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

[tribble]

The worst case is connecting directly to the QFP pins, which wouldn't be too bad... but I'm sure there are caps to AC couple the signal, which I can either connect the wires directly to those pads, or short those caps and connect further downstream.

Do you think it would be easier to replace the caps with jumpers? Then maybe you could use some of the existing wiring? Or is there other junk downstream?

[tribble]

Well, I finally bit the bullet and broke the green laser free of its glue home to re-align it. I was pleasantly surprised to find there was glue only along the front and back edges of the case, not all the way around and not any underneath. I let it warm up a bit and it came off with a little coaxing from a carpet knife blade. There is a small rubber pad under the green laser heatsink cube, perhaps to provide some give when adjusting the screws. I was a bit dismayed to find there are only two mounting screws holding the green, but at least one of them is a slot to allow for easy alignment in one axis. The fixed screw is nylon; it will not accept much torque, but you can tighten it and it will still give a little as you adjust that one axis. The screw in the longer slot is steel with a plastic bushing to allow for some adjustment as the screw is tightened.

The lack of a third screw means alignment in the other axis (tilt forward/back from the mount perspective; left-right on the screen perspective) is very much hit-or-miss. I needed some additional "backward tilt," so I ended up putting a piece of tissue paper (measured 0.001") under the case just in front of the screws (where 'front' is 'the side of laser emission'). After tightening everything down, the green and red are really well aligned, at least for the moment. I'll worry about the blue when I swap out the 405 for a 450 that should be here in a few days.

I notice the dichros are held by a single screw from below, but the mounts look really fragile and I think I would not like to try breaking them free and moving them. I'll use DogP's method for blue alignment if at all possible, just moving the diode in its mounting until things are aligned.

When I move up to a better projector, I am definitely going to get nice adjustable dichro mounts!

[DogP]

Do you think it would be easier to replace the caps with jumpers? Then maybe you could use some of the existing wiring? Or is there other junk downstream?

There are audio amplifiers that go between the DAC and the wiring, so if I removed the amps, I could short the input to output on them, and short the caps, and use the wiring... but that's probably more difficult than it should need to be.

Well, I finally bit the bullet and broke the green laser free of its glue home to re-align it. I was pleasantly surprised to find there was glue only along the front and back edges of the case, not all the way around and not any underneath

Yep, that's exactly how mine was. Did the heat seem to help any, or did it pretty much just break free from the knife?

The lack of a third screw means alignment in the other axis (tilt forward/back from the mount perspective; left-right on the screen perspective) is very much hit-or-miss. I needed some additional "backward tilt," so I ended up putting a piece of tissue paper (measured 0.001") under the case just in front of the screws (where 'front' is 'the side of laser emission'). After tightening everything down, the green and red are really well aligned, at least for the moment. I'll worry about the blue when I swap out the 405 for a 450 that should be here in a few days.

That's exactly what I did... except used a piece of receipt paper to "shim" it. BTW, which blue laser diode did you get? One from ebay, or somewhere else?

I notice the dichros are held by a single screw from below, but the mounts look really fragile and I think I would not like to try breaking them free and moving them. I'll use DogP's method for blue alignment if at all possible, just moving the diode in its mounting until things are aligned.

Yeah, I was afraid to move the dichros because I figured the holder would flex and the dichro would chip/crack if I put any mechanical stress on it. I thought about sliding a razor blade between the metal and the glue, but since I was moving the laser anyway, that seemed like the better way.

Have you tried removing the blue laser yet?

DogP

[tribble]

Did the heat seem to help any, or did it pretty much just break free from the knife?

Dunno; never tried without the heat. The glue was still fairly hard, but had just a little flex to it. Made prying it up with the knife easier.

which blue laser diode did you get? One from ebay, or somewhere else?

I ordered a bare Osram PL450B from DTR along with one in a module I want for a single-color spirograph projector project I am resurrecting. I have no idea what's in the Spencer projector now, but I'm assuming that this low-power single-mode diode can be made to work OK one way or another. I have not yet tried removing the existing blue laser... did it have any markings? I don't recall any from the pix you posted of your diode exchange. Which blue diode did you use? The PL450B should start to lase around 25-30ma so I'm hoping to not have to do any resistor surgery.

I'm also curious about the details of the red diode, but am not sure how to identify its wavelength or power. I suppose can markings are a long shot. If the blue and the green together completely overpower the red, I may be in the market for a small single-mode red to bring balance to (the force... sorry... couldn't not say it.) the colors.

ON another front entirely... I am thinking about doing some scanner tuning... some of the hypocycloids I'm drawing have elliptical centers that should be circular, which I gather could be mismatched scanner tuning. I've looked up the chinese laser ebay listings you posted earlier and will read up on tuning to see if things can be improved. If anyone is interested I can probably post a photo of what I'm talking about in terms of "non-circularity."

Did you ever get some caps on the PS rails? Did they help corner definition? I notice the wall wart supplied with the unit barely meets the capacity requirements of the galvo amp, and it has to run the entire logic board as well. Might be easier just to hook up a fatter PS externally, but +- 15 is not easy to find in the junk bin.

[DogP]

I ordered a bare Osram PL450B from DTR along with one in a module I want for a single-color spirograph projector project I am resurrecting. I have no idea what's in the Spencer projector now, but I'm assuming that this low-power single-mode diode can be made to work OK one way or another. I have not yet tried removing the existing blue laser... did it have any markings? I don't recall any from the pix you posted of your diode exchange. Which blue diode did you use? The PL450B should start to lase around 25-30ma so I'm hoping to not have to do any resistor surgery.

The PL450B may work electrically, but that's a 3.8mm diode, correct? The one in this projector is a 5.6mm diode... this is the one I got: http://www.ebay.com/itm/121002026653 . You'd probably need a new housing for the PL450B, since it wouldn't sit correctly in the stock mount, and wouldn't get proper heatsinking.

I'm also curious about the details of the red diode, but am not sure how to identify its wavelength or power. I suppose can markings are a long shot. If the blue and the green together completely overpower the red, I may be in the market for a small single-mode red to bring balance to (the force... sorry... couldn't not say it.) the colors.

I never pulled the red diode out, but I don't remember there being any markings on the violet diode. I was able to get the colors balanced pretty well just by tweaking the laser driver pots. I also think with the large heatsinking that you could lower the resistor values and turn up the power quite a bit before damaging those diodes (they don't even get warm at their stock power levels).

ON another front entirely... I am thinking about doing some scanner tuning... some of the hypocycloids I'm drawing have elliptical centers that should be circular, which I gather could be mismatched scanner tuning. I've looked up the chinese laser ebay listings you posted earlier and will read up on tuning to see if things can be improved. If anyone is interested I can probably post a photo of what I'm talking about in terms of "non-circularity."

Mine is the same way... when the circle is inside the square, it's elongated and not continuous (pic below). I plan to tune mine once I get my sound card DAC built. I've done some reading on tuning, and it needs to be done at your target speed... so before messing with it, I'd figure out what your target speed is, and maybe look into why the scan angle doesn't really make a difference. The target speed should be closely related to the scan angle that you're running at, but I'm guessing something in the DAC/amp circuit is reducing performance. An oscilloscope should help track it down pretty quickly. This document looks like it covers a lot of the tuning stuff, and is written for the beginner, if you haven't already gone through it: http://lasershowparts.com/files/Arti...20Tutorial.pdf

Did you ever get some caps on the PS rails? Did they help corner definition? I notice the wall wart supplied with the unit barely meets the capacity requirements of the galvo amp, and it has to run the entire logic board as well. Might be easier just to hook up a fatter PS externally, but +- 15 is not easy to find in the junk bin.

I noticed major ripple on the old model... the new model looks good on the straight lines and corners (where the old one showed problems). I did try putting caps on the old one, but it didn't help a whole lot. Your PS is +15V@1A, -15V@0.5A, right? I'm pretty sure that's what mine is (I'm not at it right now to check).

DogP

[tribble]

The PL450B may work electrically, but that's a 3.8mm diode, correct? The one in this projector is a 5.6mm diode... this is the one I got: http://www.ebay.com/itm/121002026653 . You'd probably need a new housing for the PL450B, since it wouldn't sit correctly in the stock mount, and wouldn't get proper heatsinking.

Hmm, well, time to break out the lathe. :-) I will see what I can come up with once I have all the parts in front of me.

I've done some reading on tuning, and it needs to be done at your target speed... so before messing with it, I'd figure out what your target speed is, and maybe look into why the scan angle doesn't really make a difference.

When I was trying to find the optimal settings in LB for the optimize function, I did discover a few things in this regard. First, if you have a long straight line and a sharp turn at the other end, you can get the scanners moving to the point where they will overshoot at large scan angles and not overshoot at smaller scan angles. So, it is possible to feed it a frame where scan angle does impact the unit's ability to faithfully reproduce the input. I guess most of the stock images are not very aggressive in that regard. I suspect that near that limit is where the maximum scanning speed is for the factory tuning. By adding more points along the line you can slow the scanners down so that they don't overshoot at any scan angle. I haven't examined the ILDA test pattern in an editor, but I assume there are some lines without 'intermediate' points along their length to get the scanners moving to their maximum speed so you can tune for that. I guess some of this is a compromise; if you adjust angle and speed for sharp reproduction for images with lots of extreme throws and sharp angles, you might be giving up performance on images with smaller smooth curves that don't have the angular displacement. I will read the document you linked to and see if some of this starts to gel; thanks for the link!

I think an corollary to all of this tuning effort is that, in the program file, I'm always going to use the same speed setting and just use the software generating the ild files to add or remove points as required to get the details rendered correctly.

I also noticed if you specify a speed higher than 25 in the program file, 'speckles' start to appear in the image. I suspect this is aliasing or discarding of some frame data which leads to little hiccups in the scanner travel, but I could be wrong. 25 seems to be the maximum usable setting in the program file, so I figure that's the setting to use when trying to do tuning. Why would one want to tune for a slower speed? I was going to try to reverse the clock rate of the DAC by playing some long frames or frame sets with known numbers of points and seeing how long it takes. I know you did some timing tests; I will go back and look those up first, but were those on the V2 hardware?

I am planning to tune for the maximum speed I can attain at a reasonably large scan angle, This is a low-powered indoor device, after all, and I'm never going to be 50' away from my projection surface... more like 10-15 if I'm lucky. 8 degrees is just not going to make a very big picture. I notice that one axis runs out of 'zoom' before the other when turning up the 'focus' knob; I will probably tune at or below that point.

The specs on my PS are +15 at 1A and -15 at 500mA.

[james]

I should also add to my previous post about generating math figures that in the [Tab] menu, option "c rendered arc step in degrees" is used in the calculations. So, if you want smoother curves, lower this number before you render!

Also note that this is exactly the kind of art that can get really screwed up if you minimize points when LaserBoy removes points that are less than or equal to the "1 insignificant angle in degrees".

James.

[james]

ON another front entirely... I am thinking about doing some scanner tuning... some of the hypocycloids I'm drawing have elliptical centers that should be circular, which I gather could be mismatched scanner tuning. I've looked up the chinese laser ebay listings you posted earlier and will read up on tuning to see if things can be improved. If anyone is interested I can probably post a photo of what I'm talking about in terms of "non-circularity."

Sometimes the problem is as simple as slightly different input sensitivity on the X and Y scanner driver amps. You will get ellipses instead of perfect circles.

If the scanners are out of tune, you can usually see this in the effect of sloppy, muli-traced lines. Well tuned scanners will always go almost exactly where the input voltage dictates. Out of tune scanners will trace a set of lines and not quite make it to the proper 2D locations, resulting in the appearance of multiple lines.

[tribble]

Well, I loaded up the lasermedia test pattern and verified I have different scanner speeds on the top-right and lower-left diagonal lines. Guess I'm going to give that tuning a whirl. I'll try to take some before-and-after pictures.

Sadly, the scanner amp does not appear to have a high-frequency damping pot, or I haven't identified it. DogP dug up some images of very similar scanner boards, but the one in the spencer projector is slightly different. I don't want to screw things up, but... nothing ventured, nothing gained.

[DogP]

I also noticed if you specify a speed higher than 25 in the program file, 'speckles' start to appear in the image. I suspect this is aliasing or discarding of some frame data which leads to little hiccups in the scanner travel, but I could be wrong. 25 seems to be the maximum usable setting in the program file, so I figure that's the setting to use when trying to do tuning. Why would one want to tune for a slower speed? I was going to try to reverse the clock rate of the DAC by playing some long frames or frame sets with known numbers of points and seeing how long it takes. I know you did some timing tests; I will go back and look those up first, but were those on the V2 hardware?

Hmm, I don't remember speckles, but it's possible (you can see the ilda test pattern on mine at speeds 5 to 40 in post 65). I believe you want to tune for your target speed, so if you're planning on running everything at speed 15, then you wouldn't want to tune for speed 25. If you just make all your files play at speed 25, then you should be fine, except you've probably noticed that the image quality goes down at higher speeds (even if the anglular speed is the same). That document I linked to talks about it a little bit at the end, under the section "A Note about Scan Speed" - "If you tune at 30K, and then display at 40K, the circle will end up inside the square. If you tune at 40K and then display at 30K, the circle will end up outside the square." I did some sanity checks on the DAC speeds with the o'scope, though I don't think I did any formal recording of the speeds on the V2.

Sadly, the scanner amp does not appear to have a high-frequency damping pot, or I haven't identified it. DogP dug up some images of very similar scanner boards, but the one in the spencer projector is slightly different. I don't want to screw things up, but... nothing ventured, nothing gained.

Yep, I don't think this has a HF damping pot, though you may be able to identify a fixed resistor that performs that function, and possibly add the pot. Fixed high frequency damping could be part of the reason for the reduced performance at higher DAC speeds (I believe that'd essentially be a low pass filter). If you follow the document I linked to, having HF damping fixed may force you to find the optimal scan speed, rather than tuning for the scan speed you want (the document explains fixing the scan speed, and having SG, LFD, and HFD as variables, but if you have HFD fixed, then the other three would probably become the variables to optimize).

DogP