of a blueish color, try projecting onto a white sheet or white paper, basically, something that has detergent in it. Detergent fluoresces blue. It will be MUCH brighter.

[absolom7691]

Lol, just read James's post after I posted :P

[mixedgas]

OK, the X-Y speeds match well based on the diagonal lines.

The green dPSS has the usual long lag in turn on, evidenced by the yellow circle not being even. DOG-P looks good.

Corners are decent. Circles show some linearity error and velocity differences that are not showing up in the straight line motion. May just be the dac resolution is 8 bit.

Some green stuff that should be blanked is not, I dont know if thats a small analog response in the green channel ripple, or just a change in the DPSS crystal temp.

Beef up the power suppy by adding some big lytic caps on the galvo neg and pos rails. Try a few thousands of microfarads. I think the wavy-ness may be 60 hz ripple. If its not ripple, the amps are not "STIFF" on the feedback loop or the command signal is filtered PWM, instead of a dac.

So I made a few quick greyscale frames to try it. It may be a 7 color ttl system. But the green trails are causing me doubts.

I can't wait to see this on the real test frame set.

Stop on a dime is a impulse response test, since I dont have the good stuff at work.

I'm impressed for 175$, some one put some work into this unit.

Try these:

[DogP]

Weird. The 405nM usually shows up on a CCD way better than we humans can see it. That one looks weak. Can you see the violet?

It is a little bit hard to see, mostly because humans don't see it very well... the reason it didn't show up well in the pics is because I was projecting it onto my tan painted wall. I pulled out my silver screen for this test, so it should be much easier to see.

Beef up the power suppy by adding some big lytic caps on the galvo neg and pos rails. Try a few thousands of microfarads. I think the wavy-ness may be 60 hz ripple. If its not ripple, the amps are not "STIFF" on the feedback loop or the command signal is filtered PWM, instead of a dac.

Will do... gotta find a good place to connect them.

Try these:

See Below... same settings as before. Of course my camera isn't the best, especially in low light... and I should probably have a tripod, but hopefully it's good enough:







DogP

[DogP]

Oh, and nothing too exciting, but I wanted to see the chips below the laser driver board, so I desoldered it and snapped a few pics. They're attached just for completeness. Unfortunately, it looks like the DAC is an unmarked IC... the other chip is a quad JFET op amp.





DogP

[DogP]

So, I was doing a little bit of searching, and came across some Chinese galvos... some of them look very similar to what are in this, so that may give a clue to what their specs should be. Trying to stay consistent, I looked for a single seller that sold a bunch of different models, and I think these are them: http://www.ebay.com/itm/20Kpps-laser...item45fcfe9d1a . They claim:

Deflection angle 	Voltage 	Speed@Mirrors size
20º Optical Deflection 	+/-15V 	20Kpps @ 7*11*0.6
15º Optical Deflection 	+/-15V 	23Kpps @ 7*11*0.6
8º Optical Deflection 	+/-15V 	31Kpps @ 7*11*0.6

They also sell cheaper galvos: http://www.ebay.com/itm/Budget-Laser...item45fd206e77 , which are skinnier, so I don't think that's them, and their more expensive ones: http://www.ebay.com/itm/30Kpps-Laser...item45f0d09207 are larger and take +/- 24V.

Of course the driver board looks very similar to the driver board included with the cheap galvos, so I'm guessing the performance could easily be limited by that. But, it looks like I know what the function of the pots are now.

Speaking of which... I did try tweaking some of the pots. In general, it looks like the factory did a good job of calibrating it, so I put them all back to their original positions.

I also spent some time playing around with software though. I was able to get the BMP converter working, but it outputs ILDA Format 1. So, I opened it in Monkey Tools, and saved it, which output Format 0. I'd like to mess with some of the other software to figure out some optimizations and stuff.

@James: Is there any documentation for LaserBoy?

You may have noticed in the images that violet was a little to the left... so I tweaked it and got it converged nicely. I tried adding caps to the +/-15V rails on the XY board, but it didn't make any noticable difference. Looking at the ripple on the scope, it looked like about 600mVp-p on the -15V, 200mVp-p on +15. It's powered by a switching power supply though, so there wasn't any noticable 60Hz ripple, and most drooping looked to be caused by changes in load (very repetitive pattern, corresponding to the display). Adding 1000uF halved the ripple on both rails, 2000uF didn't improve it much more. But like I said, visually, I saw no difference.

I have noticed that it looks a lot better now though, and I think a lot of the problem was mechanical (vibration from the fan and galvos). Before, I was just setting the projector on whatever was near (spool of wire, or box, on my carpet floor)... now I've got the projector on my sturdy desk. See the attached image.


I think the last thing I have to talk about now is the rate. It looks like it's actually running at a slower rate than entered in the file (i.e. speed "12" is NOT 12KPPS). On the scope I looked at the 103TEST file Steve posted, and at speed 5, it took 37.8ms per frame. That's about 2725PPS. Speed 10 was about 4813PPS, and Speed 20 was about 7863PPS. What I did notice was that there's a constant dead time during each loop (~4.8ms for 103TEST). What I determined is that it's probably reloading the file, so it's actually drawing at a rate of about 3100PPS at speed 5, 6200PPS at speed 10, and 12400 at speed 20, and then not drawing for a short time while reloading. The ILDA 12K test file had a slightly longer file load delay (~5.6ms), and also a slightly higher rate (about 3250PPS per 5 speed steps).

Also, the repeat parameter is probably only 8 bits... 250 takes twice as long as 125, but 250 seems to take the same amount of time as 500 or 999. I didn't feel like counting, but I'm guessing it maxes out at 255.

DogP

[mixedgas]

How the system handles color is just bizzare. Its a 7 color system for sure. Is the Green laser just so strong compared to the others that we do not see the red and blue modulation over the green?

See your email inbox.

Hit the LCD outputs with the scope. See if they just coded a "one size fits all" or jumper selectable piece of code.

That is one of the Chinese fast 8051 clones, but who knows on the DAC.



Steve

[DogP]

>Is the Green laser just so strong compared to the others that we do not see the red and blue modulation over the green?
It's my camera... it actually looks white on the screen, but my camera makes it look green. I tried tweaking the white balance, but couldn't get it any better.

>See your email inbox.
I don't see anything yet... maybe the ISP blocked it? Or maybe it's still coming...

>Hit the LCD outputs with the scope. See if they just coded a "one size fits all" or jumper selectable piece of code.
Yep, I'm gonna take a quick look at them, and if they look active, I'll hook up my logic analyzer and figure out the pinout.

>That is one of the Chinese fast 8051 clones, but who knows on the DAC.
If you have a test file that steps one bit at a time for something like 16 bits, I should be able to at least figure out the resolution of the DAC by counting the number of bit changes actually output by looking at it on the scope.

DogP