[bratan]

I'm using eBay 50W machine (SH-g350) w/ RDWorks v8. When cutting some PCB stencils I noticed a disturbing drift in X axis. The further laser hed travels in X axis, the more pads shift to the right. PCB was about 5x3 cm. so for example pads located at 50mm mark were actually cut at 52mm position.
I thought mayby RDWorks messed up imported png file, so i turned it 90 degrees and send to laser cutter, and now pads were cut correctly... Now head only traveled max 30mm in the x direction so shift wasnt noticabke. Is it hardware issue or firmware or software?

[avetek]

I have had a similar issue with a new machine I bought - it drifted in both axis and gave me oval circles in the engrave mode. I eventually traced it down to the stepper drivers which I replaced - I assume they were an early version of the 3DM580 or the firmware wasn't up to date. Anyway, replacing them fixed the problem

[bratan]

Thanks! Fortunately mine was simpler to fix. I just found this thread and it explains how to fix calibration issues! It totally worked too! I cut 200x100 square, measured and it was actually 203x100. Using RDworks software went into Vendor Settings, and using "pulse" calculator found correct steps settings. After I wrote them to firmware it now cuts precisely at any length!!! Can't believe how easy it was to fix.

[agrasyuk]

yep, it is not a drift at all. it's a result of manufacturer that doesn't care about precision and doesn't bother to preform step calibration.
in my case the deviation was almost twice on both axis. which is I guess is a very good thing since it was immidiately apparent. was able to bring X and Y to 0.05mm which is pushing the limit of my cheapo digital caliper.

I can imagine the pain of PCB pads being fractions off mm off when one assumes the machine is precise. and the mayhem when one starts to go through part datasheets and drawings while looking for the culprit.


BTW, Bratan, now that your steps are calibrated it is a good thing to verify squareness (unless you already did). axis not at exactly 90 degrees will produce paralelogram parts and will also be POS.

[bratan]

BTW, Bratan, now that your steps are calibrated it is a good thing to verify squareness (unless you already did). axis not at exactly 90 degrees will produce paralelogram parts and will also be POS.

I had no idea. But I think you are correct. My cutouts are skewed

[agrasyuk]

squaring is not hard actually even if you don't have super precise machinist square. cut 2 square shapes, i suggest to mark their orientation to avoid confusion. by rotating them against each other you can find if you have exact 90degree or which way X axis is sloped. uncouple the shaft, adjust slightly, re-tighten, repeat with test paterns untill they are exactly square.
of course if you have a precise angle gauge it will be loads better

[bratan]

Oh you are talking about squareness in XY axis, right? Should be easy to check that, thanks for the hint.
I was working about laser head not being perfectly perpendicular to the cutting surface. On the cut lines I see some sides have slant (kerf?) while others seem straight. I took speed square to the head, and it seems to be straight...

[agrasyuk]

check your mirror alignment, laser should hit the very center of the laser head opening. otherwise as you see there will be a slant in the cut as the beam is not falling onto material at 90Deg.
1/8 and below it matters very little, but at 1/4 it can be quite pronounced.

Attachment 275714

[bratan]

Thanks I didn't know that! Re-checked alignment, it was off. But for some reason I can't center it in all 4 corners I think rails/extrusions actually not leveled. For example I align beam on 3rd mirror to be in center while mirror is all the way in right front corner. But when I move it on Y axes to back, beam shifts to lower position Also when I move 3rd mirror close to 2nd mirror, it moves off center again. I spent probably an hour trying to adjust it but's just not possible to have beam at dead center at every position... I guess I'll have to leave with it...

[agrasyuk]

fisrt of all let me say i understand the frustration, it took me quite some time to identify and fix the same very issue. but its possible.
since you say "when Y axis moves..." to me it sounds like your first miror is out of alignment. adjust it fisrt so the beam stays centered on second mirror no matter the Y position. then mirror 2. in my case i didn't need to touch the third one at all.

you mentioned in other thread upgrading to red dot beam comibner. well, i did not bother to get that upgrade so not to mess with alignment again (even though my machine already has the mount for combiner lense and pointer pre-installed ).

[bratan]

fisrt of all let me say i understand the frustration, it took me quite some time to identify and fix the same very issue. but its possible.
since you say "when Y axis moves..." to me it sounds like your first miror is out of alignment. adjust it fisrt so the beam stays centered on second mirror no matter the Y position. then mirror 2. in my case i didn't need to touch the third one at all.

you mentioned in other thread upgrading to red dot beam comibner. well, i did not bother to get that upgrade so not to mess with alignment again (even though my machine already has the mount for combiner lense and pointer pre-installed ).

Actually I thought my first mirror is aligned pretty good. I always hit "bullseye" on second mirror at any Y position. But will double check again tonight just in case
About beam combiner, yeah, I was just reading that mirrors need re-adjusting when you install one... That scares me, I might reconsider and not get one, or at least postpone until "alignment pain" dulls a little