Help elimante remaing smoke residue corners (upgraded air already installed) Omtech 2028 60w redblack

I have an Omtech 2028 60 redblack and I have upgraded the air assist and have trying to get rid of the remaining smoke residue on the top (only occurs in the start and stop position now). I am thinking that I since I am already at 30 psi (8 gal California Air), I would ask for help on prehaps getting the laser to start moving slightly faster so the laser head spends less time in the start and stop point. I have already tried min power settings and that has not helped).

I have included an image of factory vs 30psi for those that do not know how much improvement you can get with upgraded air - worth the upgrade just need to fine tune it a bit right now.

Also if anyone has gotten an Omtech redblack recently would you post your cut parameters so I can see them (got 1 3 months ago that had totally different cut para then my warranty replacement i got 2 weeks ago)

2024.06.21_13.06.00 Screenshot

Settings comparison I did for a firmware update I was asked to do before warranty replacement. Start speed was higher on first then on second, as well as Max accel, Accel Factor, GO Accel Factor, and Speed Factor

How does the Ruida know to use the cut parameters or to use engraving parameters?


Although it’s not for we mere mortals to know, I think the controller has an “engrave this data” command so it can calculate the overscan distance on both sides of the pixel array making up the line.

Everything else would be a point-to-point vector “cut” operation.

But, yeah, I’m guessing … :thinking:

Again, how does it know does it know the difference between a vector cut and a vector engrave?

How would Lightburn know what you were thinking, vector cut/vector engrave in order to send these controls?


AFAICT, it doesn’t, because the result depends on the material: a setting that cuts paper will engrave acrylic.

The distinction may be a matter of sloppy terminology on our part, because we describe both raster-scanned areas of filled / image data and non-cutting vector paths as “engraved”. The controller treats those two layouts differently, but we consider only the result.

Perhaps we should use “engrave” and “vector”, but even I can’t be that fussy. :x:

When I first read about RDWorks, I had though there was an actual cut button…

I keep going back in my mind wonder if this is how the Ruida knows which parameter set it uses. I don’t have it on anything to go back and check.

Every time I go into the machine settings, I wonder, how does it know… which happens a lot :zap:

This is completely my fault… My Chinese isn’t too good… that’s why their English, no matter how bad, is better than my best Chinese…

If you think about the operation … for these to work, it has to resolve to vectors to tell it where to go… Even a fill/scan operation is only a vector path between two points…

Knowing nothing about the data stream… it’s all speculation anyway… I guess you could change some of these settings to be visibly obvious and see what affects what… Tough way to figure it out.

IMHO, something it telling it which parameter list it needs to use… fixing hardware as long as I did, this means somehow these different lists are selected. You couldn’t fix it if you didn’t have some way to select one or the other…

Wonder is @Rick would know? :pray:

Of course none of this really helps the OP. :frowning:


I think GRBL controllers work as you describe, because they do not have a separate “engraving” mode. That’s what led to the G-Code Clustering optimization, where the controller accepts and applies a block of per-pixel power variations during an “engraving” move to avoid choking on the torrent of commands required to change power levels on the fly.

For a Ruida controller, a vector “cut” has a ramped power profile on each end as the head accelerates & decelerates. Whether it actually cuts the material or just leaves a mark (which we say “engraves” the line) depends on the material, about which the controller knows nothing.

But a Ruida controller handles “engraved” motions differently than vectors. Each pass across a filled area or an image has zero power in the overscan area as the head accelerates / decelerates, then one of:

  • Constant power in filled areas
  • Per-pixel on/off power in dithered images
  • Per-pixel variable power in grayscale images

Those happen during the constant-speed portion inside the pattern or image, with the controller managing all the hocus-pocus.

Which agrees with the fact Ruida controllers handle “engraved” motions only when they’re parallel to the X or Y axes, with LightBurn simulating all that along diagonals, producing weird results in some cases. Whether that restriction comes from a limitation of the Ruida “DSP” hardware or from firmware written to handle the most common cases, we may never know.

Aye, but having thought about it a little more must count for something.

Does anyone have help for my issue?

I understand there is a lot of non relevant posts here so might have to repost.

I’ve been cleaning up the smoke marks with alcohol prep pads. So far all of my smoke marks disappear and once the alcohol evaporates, there hasn’t been any discoloration from the alcohol. It’s not going to prevent smoke marks, but it helps to clean them after the fact. Hope this helps.

Don’t duplicate your posts or double post, it’s not allowed here. Simply because it now takes us longer to read your additional posts… Add another post here is you feel ignored and it will keep the thread alive.

People are seeing your posts, there just isn’t an answer that works for you yet.

Hang in there…


I think people are seeing it but are seeing all the discussion on how Ruida works rather than what my question is. I will hope that people can read beyond that.

I understand, that’s common on these threads… I think I pointed out it’s not helping you…

You’ve got an odd problem and without any of us having a hands on, it’s difficult to try and isolate what the cause is.


1 Like

I notice more smoke marking on one axis than the other (X or Y) ? especially on the 25psi cut.
Looking at a similar test cut here I have the same “feature”, mine has more marking on the Y axis than the X. Both on top and underneath.
Question for Lightburn support, is this a software “feature” ?

While I’m not “LightBurn support”, what you see is more likely due to the diode laser’s rectangular spot than anything else.

What does the defocused spot look like? If it’s an oblong shape aligned with the axes, it cuts better in the longer direction and worse in the shorter direction.

1 Like

Although his profile says grbl, the machine settings are clearly from a Ruida.

I think his only options is to modify the start speed and min/max power.


I gotta stop paying attention to those … :person_shrugging:


Sorry I have not updated my profile in the last 3 months of owning the Omtech because I am still trying to get a machine that works correctly. Omtech customer service has been excellent as well as technical support, but sending a machine 3 weeks across a rolling ocean full of waves has gotten me 2 machines being replaced in first 3 months, third machine is in it’s way, really hope as my wife says “third time is the charm”. Again customer service has been excellent, just gotten unlucky with a motherboard issue on the first and unfixable lack of linear rail square on the second.

This return had just happened yesterday but I’ll leave this thread going for the other people that have posted on it about their issues.

We’re not picking on you, but if that information isn’t up to date … you get people trying to help you with wrong or inconsistent supplied data. You’re here trying to fix it, so you do have one, working or not.

As far as I know, none of these are built here in the USA by OMTech or Cloud Ray.

I got mine with a OMTech guaranteed arrival in two weeks.

The ship ended up parked off the coast of Long Beach, Ca waiting to unload during covid … cost me a few weeks to get it unloaded from the ship.

The plywood box it came in was about to fall apart and I could just peel away parts of it…

It worked fine, but the box and plywood is the cheapest I’ve ever seen… I don’t know how I could have returned it…

How did you return your two machines?

I’m glad you have had success with OMTech.


Thanks, I’ll look for that.
Can the elongation be controlled with more accurate height ?

As @jkwilborn pointed out, you’re using a CO₂ laser with what should be a nice round beam focused down to a nice round spot that will cut equally well in all directions. I mistakenly thought you were using a diode laser with an inherently rectangular spot.

However, that nice round beam depends on having the laser tube resonate in what’s called TEM00 mode. Failing tubes can resonate in other modes, typically TEM01 and TEM10, with elongated shapes that will cut poorly when moved broadside. More than you want to know about modes.

To rule that out, test the beam shape at Mirror 1 by putting a scorch on paper across the entrance aperture, just as you do when aligning the mirrors. If it’s not a nice round dot like these:

Then you have a real problem.