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

[tribble]

On another note... it was foggy tonight, so I had some fun with the beams, shining out my back window into the yard... here are a few cool shots I took:
Attachment 36492

That one is my favorite! Nice work!

Now that the spencerizer is giving me the images I expect when I export them out of LB, I'm going to work a little more on "beam shows" for the spencer projector.

By The Way... there is at least one other projector being sold based on this showcard! There was a recent update to a thread in the Help section about another one, the beamz hyperion bleu. Can't find it for sale anywhere; might be an older unit.

[tribble]

looks great here. pardon the low quality cell phone photo.

Attachment 36481

Cool! Is that the .ild or the .spn?

[DogP]

By The Way... there is at least one other projector being sold based on this showcard! There was a recent update to a thread in the Help section about another one, the beamz hyperion bleu. Can't find it for sale anywhere; might be an older unit.

Is that this one: http://www.tronios.com/en/lighting/l...ue-dmx-sd.html ? I think there are quite a few based on that... I'm pretty sure that's the same as the old model Spencer's projector. The new one AFAICT doesn't support an LCD, buttons, etc... the old one had headers for all that, and BazookaBo took pics of the inside of his http://www.chinavasion.com/china/who...aser_Projector , in post 106, which was the same as the old Spencer's one. If you look on ebay, there seem to be a lot of them as well.

But the new Spencer's one performs WAY better than the old one in my experience (and is quite a bit cheaper than any others I could find online).

DogP

[james]

I'm just waiting for DogP to get his LaserBoy DAC put together!

Then he can do a comparison of that to the internal DAC.

HEY! Apparently you have a bunch of op-amps laying around. What are you waiting for?

[DogP]

I'm just waiting for DogP to get his LaserBoy DAC put together!

Then he can do a comparison of that to the internal DAC.

HEY! Apparently you have a bunch of op-amps laying around. What are you waiting for?

I'm getting there. I'd like to reuse as much original hardware as possible, and make it switchable between USB and SD card... so I'm trying to make an elegant tap into the signal chain. But that takes some reverse engineering of the original design. Plus, the amp board will be completely reused, so before I start messing around too much, I'd like to tweak that, so I can easily compare before/after.

DogP

[james]

Go to a used computer and parts shop and get yourself an old printer A/B switch. It has a rotary switch with 25 individual contacts on it! You also get some spare DB25 ports!

I would suggest putting a DB25 port on the projector and using a printer extension cable rather than simply putting the USB device inside the projector. USB cables are only supposed to be 2 meters. A printer cable can be 25 feet or more.

That way you can use any ILDA standard output DAC.

In the future you might want to get a sound card with 8 channels of output rather than 6. That way you have room in your waves for stereo audio tracks!

Besides that, you will probably also want to replace all the diode drivers to achieve analog modulation for more than 7 colors. The red and blue will be fairly easy. The green DPSS will be a challenge.

[DogP]

Go to a used computer and parts shop and get yourself an old printer A/B switch. It has a rotary switch with 25 individual contacts on it! You also get some spare DB25 ports!

I would suggest putting a DB25 port on the projector and using a printer extension cable rather than simply putting the USB device inside the projector. USB cables are only supposed to be 2 meters. A printer cable can be 25 feet or more.

I said I want this to be small. And actually, the USB cable length limit is 5 meters (more with an active extension). And I'll probably do something a little more fancy, like USB cable detection switching an analog switch or something. I got a lot of them laying around here too. :P

In the future you might want to get a sound card with 8 channels of output rather than 6. That way you have room in your waves for stereo audio tracks!

Yeah, I was looking for an 8 channel card, but I wanted it to be as small as possible, and only really need 5 channels... I'd probably never use channels 6-8, but if I need to, I can always remove the CM6206-LX (5.1CH) and swap it for the CM6206 (7.1CH)... they have the same pinout, except the 7.1CH adds the extra two channels and headphone output.

Besides that, you will probably also want to replace all the diode drivers to achieve analog modulation for more than 7 colors. The red and blue will be fairly easy. The green DPSS will be a challenge.

Is there a page that describes laser modulation well? I've never quite gotten the whole idea of how it works.

If I drive it with a square wave, I assume it'd be "dimmer", but the square wave would need to be a much higher frequency than the slew of the deflection, or it'd look like dashed lines. But I assume at some frequency, the laser doesn't respond anymore, and it just sees a voltage proportional to the duty cycle, which doesn't really help (though that frequency is probably pretty high, assuming the drive circuit can do it).

If I try to actually dim the laser by controlling the DC to it, I assume the dimmest I can go is the lasing threshold. That's about where I'm driving my blue laser, so it doesn't give me much adjustment there (of course a 1W laser would have much more room to adjust, but I don't have nor want one). And then, yeah... I assume green is much harder.

Or is there some other technique that I'm not thinking of? I'm really just looking for the basics of how it works, so I can wrap my head around it.

DogP

[mixedgas]

Dog P,


Your right about the oddity of not updating X-Y at the same time.

"Double Buffering" is what is done on more sophisticated DAC chips in internal logic so the X-Y vector updates at the same time. At 30K, you will see artifacts in the image from the lack of double buffering on fast galvos.

If you installed a faster power amp, you would not see much improvement unless you re-tuned the loop and/or increased the PSU voltages. You might see less cross-over distortion or better settling time. How many bits is the DAC again? That has something to do with observation. On a 8 bit image, things might not be so noticable.

I can see the colors updating a few instruction cycles apart, on such a simple processor. That is logical. 1/PPS = the point time. So 600 nS out of sync is not too terrible at 20K, because various interrupts should cause that to slide around a bit, smoothing the color in the image. From a sheer lack of memory, I can see why they did not lead the green signal, properly.

I do not understand what they were trying to do on the color update "steps". It may be a processor hardware or code artifact. I do not know.

As for refitting analog modulation drivers. The red and blue diodes will respond with very respectable linearity to changes in current. So the analog driver has a offset potentiometer to bias the diode just below lasing threshold. The analog signal from the controller is then added to the threshold current with adjustable scaling as needed. Above threshold most smaller diodes are pretty good for this type of drive. If your young enough to remember CRTS, the diode lasers pretty much will then behave like CRT guns.

The green however, will have thermal time constant issues, caused by heating in the lasing crystal and various things that occur in the IR pump diode that drives the whole process. Small greens like yours usually hack fairly well. If you mod it, You'll experience what is called "jellybeaning". Jellybeaning is where the turn on/turn off times are different as the animation of the image changes the average heating in the crystal. The adsorption of the IR pump laser is very dependant on how the crystal is heated and cooled. ~ 0.5 watt of pump diode focused into a 250 uM spot causes a lot of heating in a insulator like that optical crystal.

PWMing the diodes would result in very good color linearity. However you must adjust the peak current for the best white at full drive. You also would need to have a PWM update rate of 3 to 6 times per point, and at 30K, that is very fast, on the order of 100s of Khz to low Megahertz. The diodes can respond that fast, but no low cost, commercial driver, would do that.

I did PWM for laser color, long ago. However I had Acousto-Optic modulators with 7 Mhz bandwidth and a video rate digital PWM circuit. It looks great,better then analog current modulation, but it's a lot of work, clocking a A/D chip at 6 Mhz.

PWM would probably drive the Spencers Green laser without the crystal TECs bonkers. Higher end lasers have active temp control of the two crystals and the pump diode in the DPSS.

I'm still impressed at what you get for 129$.


Steve

[kecked]

Especially when you consider even the picaxe can multithread. Setup on A dac pair can easily be simultaneous via two data pins and one cs to a single shift reg with a single cs on a Master clk. dac hips tell you how right in the cut sheet. Just did this using spi. Have done it on i2c nd micro wire as well.

[DogP]

Your right about the oddity of not updating X-Y at the same time.

"Double Buffering" is what is done on more sophisticated DAC chips in internal logic so the X-Y vector updates at the same time. At 30K, you will see artifacts in the image from the lack of double buffering on fast galvos.

Yep, I've used several DACs that have had a Load DAC strobe to synchronize output... but the TLC7528 doesn't do that, so I guess that's why there's a delay.

If you installed a faster power amp, you would not see much improvement unless you re-tuned the loop and/or increased the PSU voltages. You might see less cross-over distortion or better settling time. How many bits is the DAC again?

It's 8-bit... and yeah, I decided to install the rest of the upgraded amps, so as I mod and tune, hopefully I won't run into performance problems from the low quality amps (I'm guessing I probably wouldn't have anyway though).

As for refitting analog modulation drivers. The red and blue diodes will respond with very respectable linearity to changes in current. So the analog driver has a offset potentiometer to bias the diode just below lasing threshold. The analog signal from the controller is then added to the threshold current with adjustable scaling as needed. Above threshold most smaller diodes are pretty good for this type of drive. If your young enough to remember CRTS, the diode lasers pretty much will then behave like CRT guns.

The green however, will have thermal time constant issues, caused by heating in the lasing crystal and various things that occur in the IR pump diode that drives the whole process. Small greens like yours usually hack fairly well. If you mod it, You'll experience what is called "jellybeaning". Jellybeaning is where the turn on/turn off times are different as the animation of the image changes the average heating in the crystal. The adsorption of the IR pump laser is very dependant on how the crystal is heated and cooled. ~ 0.5 watt of pump diode focused into a 250 uM spot causes a lot of heating in a insulator like that optical crystal.

PWMing the diodes would result in very good color linearity. However you must adjust the peak current for the best white at full drive. You also would need to have a PWM update rate of 3 to 6 times per point, and at 30K, that is very fast, on the order of 100s of Khz to low Megahertz. The diodes can respond that fast, but no low cost, commercial driver, would do that.

I did PWM for laser color, long ago. However I had Acousto-Optic modulators with 7 Mhz bandwidth and a video rate digital PWM circuit. It looks great,better then analog current modulation, but it's a lot of work, clocking a A/D chip at 6 Mhz.

PWM would probably drive the Spencers Green laser without the crystal TECs bonkers. Higher end lasers have active temp control of the two crystals and the pump diode in the DPSS.

Ah... thanks for the info. I looked more closely at the CN5611 datasheet, and it says the frequency of the PWM signal should be <10KHz... which is barely sufficient for on/off control, so it doesn't look like PWM brightness is gonna happen with this driver. :P

I'm still impressed at what you get for 129$.

Same here!

DogP