LaserUser.com Forum (formerly RDWorksLab)

See photo below.

https://drive.google.com/file/d/19iCYJV ... sp=sharing

These tests were run at 150mm/s - X scan and Y scan are pretty smooth, but when I do 45deg and 135deg scans the lines get small vibrations.

I have tightened just about everything I can think of, replaced the bearing blocks with HiWin, replaced all the belts, lowered acceleration and speed settings

I know the Y-axis drive shafts and pulleys aren't perfect but trying to decide whether or not to keep putting money into this machine.

I am interested if this type of backlash can be due to anything power related such as EMI, or if it could be a bad stepper or driver.

Looking for any kind of answers for this

Yep, you do- that is something in the mechanical X-Y setup [or possibly the lens loose in the lens tube?]. I am curious why you scan at those angles- you are fighting the mechanics.

I agree with Gene - something is loose in the optical path - mirror or lens. Or perhaps even the tube itself, or the bed.

EMI is an interesting idea, but I think it's unlikely. Here's how you can test it: you can clearly see some "wiggle" in the horizontal scans (upper left in your image). Do a single-line scan (one in each direction) with the Y stepper UNPLUGGED. If you see some wiggle in the Y-direction, you'll know that it must be unrelated to the Y-stepper. Conversely, if it's perfect, then something IS wrong with the Y-stepper.

Gene - I thought running angled scan tests would show me how the machine is performing at an angle, but I was also realizing this is never going to happen because its never swinging the motion that direction it's following a cut or engraving in X.

I've checked everything in the mechanical system including the lens and laser tube. I created Russ' 1st version tube mounts and the tube feels pretty solid to me. There is some strange wiggle on the X bearing block mount when I try to move it by touching the original head mounting plate. It seems like it is just the bearing block slightly lifting off the rail but it is pretty easy to wiggle.

I also have the acrylic head designed by Russ with his mounting bracket that has a bit of wiggle so that could be causing the backlash on the X scans.

Jeffrey - I will try this tomorrow thanks for the idea!

All in all the machine works decent for what I use it for and I have come to realize that the tests I run may just be flawed and I'm taking the results as issues with the machine.

The diagonal test may be "flawed" as you put it. However, the horizontal line test in the upper left is a very common test, and you should NOT be seeing the oscillations at the end of (alternate) lines. There's a problem, and it's worth fixing. These machines may not be capable of perfection, but they're capable of better than what your test shows.

By the way, are you sure the bed doesn't jiggle when scanning?

I tried unplugging the Y stepper and still had a little wobble in the lines.

Here Is a video of the head wiggle I was talking about.

https://drive.google.com/file/d/1Evi2dO ... sp=sharing

This is from the bearing block on the X rail. Is this normal?

I just replaced the block with a genuine HiWin so it could be a bad rail.

The bed doesn't jiggle as far as I can tell and I've shimmed the edges with card stock to stabilize it.

I would say you have found a problem, if not the problem. I would question if the aftermarket rails are exactly the same as the Hiwin bearing block- did you have the same problem with the first bearing block?

I have been in the middle of a big job so still haven't had a chance to put the old bearing block on to check the fitment on the rail.

See video below for another potential issue.

https://drive.google.com/file/d/1LbvpfR ... sp=sharing

The Y-axis driveshafts wobble and I can imagine this vibration gets sent through the belts into the head as well.

The head wobble definitely needs to be fixed first, but I am able to cut most of my jobs without much of an issue.

Circles and curves have a little bit of wobble in them but it's not very noticeable.