Bump on engravings

I didn’t have any luck finding a similar post about this, but I also don’t check these forums too often anymore.

Basically I am engraving some stamps for use in ceramics, and what I have noticed with every stamp is that as soon as the “geometry” (white areas) stop and the black background begins again, the engraving leaves this ‘bump’ behind. I am not sure if it’s a power issue, or a focus issue. I would assume the focus would stay the same throughout the engraving. Here are some examples of what I am taking about.

As you can see, it only happens at the top of the engraving, and not at the bottom of the “geometry”. Here is the photo that I am using to engrave these.

As you can see it is a very high contrast etching, but it only creates that ‘bump’ at the top right when the white area is complete.

Here are the settings that I am using.

Material: Proofgrade Thick Clear Acrylic

Engraving settings: Manual Engrave
Map Grey’s to Power
Speed: 250 (I have also tried 300 and 500 with the same results)
Power: Min/0 Max/100
Passes: 1 (it also happened when I tried a faster speed with 2 passes)
Resolution: 450
Focus: Default focal length for this proofgrade material

Cut Settings: Default cut settings for this proofgrade material

These were all cut today so I should have the latest software/ firmware settings.

I haven’t tried these settings with another material yet, but if you have any thoughts, let me know.


What file format are you using (svg, bmp, etc…)? My first thought is that it’s probably an invisible layer in the original file.

I am using SVG. Basically I have the above image as one layer, and I have another layer on top of the image that has the cut lines. The bump doesn’t seem to correlate with any of the cut lines or the images.

The tube is overheating. It always happens in a transition after a section with breaks in it to a section without breaks in it.

Basically going from less engraving per pass to more engraving per pass causes it to heat up very rapidly and lose output power. Is yours a basic or pro?


Prior posts indicate a basic.

1 Like

Yes, it is a basic. That makes sense. I did however just try the numbers again and it didn’t happen this time. The first time I did the numbers I used a speed of 500 with 2 passes. This time I did a speed of 300 with one pass and it didn’t show up.

The second set of numbers was cut directly after I did 2 larger stamps with the Settings listed above.

1 Like

If your system has been running for a while or during a hotter period of the day it’s going to closer to that threshold/happen sooner. If you just fired it up for the first time in a while, the water cooling system and radiator it is cooling from will be at their coldest and able to mitigate those large power jumps much more easily.


I’m just amazed always at your diagnostic acumen. It sure helps to have someone with your experience and discipline on the forum.


That makes complete sense, but for some reason I noticed just the opposite effect.

It’s gotten a bit chilly the past couple days, so when I first turned on the machine it was about 63 degrees F in the room. I did the numbers first (with the 2 passes) and noticed that bump. Total time to etch and cut was about 20 minutes. After the numbers I then created 2 larger stamps using the settings listed in the first post. Each “large” stamp took 30 minutes aprox. After I got done with the 2 large stamps I tried the numbers again with a single pass at 300 speed, and it didn’t show up.

Between each job I took about 3 minutes to set up the next one. The set of numbers that didn’t have the bump was actually when the laser and room (now at 66F) was at its hottest.

It’s almost as if it got more stable to the changes at a higher temp.

Ha, thanks! I spent a while trying to figure this out once I got my k40. I was just using a vat of water to cool my laser, and it sat right by the window soaking up the sun. Once I increased the vat size, and changed the material it was in, the problems were much less worse than before.

For those more interested in WHY this actually occurs, it is because laser tube mirrors are at a specific distance apart, to increase the amplitude of the beam via constructive wave interference. (see the left side of the graphic).

However, when the mirrors change temperatures quickly, they change shape causing the waves to go out of phase and cancel each other out, or not amplify at 100% efficacy. (right side of the graph) This is where the laser loses output power based on how far the mirrors move. If they move half the distance of the wavelength of the laser it could cancel out the beam instead of amplify it.

This is why its important to keep temperature as stable as possible.


Being too cold can be an issue too. Having rapid temperature fluctuations is the ultimate issue here. I actually just posted a little bit about how that works above. What you really need in a CO2 laser is as much temperature stability as possible.

Im not sure if theyre doing any sort of predictive flow rate control based on expected temperature changes due to laser engagement per line, but if not, it might help. If theyve built that ability into the laser its something they could tune over time. Id be curious to know if thats the case


Thanks for the in-depth knowledge! I have repaired the K40 we have at the makerspace a few times, but I never looked too much into the theory behind it. I am also a little curious how stable the tube stays. It would be nice if there was that info in maybe a debug panel of sorts. I am assuming the software will still stop it if it gets too warm though? I know they disabled some of the thermal errors when I got my glowforge.

Thanks for your observations.
Not got my pro yet, but I will definitely give more consideration to a/c in the garage, or whether to switch to a house room, and cope better with any noise issues.


Interesting but why is it on change of temperature rather than just absolute temperature moving out of a sweet spot?

I.e. it seems the rate of change of power causes an issue but I would expect tube length to vary with the integral of power.

1 Like

Thermal expansion of glass seems to be about 5um / m per C and as the tube is around 1m length it will change by about half a wavelength per degree C. Do the mirrors really need to be aligned on wavelength boundaries?


Good point. It seems that extremely precise temp control would be needed to align on wavelength boundaries (a small fraction of a degree), and the tube must have much higher temp swings than that.

I tried googling the topic, and rapidly realized how little I know about how lasers work!


Hum, this doesn’t seem right to me. IIRC, the lines in an emission spectrum of a gas are not infinitely thin spikes, but rather each line looks something like a Gaussian. The length of the tube determines the frequency that is picked out of the Gaussian. Stimulated emission then enforces that frequency. So, as the tube lengthens the frequency changes a bit to compensate. Otherwise you would have to manufacture tubes accurate in length to angstroms to get it to work, and as it lengthens as it heats up you would see big variation in output power. Neither of which is seen.

@palmercr @AhnoldZ @Scott.Burns From what I’ve read it’s not the glass that is heating up, it is the mirrors mounted inside the glass, and the difference is only a few nano-meters - still enough to change output though. This will leave you with a lot less play since you are only accounting for something incredibly small instead of a length of glass expanding and contracting. And yes, there is definitely a sweet spot in there. However, as the tube is many times the length of the wavelength being amplified, it does allow for some slop.

To check I understand the problem:

You engraved the image shown in the orientation shown and the GF works up from the bottom. It makes a deep pocket in the acrylic leaving the letters raised but after finishing the letters the pocket has a bump in it top left.

How far along does the bump extend and does it fade out or finish abruptly as you go right?

AFAIK the GF engraves in both directions but you are saying you get the bump on the left only?

That’s correct; it engraves in both directions on the X-axis.

1 Like