Any help for figuring out how to tune a WonSung galvo set?

[jsibre]

Update - I've got things looking a bit better than the pictures above now, but still have the fundamental issue that the circle is off center.

It's a bit less jittery now. I'll try to get some new pictures this evening.

I think I may have to re-tune (again - I'm getting a lot of practice!). In order to get to the point I'm at presently, I've had to adjust the IS and PS pots. I'm having a hard time deciding if the LIN pot actually does anything. But by messing with the IS and PS pots, I'd venture a guess that the +/- 32768 is probably not properly adjusted within the limits of the galvos.

I'm thinking my process should look like:

1. While outputting the signal for a square that reveals the limits of my DAC, double-check that output from my amplifier are actually at +/- 5vdc and centered properly (this is what gets fed to the galvo drivers). I've done this before with my 'scope, but I think I should double-check this first.
2. Hook that amp output back up to the galvo drivers' input, and switch to a grid pattern, and ensure the grid looks right (not bumping into the limits of the galvos or the output range of the drivers, etc.). I expect I may have to adjust IS to get the size maximized, and PS to make sure it's centered in the range of the galvos. (Does this sound sane??)
3. From there, I'll just work my way down Ben & Buffo's 'ladder' a few steps (HFD, LFD, SG) and then work my way back up again, using the three test patterns.

Currently I've been doing this at 20kpps. Any reason I should go down to 12k or [try to go] up to 30k? I know that's what the one test pattern is designed for, but honestly, I'm not sure that it matters, as long as I tune it at the speed I plan to display IDLA files at. (Further, since I'm not really tinkering with this to 'display ILDA files', but rather to tinker with controlling the system with the Arduino more 'raw'ly, the tuning process is really more a matter of me trying to make sure the drivers are adjusted as optimally as possible. I'm not positive that 'optimal' really even means anything in this context, without nailing down some other parameters, such as PPS, or max Rads/sec, etc.)

Also, am I correct in my idea of matching the DAC range to the galvo/driver limits? Or is it more important to take the 'increase the IS until the circle stops getting bigger' approach, regardless of where i end up with regard to the limits (well, I guess I need to be 'within' the limits, but maybe don't need to be 'at' the limits)?

And is there anything else I should do in the midst of this process?

Thanks!

I'm not exactly sure where I'm gonna take this next... Waiting on my RGB laser to show up, still. Hopefully once that arrives, I'll have much better blanking response. Right now, it's pretty crappy. I'm also approaching the limits of what this microprocessor can do in real time while also trying to control the galvos. And it doesn't have enough RAM to hold much in the way of data points. Just holding the three ILDA test files in memory at the same time is a significant burden. About 10k at the moment, I think, but that'll increase quite a bit once I include the color info (right now, I'm only paying attention to X, Y, BLANK). Including the color bit may also impact my ability to draw at 20kpps, because that'll mean two (or three, if I try to honor the blanking signal in addition to the color info) more outputs that have to be managed.

Looking forward to playing with that Not sure if I'll try to PWM the RGB or not. I guess I'll probably try it eventually...

Oh, any recommendations on good safety eyewear for a 500mw RGB laser?

And... Thanks again

[jsibre]

Hi,
I didn't have the time to do it yesterday... So, better late than never.

Quadrature:


ILDA Test Pattern:


LaserMedia Test Pattern


I didn't notice the overshoots in the first image until after I took the picture, because they only showed up after fine tuning on the latter two.

So, with that in mind, a question: Are those significant?
Back story: After much mucking AFTER seeing the images (and before uploading), I realized I couldn't seem to get rid of them, AND have the circle touch [one of] the sides of the square. As I keep going to higher and higher gains (and LFD, and HFD) I eventually reach a point where the HFD is so high that my straight lines are wavy. (which you can already see, in a minor way, in the LMTP image above).

Thanks!