Darker etch on short runs

I’ve noticed that whatever my speed and power settings are, if the laser is only doing a small area, say a 2" image, the etch will be much darker than say, lining up a 6 2" images side by side so the laser has to travel 12" in each direction instead of 2". This makes it difficult to do a test etch. You think your speed and power settings are good with the test, but when you make multiples on a sheet, everything comes out much lighter. I’ve seen this happen time and time again. Is this something inherent with diode lasers or am I missing something?



  • Not enough overscan, so that the laser must travel slowly across the shorter distance

  • A very low acceleration, so that the laser head covers a vast distance while getting up to speed

Other than that, upload the *.lbrn2 file so folks here can poke around inside.

Acceleration. That’s a very interesting thought. I wonder if it’s because, on a small image, it’s actually going slower because it needs more time to get up to speed. I’ll prep a file and upload it.

I have my overscan set at 2.5%/.14". What is recommended? Should the amount of overscan be changed depending upon the speed set in the layer?

Exactly: starts slow, speeds up gradually, might never get to full speed before it’s time to gradually slow down.

It’d have to be unnaturally low for that to kick in, but if you switched units after setting it, it might be off by a factor of 25.4 or 60 or some combination thereof.

Once you get that settled (if it needs settling), then you can increase the overscan if the edges look darker than the middle.

Increasing the acceleration reduces the required overscan distance, but adding too much acceleration eventually means the motors can’t shove hard enough to get the laser head / gantry up to speed as fast as you want. It’s a tradeoff!

I’ll upload a file when I get home tomorrow night, but I don’t think that’s the issue, otherwise I’d see the images on each side darker than the ones in the center of the run. All of the etching on a long run, end end, is the same color. All uniform. No dark edges. Yet, when I etch a small area where the head it just zig zagging back and forth quickly, the result is very dark and much more defined than the result from, say a 12” run. It’s as though the power is intensified when etching a small area.

While you’re there, run the GRBL command $$ in the Console window, then copy-n-paste the GRBL settings in a reply so we can take a look at them.

Below is the gcode. The lbrn2 is attached.

I gave it another shot tonight and got the same result. When doing the small area (top row), the result is clearly defined an nice and dark. If I run a full line of them (second row), they all come out lighter. Argh…

Thanks for the review and help!

Sample.lbrn2 (304.0 KB)

Waiting for connection…
Waiting for connection…
[xTool D1Pro:ver B2]
Starting stream
Stream completed in 0:00
[MSG:Pgm End]
Starting stream
Layer C03
M8 N1
Stream completed in 0:47
[MSG:Pgm End]

Well, at least it’s consistent, which is not to be sniffed at when tracking down a problem. Diagnosing erratic results, now, that’s hard. :grin:

All of the settings look reasonable.

The Prusa acceleration calculator suggests the machine gets up to speed within 5 mm, ruling out my notion of absurdly low acceleration. You might want to increase the overscan setting, but it won’t affect the results here.

Maybe material heating is the basic difference between the single and multiple patterns. The plywood (or whatever you’re engraving) may be getting and staying hotter due to adjacent passes in quick succession with the single pattern. The multiple pattern layout allows more time for the wood to cool between adjacent passes, so it doesn’t darken quite so much.

If that’s the case, then running two patterns should produce a slightly lighter result than just one, but darker than three. I’d expect three or more patterns would be as light as eight.

Another test would be to align the patterns vertically, rather than horizontally. If it’s material heating, then they should all come out nice and toasty brown, because the horizontal scanning will heat each one the same way it does a single pattern.

Try those variations and maybe smell the results. :grin:

Thanks. Good idea. Kind of difficult to get consistent results if the length of what you’re making varies. Not sure if that’s the same case with a CO2 laser, but I can only imagine it is.

I guess if I want to try to get “normal” results, I’d have to force the laser to travel 6” out after each line and then come back for the next pass. Or just etch each one individually. That might make for an extended project time if I’m making 100 of these things. I’m not even sure if that’s even possible if I have an array of 30 or so on a sheet.

Assuming localized heat is what’s causing the difference, the larger arrays are the “normal” condition and a single pattern is “abnormally” overcooked. It all depends on how you look at it!

Limiting case: If the pattern was 5 mm across, we’d expect really dark results.

Because you want a darker result across the entire array and the laser is already running at 100% (assuming the power transferred correctly when I opened the file; diode vs CO₂ may differ), the only knob available is the speed. Reducing the speed will increase the energy deposited at each spot, which may compensate for the cooling between passes.

Try reducing the speed by a factor of four and engrave a 3×3 array. If that looks good, I bet it’ll be the same for 5×5 and larger, so you can run the entire platform at once. Slower speeds will increase the overall time, but once you figure out the exposure for the result you want, it should work the same way until the laser diode wears out.

This is all part of figuring out how the material you’re using interacts with the laser you have. A material test can give you an idea of where to start, but it cannot cover all the variations: the only real test is an actual pattern in an actual layout.

Experimenting on a cheap material (I loves me some good chipboard action) can help weed out layout and sequence goofs, so start cheap and work your way up.

This topic was automatically closed after 30 days. New replies are no longer allowed.