"Overshoot" spots during engrave

[omegait]

Greetings all! I've been working with a new laser system, looks just like the LSR_LO12900ZR130W, and has the LO-X7 controller, but there's been a few small issues. The issue at hand is that when engraving colored anodized aluminum there is a scattering of spots in the direction of travel. My attached image shows one of our orange parts, but the same happens on blue, red, and everything else.

In the attached image I show two parts: the one on the left that looks like a mess came off our machine. The one on the right that's clean was produced by a laser guy who we were outsourcing these parts to before getting our own laser. The text is about 7.5mm tall.

Detail image of overshoot spots

Laser overshoot issue detail.jpg (120.97 KiB) Viewed 4405 times

Our machine was upgraded by our supplier to a 150w laser tube - at least that's what the documentation states.

Settings used:
* Mode: Cut after Engraving
* Cut settings:
** Min Power: 25% (so should be about 37.5 watts)
** Max Power: 35% (about 52.5 watts)
** Speed: 700 (not sure what the units are, possibly mm/s?)
* Engraving settings:
** Power: 30% (about 45 watts)
** Speed: 700 (again, not sure of the units.)
** Gap: .07mm

Setup:
My parts are laid out in an acrylic jig across a flattened table - I replaced the mesh with a piece of 2cm granite to improve the flatness. I then set the focus height and run the program across all the parts in the jig. I do not coat any

I've determined the proper focal position experimentally, by cutting across an inclined plane and measuring the height at the tightest spot of the burn mark. I'd have to say that the lens was an F50.8mm, and that corresponds to what I find is the most common focal length lens to be shipped with a machine. I'm purchasing an F75 and an F38.1, along with some accessory hardware, for cutting foam and finer detail engraving respectively.

I've already aligned my machine end-to-end, and I engrave without the air focusing tip: while my beam is at the correct angle to stay true across the machine, it isn't exactly in the center: there's still a translation issue even though the rotation has been corrected. Thus the beam wastes some of its power hitting the edge off the tip, so I just remove that and am making sure the lens stays clean or gets cleaned between setups.

I've verified that none of the mirrors are rattling, I've made sure the lens and tube are secured. I'm stumped.

[Techgraphix]

It seems to me that the left is a bitmap (jpg for instance) and the right one is vector.. correct?

Kees

[Tech_Marco]

Take out the focus lens and post a picture of that, both sides
By the way, you shouldn't use 130W for engraving project but should consider lowe power laser tube like a 60W or 80W. A 150W will make things worst.

To improve engrave quality, you should usr 38.1mm focus lens instead of 50.8mm

Marco

[omegait]

@Kees: Both were vector art. I'm not 100% certain of the other's guy's process but I do know he used CorelDraw to send the info to his machine. In my case I exported from Inkscape SVG directly to DXF and loaded that into LaserCAD directly.

@Marco: The lens is convex side up in the machine, I'll get an image of the lens when I can: the machine's currently in use making product, under-quality product, but product nonetheless; since it's only cosmetic our customers mostly don't care, especially since they'll be covering the resulting tool with granite mud, but both my boss and I aren't happy campers about it. It's a ZnSe lens, 20mm in diameter. When I last had it out for cleaning, it had collected a tiny amount of carbon, I carefully cleaned it with an alcohol prep pad from the medicinal cabinet. I'm switching over to using a lint-free cloth and isopropyl alcohol now that I have them. Cleaning the lens hasn't change the result. Neither has cleaning the mirrors. I've noticed a small nick in the beam combiner, and I've not yet eliminated that from the beam path as a test.

[DonL]

Focus is important, When I see the grainy image it tells me the focus is bad, looks like someone sprinkled a tiny bit of sugar on the letters, See my example and you can see that image 3 is grainy, now I used a slower speed and less power so it is not as bad but see the similarity is evident, plus see how much a little change in focus affects the engrave. The top image is razor sharp but the camera on my phone is crap so it is a bit fuzzy in the photo. plus to reduce glare it is on an angle... so the top image looks smaller... but getting past that, the letters are 12mm high and it was set to cut after engrave. 50% powe at 400mm/s on a 35 watt tube.

If you have 4 rollers on your laser head mount, make sure all 4 are touching the same, if one spinns because it is not in contact with the gantry, the laser head wil bounce around ever so slightly and cause this spotting and jagged lines. sorry... 1.5" focal length is 38.1mm since we are metric only here I get confuse when converting to inches

[omegait]

@DonL: Thanks for the test! Yeah, I expected an F38.1 to be a lot more sensitive to correct focal distance than my current F50.8. The current lens has a focal range of about +-2mm, and I've had good success in that space, but the issue here isn't the focus, it's the white spots to either side of the characters.

@Marco: Got an opportunity to take out the lens - it was still clean from my last cleaning:

Lens surfaces

Again, I operate the lens convex side towards the laser source, concave side towards the parts.

[DonL]

Your lens looks to be very high quality, I never pay attention to the concave or convex side, does it make a difference? which is correct? Convex up or down...

[omegait]

A possibly second "problem" is the wobble you can see in my first image: at any speed over 10 (?mm/s?) the laser can't run a diagonal or curve smoothly; all cuts at high speed wobble the lines.

The speckling is always in the primary axis of travel. In the first image I posted that was using x_swing, and in this image it was y_swing.

[omegait]

Your lens looks to be very high quality, I never pay attention to the concave or convex side, does it make a difference? which is correct? Convex up or down...

Yes, I did a bit of research into it and found that if you don't put it in correct side up you won't get the correct effect or will have more error on the focus. Something to do with the way optical systems work. IIRC you can find posts about it on these forums, but otherwise some web searches will show the science behind it.

[Tech_Marco]

Check the beam spot right in front of the laser head inlet hole. I think the beam may be a bit of scattered.
Test it with a piece of paper to verify it.

Marco

[omegait]

I positioned the head at the front-right extreme; the position with the longest laser path.

The beam combiner has been taken out of the beam path for several parts now and there's been no change.

As you can see, there's a very slight misalignment, slightly more than there was when I last aligned it, but still pretty good.

[electronrancher]

I had those little random nicks in engraving edges when my lens mount was loose. Check for a loose lens holder or mirror holder. Tightening mine fixed the random spots instantly.

Also, you may consider running at a lower speed - 700mm/s may rattle something that a lower speed wouldn't.

@DonL - Just FYI, A laser lens goes convex side up, flat side down.

[SWMS]

A possibly second "problem" is the wobble you can see in my first image: at any speed over 10 (?mm/s?) the laser can't run a diagonal or curve smoothly; all cuts at high speed wobble the lines.

wobble and speckle.jpg

The speckling is always in the primary axis of travel. In the first image I posted that was using x_swing, and in this image it was y_swing.

I get this same problem

This could be...
Vibration caused by the stepper motors at high speed in the frame.
Motor missing steps (try a higher number of steps on your drivers?)
Voltage too low on the drivers
Loose wheels

I have yet to eliminate all these things.

If it is stepper motor resolution. If all else fails you could try getting nema23's as opposed to nema17's. As far as I am aware these can handle higher micro step resolution.

I'd also consider a closed loop system. Servos are closed loop but if you look you can get closed loop steppers. Closed loop means the steppers basically know where they are in relation to your work area

With regards to your engraving. Note that when engraving with y swing the machine is having to move the entire gantry as opposed to just the head and with friction and mass acceleration and deceleration will be slower. At high speeds there may be overshoot due to the mass of the gantry.

There could also be some backlash occurring.

Seeing as you have the same problems as me, pm me and maybe we can eliminate them together!

[DonL]

I had those little random nicks in engraving edges when my lens mount was loose. Check for a loose lens holder or mirror holder. Tightening mine fixed the random spots instantly.

Also, you may consider running at a lower speed - 700mm/s may rattle something that a lower speed wouldn't.

@DonL - Just FYI, A laser lens goes convex side up, flat side down.

CHEERS!
I did not find that little tidbit anywhere so I have been just popping it in any old way and it seems fine but I will note that for future...
Convex towards Laser Source...

[omegait]

I had those little random nicks in engraving edges when my lens mount was loose. Check for a loose lens holder or mirror holder. Tightening mine fixed the random spots instantly.

Loose optics was my first thought. I know the lens isn't loose, and I'm nearly certain all the mirrors are locked down, but I've also not 100% verified.

Also, you may consider running at a lower speed - 700mm/s may rattle something that a lower speed wouldn't.

At the rate it is currently moving my boss thinks the machine's running far slower than the others he's observed. Others that are an order of magnitude or two more expensive...

With regards to your engraving. Note that when engraving with y swing the machine is having to move the entire gantry as opposed to just the head and with friction and mass acceleration and deceleration will be slower. At high speeds there may be overshoot due to the mass of the gantry.

It is, but when the pattern is significantly Y-axis dominant I've found it's faster - and the overshoot patterning doesn't seem to be affected by running x or y: both result the same. My programs are quite colorful from all the optimizations I put in to decrease time: if there are small patterns/fonts, those become cut-only; if there are sections of the artwork that are very Y-dominant, they get y-swing motion; otherwise x-swing motion. Multiple parts are done individually and in the order of front-right to top-left so that I can extract parts safely as soon as they are individually completed. Stuff like that.

As to adjusting the drive/beam parameters I've not yet explored that area so I'm at this time clueless in that area.