Laser misfiring - any ideas?

The attached image shows results on the back of a mirror using LaserEtch by Jtech Photonics.

Note that in the second half of the etch/burn, the laser is occasionally left on in what should be whitespace. I believe the burn is not as complete in these areas as the intentional burn areas move at 240mm/min and moves between spaces are 6000mm/min.

Running these at different speeds may result in different, fewer or no errors, so I do not think it’s the gcode LaserEtch is producing. I get no communication errors in the repetier console, and no problems in the image registration.

Guesses:
–Maybe noise on the TTL line? I’m using shielded 18 gauge cable.
–Cheap driver overheating? That would make sense considering these errors occur later in the burn - how should I check that?
–Using two separate ATX power supplies, on for the MPCNC, one for the laser. Maybe it’s angry that I haven’t unified the ground between the two supplies?

Thanks for your thoughts on this!

–Matt

Could be a lot of things.

Try using the RC8 firmware or comment out the screen portion of the firmware if you are not using an lcd. Running that fast over usb might be the issue. An Sd card is recommended. At the same time those are pretty fast moves you might want to back down a bit this is a heavy machine. A laser only machine has much less moving mass.

I get overtemp shutdowns on RC8, not sure why, so I’m running an older version of Marlin tagged “THERM”.

I’ll bring those travel speeds down and see what happens.

Thanks for the ideas.

–Matt

If you aren’t using an extruder use RC8 and change the extruder themistor to 998 instead of 11. That therm firmware is too old.

If you aren;t sure what to do I can change it for you but it is just a number in configure.h

Alright, once the current job finishes, I’ll reflash to RC 8 and change the thermistor # to 998. Looking forward to going to modern firmware. I’ll also comment out the display, as I don’t have one.

Thanks again!

–Matt

Looks like I’m in business with RC8! 3/4’s of the way through the same code without a single out-of-place mark!

Thanks so much! I hope to be back soon with write-ups of my fledgling air assist project for Leo’s Laser, and managing the huge 2’ by 6’ work area machine for precision laser work in a smaller footprint.

I owe you another tip for keeping me from chasing my tail in the dark. Thanks again!

–Matt

Finished piece still had a few ‘misfires’ but the final was still good enough for a gift! Now running a different one at 2400mm/min travel speeds, we’ll see if that clears it all. Can’t wait!

Slower speeds made the errors burn deeper and more easy to spot, they are all in the second half. Some are ultra small pulses, kinda want to measure them and clock the times… Ordering the Full Graphic so I can try the SD card, that must be the missing link.

Did your computers screen turn off on the second half?

The screen does turn itself off, that’s a variable I need to quash too. Using a ten-year old laptop with Ubuntu as my head unit.

I’ll take a run at that today and see what happens.

EDIT: I have an octoprint install with camera out in the shed I use for monitoring and time lapse (Pulling snapshots using free SeqDownload). I’ll hook that up and run a job on the R-Pi instead of the laptop and see how it goes. On scrap, of course…

Boom! R-pi works, (Octopi) and having Ubuntu keep the screen on works too! I would have figured that out in a month or two.

Gonna switch to an SD card for my data transfer, since that seems to be best practice.

Thanks Ryan!

–Matt