Skewed gantry- duck waddle


#1

After a bumpy start getting the forge online, I downloaded a simple test file (an oval with a box around it).
Successfully uploaded via the laptop and ran the job. Just a simple score on some heavy 0.20 mm card stock. Forge, scanned processed and reported ready to print.

Unfortunately, it looks like the forge has a, for lack of a better description, dragging right stepper. While the job is running, the right and left side of the gantry appear to be out of sync. One moves before the other. Kind of like a duck waddle.

Below is bed image before print.

Below is result of 2 attempts (rotated stock between runs)

I have noticed that whenever I turn on the machine, it goes through its calibration “dance” and the final move always results in the right side of the gantry being skewed tight against the back of the unit, while the left side is about 2-3 mm away from the back. The effect is that the gantry arm is not parallel to the chassis. I have followed the re-set instructions several times, centered head etc., always with the same result.

Help!


#2

Yea… You’re going to have to send that back. Curious if the box showed any signs of damage? That seems like it would take a big hit to cause


#3

I had box held at UPS depot and picked it up myself.
Had three handles, a bit of split paper tape, but otherwise quite good shape considering we live out in the middle of Colorado.

REALLY hoping it can stay…


#4

Well I hope you are lucky and it can stay! Perhaps something is just loose and support can walk you through checking it.


#5

I would check all the V wheels are still on the rails, that the belts are over the pulleys at both ends and not slack and check the pulleys are tight on the motor shafts.


#6

Done, done and done. When the steppers are energized, the gantry is locked up tight. Can’t move either side. When I did the test cut, it really looked like the right and left were being driven out of sync. When the machine goes through calibration, it adjusts each side separately with small bumps, then homes. I will be trying a longer test cut later today and try to video.


#7

Those checks, nothing on the rails - definitely sounds like something for Support to throw some troubleshooting and input on. Hope it’s resolved quickly.

Did you print a founders ruler, just to weigh against a positively known, good variable?


#8

Actually don’t have long enough exhaust hose, so was venting full blast into the dining room, (the fan is AWESOME by the way, kind of like a 747 on the St Martin ((before the hurricane)) runway.) thus the very small test. It was actually a score, 500/1. Going to Home Depot this morning…


#9

That’s the best problem report I’ve seen…the title is clear (and clever) and the description is spot on.

Good luck getting it resolved.


#10

Haha. Yes, I forgot to drop my hose out the window. That thing was flying around like a hose that a fireman dropped.

Support is (probably) going to ask you to run it against a known good file and material. So a founders ruler and draftboard, or whatever proofgrade (draftboard being the most economical to test on). Might as well knock that variable out just to save some time.

It’s at least mildly interesting to me that both of the cuts appear to be uniform in their skew.


#11

The ruler isnt going to be a great test based on what he posted because the straight lines in his test look pretty much fine.

Curious what just a simple circle looks like? This type of problem is a common one on DIY cnc machines with a gantry.


#12

It’s a known working file - as are any of the catalog prints. No doubt their is another catalog file that may illustrate it better. The ruler does have combined X-Y movements for both engraves and cuts that will illustrate the issues.


#13

As a few people said, it would help to do a test with a more known quantities. After you get your exhaust set up, would you please:

  • Turn off your Glowforge (this is important to avoid damage to your unit)
  • Open the lid and, using both hands, gently move the laser arm to the center of the bed
  • Gently move the head under the lid camera
  • Turn your Glowforge back on
  • Place Proofgrade™ draftboard in the center of the bed
  • Print the Founder’s Ruler using the default settings

Let us know if you see the same issue with the duck waddle! If you do, unfortunately, I will likely suggest that we replace your unit.


#14

Doesn’t look good. Made up this test file, ran it…

Keep your eye on the right side of the gantry in vid below…

scan of actual test result…

did the founders rule on card stock, also not good…

ran founders rule 2nd time scaled up… results…

https://discourse-cdn-sjc1.com/business5/uploads/glowforge/original/3X/5/e/5e5ecb90401242bd0d10a4df77c56f2c5eb56b15.mp4


#15

You might want to run just one test on the Proofgrade draftboard that they provided. I hear that weird things can happen on non-Proofgrade materials. They have to see it tested on Proofgrade to help make the diagnosis. :slightly_smiling_face:

(But it doesn’t look good.)


#16

That is ridiculous. The Y motors should always move in unison after it it has homed. It is a fault if they don’t regardless of the material. How can the material on the bed cause the Y motors to misbehave?

There should be no reason to reproduce bugs in PG unless it is something that can depend on material properties and even then this is not a laser that only works with PG. It has to work with normal materials suitable for lasers.


#18

While that’s entirely true, and the machine is obviously whockerjawed, they have a standard “yep, it’s broken” test (which we know will fail), and they even give people a piece of material to run the test on, so what’s the big deal with just doing it?


#19

Because it is a complete waste of time and material. It will either do exactly the same or it will magically work. In either case the machine is faulty. It is no good if it only works on PG materials and does a duck-waddle on non PG materials.


#20

Lol! Reminds me of calling tech support.

Me: “My computer will not power on.”

Tech Support: “Ok, Sir, I can help you with that. First thing, can you save all your work and reboot your computer for me? That does seem to solve many problems.”


#21

And that’s not the issue here. I will be right there in the angry mob if, at some point, someone has a machine that only works on PG materials and Glowforge tells them tough luck.

Maybe I’m old and have no fight left in me, but despite decades of experience programming and troubleshooting all kinds of technology, when tech support asks me to try turning it off and on, I just do it and move on. Even if I’ve already diagnosed the problem myself and am 100% sure it’s a faulty interocitor part. I’m on the phone with someone who’s trying to follow a process and get to where I need to be: I can argue with them or I can cooperate. I’ve also been on the other end of those calls, and I can tell you few things were an indication that I was about to have a bad time diagnosing what was very likely to be a trivial problem than the guy who starts with “I’m an engineering manager at Google and I’ve been building computers since before you were born, so don’t ask me if it’s plugged in”. That’s where I stop listening because it’s probably not plugged in.