Motor Failure? Odd failures during second cuts

Troubleshooting
1

During my second cuts with the machine, I experienced an odd series of failures, but I’m not sure if they’re related.

At the time I was running v1.17.

While cutting, the machine disconnected from Wifi, but continued to cut, after a brief pause, although it never reconnected to my browser.

logs at this point:

Maslow-serial.log (4.5 KB)

Then a few minutes later, it paused again, and stopped cutting. At this point it reconnected to wifi, but had lost its positioning and told me I needed to retract the belts to zero.

Logs at this point:

Maslow-serial(1).log (2.3 KB)

Once I retracted and extended the belts again, when I went to jog the machine, I noticed a really strange behavior, the lower two belts were unspooling extensively, but the sled was not moving. At the time, I was trying to jog to the upper right direction. Initially I thought the issue was with the lower 2 motors being controlled improperly, but realize that the upper right motor was not pulling in any belt even though I was trying to move in that direction. This motor was completely unresponsive, and felt warmer to the touch than the other 3 motors.

See video of attempted jog in top right direction:

I tried updating the firmware to 1.18, and immediately after the motor did show signs of life for about a minute before it became unresponsive again. When it was moving, it had a higher pitched shriek than the others, and when retracting the belts, was significantly slower than the others.

Could this be a software or electronics issue? or more likely a failed motor? surprised since it has so little run time on it

I risk bought a new motor from the shop just in case because I’m a bit behind on this project

2

I did try unplugging and reconnecting the encoder and motor wires, and moving them around during commands to see if it was an intermittent failures due to harnessing, but it didn’t fix the problem

3

v1.18 was released yesterday which has significant fixes in these areas.

it also shows the machine position both relative to where you have defined home,
and relative to where the machine thinks the center of the machine with the
router all the way down is

it’s very possible that the unspooling of the bottom belts is either fixed in
this release, or is due to the Z height being wildly off from where you think it
is.

David Lang

3 Likes
4

Hey David! thanks for the reply

Initially this failure mode occurred when I was running 1.17, but as an attempt to fix it I did update to 1.18 and observed the same behavior. In the preview window, it shows the machine location moving up and to the right of the work area, but in reality the machine basically stays in place and unwinds the lower 2 belts.

I don’t know if I checked to see if the machine had lost its z-zero significantly between power cycles, but if it did, that might lead to a failure mode like this?

5

This seems like an issue. How does it feel when that belt is retracting or extending?

6

It seems to occasionally work when extending (by hand) or retracting without load (IE not on the frame). Any time its on the frame and commanded to pull the sled upwards it is unresponsive, and occasionally it is unresponsive altogether.

I wouldn’t say it ever felt “hot”, but rather just warmer than the others, which didn’t seem warm at all.

Is there any assembly problems that might lead to this (loosened bolts, poor connections, etc) that might lead to a problem like this? Wondering if theres anything I can troubleshoot in the meantime while waiting for the new motor. Thanks!

7

mfinn wrote:

Is there any assembly problems that might lead to this (loosened bolts, poor
connections, etc) that might lead to a problem like this? Wondering if theres
anything I can troubleshoot in the meantime while waiting for the new motor.
Thanks!

if the gear is too tight to the plastic (rather than spaced out a mm or so) that
could cause problems.

on the 4.0 kits, we had a problem with the motor being too close to the idler
sometimes (the motor mounts have a little slop, if the motor was too close to
the idler it would bind, loosening the bolts to shift the motor a mm or so
further away from the idler made a huge difference)

David Lang

8

is there a way to jog the motor manually in the software? Without needing to fully reassemble the hardware and do the retract/extend/tension routine? Would be helpful during troubleshooting

9

mfinn wrote:

is there a way to jog the motor manually in the software? Without needing to
fully reassemble the hardware and do the retract/extend/tension routine? Would
be helpful during troubleshooting

not really, there is a hidden feature to puse a motor for a very short time, but
I don’t think this would help you

@bar I think something like this (in a different tab/menu, not easy to trip
over) would be useful. I’ve started work on something like this a few times,
would you accept it?

David Lang

2 Likes
10

I think I’m going to go bankrupt sending out replacement parts for all the bits that get broken if there is an “override safeties and drive motor” button

2 Likes
11

Bar wrote:

I think I’m going to go bankrupt sending out replacement parts for all the
bits that get broken if there is an “override safeties and drive motor” button

what if it keeps the current limit safeties?

the biggest things would be to ignore:

  1. failed tests on other arms (do not allow driving the motor if the encoder on
    the same arm isn’t working)
  2. kinematics rules (and state would become ‘unknown’ as soon as you do this)

David Lang

1 Like
12

Maybe just make it cumbersome to enable, with many button presses and warnings for confirmation

Would make this troubleshooting much easier!

13

@dlang @bar I think I found the issue:

I disassembled the bad arm, and found that the spool had been binding on the arm halves. This was only present on the one arm exhibiting the issues in my previous post. See photo for wear surface:

IMG_1961
IMG_19611080×1542 257 KB

I bought some silicone grease and hope this solves the problem. The new issue is that I can’t seem to test the fix as my machine is having some software issues:

  1. After reassembly of the arms, it seems like the machine is upset about the belts being a different length(?) Getting the error message:

  • [MSG: INFO: Requesting state change from Belts Extended to Taking Slack]

  • [MSG:INFO: Succeeded]

  • [MSG:INFO: Measured waypoint O]

  • [MSG:INFO: Center point deviation: TL: 0.000 TR: 0.000 BL: -51.349 BR: 2.817]

  • [MSG:ERR: Center point deviation over 12.000mm, your coordinate system is not accurate, maybe try running calibration again?]

  • This is preventing me from moving the machine or running gCode to test if the motor is still failing

  1. I tried to recalibrate, but it keeps failing. I think the issue is this message ([MSG:ERR: Emergency stop. Update function not being called enough. 101ms since last call]) but I’ve included the logs as well if it helps diagnose.

Maslow-serial(4).log (10.4 KB)

Appreciate any help troubleshooting, I’m definitely less skilled when it comes to the software issues. Thanks!

14

then you start the machine, you retract the belts all the way, that sets the
Zero length on each arm. Then you extend it.

do the extend/retract cycle a few times, see if you consistantly get near zero
results from the retract (after the first one). It may be that when you
reassembled the arm the magnet got in the wrong place.

David Lang

mfinn wrote:

15

@dlang just want to make sure I understand correctly, when you say the magnet got in the wrong place, do you mean:

  1. I didn’t put the magnet on the correct side of the arm so it is right over the encoder sensor
  2. The magnet is right over the encoder but it is rotationally clocked differently than it was when I first built the machine

Obviously #1 would be a problem, but is #2 also a problem? I didn’t keep track of its rotational orientation when I took it out and reinserted it

16

also, where do I check if its getting “near zero” results after the initial retract? Not sure which values to be checking

17

mfinn wrote:

  1. I didn¢t put the magnet on the correct side of the arm so it is right over the encoder sensor

This would be the issue, if the roller with the magnet is in the other roller
spot or upside down you will have a problem

  1. The magnet is right over the encoder but it is rotationally clocked
    differently than it was when I first built the machine

This doesn’t matter, when you retract the belts, it resets the zero for that
belt

David Lang

18

mfinn wrote:

also, where do I check if its getting “near zero” results after the initial
retract? Not sure which values to be checking

when you do a retract-all, in the log window, it reports as each belt pulls
tight. As part of each of those four reports, it tells you how long it throught
he belt was when it pulled tight. These numbers are typically 0.1mm or less. But
if the magnet is in the wrong place, or is slipping in the rollers, these
numbers will be far larger.

so extending the belts and retracting them a few times to make sure these are
all very small numbers will help identify if this is an issue.

David Lang

19

I had that problem, and lubrication didn’t solve much. I had to sand the entire reel where there was friction until it spun freely (I did that on all four of my reels). Today my retraction force is 600.

1 Like
20

@dlang ran the test a few times, see values below:

[MSG:INFO: Bottom Right pulled tight with offset 0.043]
[MSG:INFO: Top Right pulled tight with offset 0.075]
[MSG:INFO: Bottom Left pulled tight with offset 0.161]
[MSG:INFO: Top Left pulled tight with offset 0.032]
[MSG:INFO: Bottom Right pulled tight with offset -0.193]
[MSG:INFO: Top Right pulled tight with offset -0.118]
[MSG:INFO: Top Left pulled tight with offset -0.107]
[MSG:INFO: Bottom Left pulled tight with offset 3.199]
[MSG:INFO: Bottom Right pulled tight with offset 0.043]
[MSG:INFO: Top Right pulled tight with offset 0.011]
[MSG:INFO: Top Left pulled tight with offset -0.011]
[MSG:INFO: Bottom Left pulled tight with offset 2.244]
[MSG:INFO: Bottom Right pulled tight with offset -0.011]
[MSG:INFO: Top Right pulled tight with offset 0.011]
[MSG:INFO: Top Left pulled tight with offset -0.054]
[MSG:INFO: Bottom Left pulled tight with offset 0.827]
[MSG:INFO: Bottom Right pulled tight with offset -0.043]
[MSG:INFO: Top Right pulled tight with offset 0.011]
[MSG:INFO: Top Left pulled tight with offset -0.011]
[MSG:INFO: Bottom Left pulled tight with offset -0.204]

Full log here:

Maslow-serial(6).log (812.6 KB)

There was a bunch of other print statements that I don’t recall usually seeing, like hundreds of lines say this: [MSG:ERR: MaslowKinematics: Failed to compute X,Y from belt lengths, using (0,0)], not sure if related to my issues

Thanks!