Laser engraver m67

Hi,
I just started testing Chinese 15w laser engraver and can't figure out why this is happening:
I try to engrave text and one half of the letter burns deeper.

Controlled via m67, gcode generated by Laserweb4 in customized Marlin mode.

Attached files links from google drive, too large to upload here
Engraved: drive.google.com/file/d/1ywLHomE1Rkr7NMo...wc3/view?usp=sharing
Gcode: drive.google.com/open?id=1A5wAClJs0RRogHEd6n1PbrOFDlck6YN6
Picture: drive.google.com/open?id=1mULb52oydrkEU2Lnah2ag33eD0eEBo0O

it looks like it is an acceleration issue. The places where your burning darker/deeper are where you are lasering right next to a direction reversal in your rastering. What you should probably do (if you can figure out how) is make your raster move over shoot past where the laser turns off far enough for the axis to slow to a stop and accelerate back up to full speed before the laser turns back on.

Just thought of a way to keep laser intensity more of less constant at low speeds.
In hal multiply the analog value fron m67 with actual speed/some reference speed.That should give full intensity at full speed and reduce intensity as the speed gets lower.
You should see a big improvement with this.
I am away from any decent computer for a while so can t help with hal variables.

That might work if the output of the laser was linear. (I'm pretty sure most are not.)

The value from m67 comes into hal as an analog value. Just scale it. For a stepper motor the current speed is found in stepgen....velocity-fb somewhere. But should be resultant of x and y axis speed

Todd wrote:

it looks like it is an acceleration issue. The places where your burning darker/deeper are where you are lasering right next to a direction reversal in your rastering. What you should probably do (if you can figure out how) is make your raster move over shoot past where the laser turns off far enough for the axis to slow to a stop and accelerate back up to full speed before the laser turns back on.

Seems that was a problem and now are much better results:
drive.google.com/open?id=105sRZENT_fCcZGFyDwU1rEFlHT3YAbVt

Also, beam size and power test:
drive.google.com/open?id=1UG8PiZCtEuZvh-y1ge8UOoiFONjPAd4n

Looks like need play with beam size between ~0.1->0.15 area, with 0.1 may see burned outline like its overlapping or something like that and 0.15 starting to make pixels

The value from m67 comes into hal as an analog value. Just scale it. For a stepper motor the current speed is found in stepgen....velocity-fb somewhere. But should be resultant of x and y axis speed

I will not be able to do something like that by my self

Many thanks, guys, hope you will share more your knowledge

I am home with my machine again. Will try intensity reduction in the weekend to see if I
believe in it myself. I can't laser engrave in that machine but know what to look for.

Question to hal experts: I need to calculate a * sqrt ( b^2 + c^2) / d. There is no sqrt() hal module, is there?

No, but there is the "hypot" component that will do what you need in conjunction with the scale component

Great, even better. Thanks.

Kaspret, can you attach your machine's hal file?

I looked a bit closer in the G-code and the pictures. The darker sections clearly
comes from "overexposure" at direction changes. Due to low speed, too much light is fed to the material.

The G-code has a lot of intensity variation already. Typically starting at Q0.9 intensity
and over a millimeter or so reducing down to zero.
It looks like there is already velocity-compensation although it doesn't
make sense because it is not done in conjunction with a direction change or motion stop.
Did you do some intensity-gradient in the picture? Or is that a result of laserweb4?

It shouldn't be too hard to have the intensity velocity-compensated, although there may be quite some lines in the hal file.