I'm getting bad results with wobbly lines, what could I do to improve things? [Solved]

I was able to complete a relatively large job (single plywood sheet desk, all files available here.

Unfortunately the results are pretty bad. Below is what I mean by bad. Some things may be causes and not issues, so advice and thoughts would be very helpful

  • Many straight cuts are crazy wobbly. The vertical line is 4", I’m sure it should not be that off.

  • Scale of the CAD is not respected (although I think that relates to this issue). This error is over approx 30"

  • One arm seems to have partially seized immediately after the job. First would not retract, then made a horrible rubber friction-ey sound and partially retracted after reboot. That motor is also pretty warm.

My biggest hangup is the wobbly line thing. Some details on my setup and the job:

  • Firmware 0.83.1
  • Horizontal orientation
  • 0.25" upcut mill
  • 6.5mm step, router speed 2 (seeing chips)
  • feedrate: 3000 mm/min
  • 0.5" Sande ply
  • 6.5mm step down
  • Frame specification (from maslow.yaml)
    Maslow_tlY: 2237.500000
    Maslow_trX: 3639.500000
    Maslow_trY: 2256.899902
    Maslow_brX: 3674.699951

Cheers all!

2 Likes

the fact that you are saying that one arm is having an issue can account for
this. Please take a video of the cutting next time, it makes it much easier to
see what’s happening

are belts getting loose as you are cutting?

David Lang

Hmm… I have some partial video of that job as I was excited that the machine appeared to be getting through it. They are uninteresting though.

I am wondering if the problem is caused by the z axis. I like to re-set z bottom, then set z home before doing any jobs to prevent belt slack.

I noticed that there is always a slight (0.5mm) difference between the base of the arms on the lead screws, which could cause the mill to sweep an arc if the base twisted during cuts, which I have seen it doing.

That seems like a pretty aggressive federate and step/cut depth. I’m having mixed results with 1000mm/min and 7.35mm step on a single flute. Certain parts of the board worked with those settings, but not near the edges.

What does the bit manufacturer recommend for federate?

Also worth asking if your belts are tight? I had software issue where on jogging one belt would go a little slack. Hard to notice at first, but caused accuracy issues.

Coming from the 3D printing world, slowing down can make a world of difference in quality. The settings are usually a trade off of quality vs speed.

Michael Barrow wrote:

Hmm… I have some partial video of that job as I was excited that the machine appeared to be getting through it. They are uninteresting though.

I am wondering if the problem is caused by the z axis. I like to re-set z bottom, then set z home before doing any jobs to prevent belt slack.

I noticed that there is always a slight (0.5mm) difference between the base of the arms on the lead screws, which could cause the mill to sweep an arc if the base twisted during cuts, which I have seen it doing.

reset the Z stop by running the machine all the way to the bottom (this will
also even out the two lead screws)

David Lang

I use a feed rate of 30"/min.

I would draw a rectangle with the x,y coordinates from the yaml file and measure to make sure that they match.

1 Like

Thanks all for the suggestions.

I think much of the trouble can be eliminated with replacing parts and fine tuning.

  • I replaced the arm that was getting hot with new parts. I had a kickstarter phase 1 kit and I found the drive train tollerance was way better with the new arm. Maybe the moulds wore in.
  • I increased my cut tab widths to make pieces more stable

This did help but I now have another problem where cuts are getting progressively deeper and the safe depth is not high enough to precent gouging. Pretty sure the bit is walking loose, which would also add wobble to cuts.

Next thing Ill try is cleaning the mill shank with alcohol and thoroughly cleaning the chuck.

Ideally I would use the two spanner method to tighten the chuck, but thats not possible. I also dont want to put any torque the base since I’ve noticed the nut holes in quite a few places are basically beginning to disintegrate after repeated disassembly and reassembly. @bar, im pretty bummed about this since Ive pretty much bought two kits but still not getting the results I need. Im trying to start a businesses but the new kit is performing worse than my old M2 did which is very disappointing.

1 Like

Thanks for sharing, I’m having issues with wobbly lines too but with way lower feedrates and depths (more like 400 and 1-2mm depth). I’d love to blame the speed but, I’ve not had enough success to dare take it to the sorts of speeds you’re at.
I’m also interested if the z stop/home settings are part of this, I’ve seen a few people mention it including yourself.
I expected to change z home depending on the bit i’m using, but wasn’t so sure on changing the z-stop. Are people doing that every fresh start?
One problem I had on my last calibration was that the cable for the top z stage (opposite side to the dust extraction) got jammed under the bearing carriage, that resulted in a difference in z height on the two sides.

My reason for doing z stop with every job is I have to power cycle the machine and it fails 95% of the time mid job (I will pause the job or hit the stop button with the router at some job defined z position). I did not inspect the firmware but if belt positions are volatile on a power cycle, I assume the z position would likely be too.
Since belt tension factors in z position, I want to ensure that is accurate so always set z stop at the start of a job before z home.

1 Like

Aloha,

So…my Maslow 4 was finally cutting a few days ago, but it was throwing the current warnings and occasionally stopped because it lost its location, and when it did cut, it sometimes cut wobbly lines.

About pulled my hair out, then decided to try something.

I’m using a 1/4" compression bit and was cutting 1/8" passes. I reduced my pass depth to 3/32" and voila! I’m not getting warnings, I’m not getting errors, i’m not getting wobbly lines.

It’s a miracle!

Your mileage may vary…

Yep I’ve got one ‘sticky’ arm as well. Sometime soon I’ll be pulling it apart to explore ‘fine-tning’ it.

1 Like

If you are routinely getting current warnings, it probably indicates a problem
on an arm (screws tightened too much, motor too close to the idler, etc)

David Lang

I cant solve the x y aspect ratio issue but otherwise I was able to get good results after replacing the arm and a few other things. The full bin list:

  • Replaced arm with current warning. This arm was manufactured with too tight tolerance on the metal idler and motor sprocket. Motor adjustment was not the solution. I had very wobbly straight lines before this.

  • Increased step down to 6.5mm, decreased speed to 2000mm/min

  • Set router speed to 2

  • Cleaned router collet and nut interface. It was jammed with ultra fine sawdust which despite extra snug collet tightening, caused the bit to wobble (best case, only percievable with a shrieking sound during cuts) or walk (worst case, bit would drop 1-3mm down in the collet on each pass).

1 Like

Is there a thread anywhere about rebuilding the arms? How to judge whatever it is that’s too tight. I have one that gives me current warnings and I guess the reason I never had a job complete until the latest firmware.

@Matthew_Hart maybe, but I replaced mine because the idler and sprocket fit was so tight that the idler could not freely slide to the bottom of its metal axle during assembly and I had to use some force to get the idler on.

I could not wiggle the motor to relieve compression on the gear teeth because the arm casting had no play in the motor mounting holes.

I assumed this was by design at the time to prevent backlash, but the idler and sprocket fit in the replacement arms I purchased was looser to the point where the idler could free fall down its axle shaft during assembly.

I hope this helps.

Michael Barrow wrote:

@Matthew_Hart maybe, but I replaced mine because the idler and sprocket fit was so tight that the idler could not freely slide to the bottom of its metal axle during assembly and I had to use some force to get the idler on.

I could not wiggle the motor to relieve compression on the gear teeth because the arm casting had no play in the motor mounting holes.

I assumed this was by design at the time to prevent backlash, but the idler and sprocket fit in the replacement arms I purchased was looser to the point where the idler could free fall down its axle shaft during assembly.

the idler should slide easily on it’s shaft.

David Lang

1 Like

I think that idler gear is 100% a weakness of the design and it needs some attention. That is high on my todo list.

1 Like