Galvo distortion with Lasershark

[nicobux]

Hello,

I am having distortion issues with my scanner system. The distortion is most pronounced at higher scan speeds. The distortion is also accompanied by odd noises from the galvos, which sort of resembles a repeated scraping sound. The galvos are from Spacelas, and I have been in contact with their representatives. After sharing pictures and videos with them, they sent me this email:

"Hello ---
 
There are too much points for output, you need to cut the some lines by using a DAC card.
 
It is no problem of the galvo set yet."

If anyone has encountered such a problem and can shed some light, please do. I was prepared to simply get another scanner system but they insist it's not an issue with their hardware.

Thank you,

[Macpod]

Good evening Nico,

Galvos systems are sold with a KPPS (kilo-points per second) rating and max angle of deflection. The KPPS rating defines how many times the galvos can move from one max angle of deflection extreme to the opposite max angle of deflection. From my experience the ratings that are advertised are often higher than the galvos are actually able to perform at unless you purchase from a reputable company (i.e. Cambridge Technology).

If you are observing distortions and especially if your galvos are making odd noises (i.e. scraping noises) you are driving them at the KPPS rating higher than they are capable of being run that. This is what Spacelas is describing.

You should fix this! While over-driving the galvos will not harm the Lasershark it will (and currently is) damaging your galvos.

If you are using Jackctl, you can change the rating to something lower until the galvos stop making noises. If you are using lasershark_stdin, at the start of output you can send a line which will inform lasershark_stdin what rate to send samples at.

[nicobux]

I understand that they might be over advertised. However, even when running the idla test pattern at lower (ie half the rated speed) speeds the distortion is present. I can't really imagine that they would overstate to that extreme, but perhaps.

I have also noticed that the noise is only produced by the Y galvo. Swapping the X and Y outputs on the lasershark confirms this. I did sort of fudge the soldering job on the Y connection (overheated the pads a bit) but it still seems ok. Is there any way I could diagnose whether its an issue with a physical connection?

[Macpod]

It depends on where you buy them, I have seen sub 10kpps galvos  sold on ebay as 30kpps gavos.
You can spot these by their rectangular mirrors (mass limits quick changes in direction so generally faster galvos will have smaller mirrors and those with rounded/chopped corners).
You can also identify them by checking the wiring between the galvo amplifier and DAC (i.e. the lasershark). If the X and Y inputs only have two wires, the amplifiers aren't interpreting a bipolar differential signal which is important for position stability and to prevent distortion at higher frequencies.

Don't forget the ILDA test pattern is supposed to put a little stress on your system too.. otherwise it wouldn't be a very good test pattern

If your galvos aren't making terrible noises but you still see distortion it is possible your galvos are not tuned to be in harmony with OpenLase's default settings. Tuning can be performed in two spots...
1. The galvo amplifiers have a bunch of multi-turn pots which impact the galvo amplifier control loops.
2. If you look towards the top of some openlase executables you will see they have custom tuning attributes. playilda does not use custom settings, but the simple.c example does:
https://github.com/marcan/openlase/blob/master/examples/simple.c#L87

When you tune a system, remember that you are tuning it according to your current setup.. i.e. if you have adjusted the output size/keystoning/kpps rate in the output app, tune the ilda output, and then adjust the output size/keystone-ing again.. you will have to re-tune the system!
Also keep in mind that while you can get close, it is unlikely you will ever get a perfect ilda test pattern.
http://www.laserist.org/StandardsDocs/ILDA_TestPattern95_rev002.pdf

[nicobux]

Thanks for the informative reply.

The X and Y lines have three wires each.

Perhaps the issue is tuning. Do you have a link to any good tutorials? I have watched videos explaining the tuning process for the ILDA pattern, but I've been hesitant to touch the pots.

Do you recommend tuning exclusively to the ILDA pattern? Or would it be beneficial to tune to what I want to display (since I just need a static image).

[Macpod]

If you are just going to display a static image I would just tune to this and only after you determine the size of the image you want to display as well as the galvo speed (kpps setting). If you are using openlase, definitely tweak these params first. If they do not work, disable tweaks and then use the pots.

I cannot say I have a difinitive guide to tuning galvos and tuning will vary a little according to how many pots you have (higher quality galvos will have more adjustments). You might want to scan through the photonlexicon.com forums. In fact check that forum out regardless, there are some very knowledgeable folks over there

I would also reduce the complexity of this static image and ensure the paths are optimized. By this I mean I would reccomend drawing the shape in a manner that requires the galvos to perform the least drastic direction changes (i.e. if you are drawing a box, don't draw the left side, then the right side, then the top side and finally the bottom... draw the right side, bottom side, left side, top side).