UR201 sudden movement

[cncdrive]

Could you please make a short video of it?

[Matt]

Could you please make a short video of it?

I'll absolutely try. It's impossible to reproduce on demand so far and aiming a camera at it while in use all day isn't practical. Of course, when it happens, shortly after I'm fiddling with the MPG trying to replicate it for this very reason. If I can somehow replicate the condition I will absolutely get a clip for you. I'm itching to figure it out and restore confidence. So many edge cases and weird bugs, so little time.

I had one idea cross my mind. Would it be possible that the MPG dial gets stuck between detents and somehow internally sends off what is essentially as rapid-fire burst of movements? As far as I can recall this spontaneous long travel generally happens when I'm dialing in that last small movement and I'm being very slow on the dial. To the point I wouldn't be shocked if I was actually holding it over top of the detent.

As it sits right now the only other weird thing that seems to pop up is I'll get that "no connection" message sometimes even though I'm a mere 3 or 4 feet away. But around here if you so much as sneeze you could lose cellular reception so I've generally just admitted defeat and assumed wireless stuff is always a mess.

[cncdrive]

I think it is not that hard.
You can put a camera next to the machine start it and stop it when it happens and then cut the end of the clip out.
And also then you could maybe make some video of the UCR201 screen to see if there is anything irregular after the issue happened.
I would like to see it, because we just can't reproduce this problem at all unfortunately and maybe we could get some ides if we see how it happens.

[Matt]

I'll absolutely try! Thank you for being willing to explore the idea. The struggle in capturing a video is that I'm running this for 12-14 hours a day cutting almost continuously and with the wireless pendant I'm all over the place. It's a very large table. A fixed camera wouldn't capture the UCR201 screen well enough. I'd need to get something like a phone video at the time... but it all happens so fast I'm more in a panic to grab an estop and not so much thinking about video. I do have security camera footage but even at 4k it's nowhere near clear enough to make out anything on the UCR201.

I can tell you though that the screen on the UCR201 (during these phantom movements) displays the coords and updates like it normally would. Nice and fast. I love that part. They increment/decrement just as they would if the machine was being moved manually with the pendant. It doesn't flash anything about connectivity issues or do anything weird like screen flicker. Once upon a time I even enabled that safety button enable option but I recall it still did make a sudden movement on me with that, so I turned that back off for now.

Anyhow I'm now on a mission to get you a clip if I can. Thanks for your offer to look into it for us!

[Matt]

Today it happened again. I couldn't get a video, only a photo as it all happened too quickly.

UCR201 screen when X decided to go X- until hitting soft limit

I was jogging X+ by 1.0 being very careful of the number of clicks felt. I needed to move X+ by 1 unit to get closer to my part. It moved X+ by 1, but then immediately went X- until hitting the soft limit which was hiding just to the side by about 3 units. I quickly pressed the STOP button on the UCR201 but my reaction time is not what it once was and according to the UCCNC software I hit the softlimit first. The log shows me "Softlimit reached;" Thank you, thank you, thank you whoever first implemented soft limits!

I'll add that during this move there was no other large machinery operating in the building (I've always been suspicious of a VFD or some giant AC motor introducing interference). I can say today at least that that wasn't the case.

I was a mere 3 or 4 feet from the USB receiver, line of sight, when this happened today.

Yesterday I had Z act similar but at the time I brushed it off as possibly being my mistake, so I didn't want to bug you. I was moving Z- by 1.0 for 2 detents when it completed the requested move then decided to move toward Z- until I hit the estop. I didn't mention this at the time because I was concerned perhaps my sleeve moved the wheel (which I'm normally very careful of) but nevertheless it would have only taken 5 detents to cause it to my Z max at the time.


Would it be possible to rig up a diagnostic tool for the pendant that logs all I/O from the UCR201? Perhaps I can leave it idle, machine offline, and see if it's actually getting some phantom signal from somewhere. I haven't experienced any of this movement with UCCNC alone so I'm fairly confident it's happening due to the pendant somehow.

edit: it happened yet again just now. Was moving X+ in 1.0 increments by approximately 20 clicks. Once it reached the commanded spot, it stopped and moved X- until I hit stop. There's no possible way I or anything told it to move 20 units in X-. However, I did set the pendant down at what seems like the same time as I was foolishly confident it was going to obey my simple movement request. If perhaps there's something inside that's insanely sensitive to slight bumps (and I do very much mean slight, it wouldn't have cracked a damaged egg on concrete) that might explain this. A wonky coil or something perhaps in the pendant?

[Matt]

Another oddity just happened with the UCR201. While the X axis was selected, pressing 'Home' at first completed the homing process for XYZ as designed in the configured order Z, X, then Y.

Attempted to run it a second time once this was complete and the UCR201 display says "RUN" but it did not actually trigger the homing process. Machine remained stationary. Pendant remained showing "RUN" until STOP was pressed on the UCR201. Subsequent 'Home' button presses all worked as expected. Almost as if it thought it sent the request to home the machine and changed the display before getting confirmation back from the software.

Then earlier today, pressing 'Home' caused Z to home and X and Y to be forgotten about entirely. I don't know if it updated the machine coords or not, but the only axis that moved at the time was Z.

[mickecarlsson]

Hi Matt,
I have also observed unusual movement when I lay the pendant down. In version 1.2115.
It’s been a while since I used my machine due to winter here in Sweden, but one thing I learned to do was:
Wait until screen in the pendant goes blank, then lay the pendant down.
I also change the axis with the button on the pendant, that seems to cure the unusual movement.
I jog the X axis with 100%, then change to 1% to get to my spot, then change the axis to Y on the pendant before laying the pendant down.

[Matt]

...then change the axis to Y on the pendant before laying the pendant down.

Thanks for the tip! I'll need to try this today. I can't reproduce the problem on command just yet but weirdly maybe it does have something to do with pendant orientation. I never would have guessed. But now that you mention it, none of my axis have moved randomly when the pendant is stationary, it's always been when it's being moved around or setting it down on the desk. Curious! Thank you again for the reply and suggestion, can't wait to test this!

[Matt]

Good news, I've got that video you wanted! But this is likely another issue altogether because the UCR201 was sleeping when this happened. Waking using the safety buttons on the UCR201 just showed the ESTOP state.

What happened:

- Vibrations triggered a physical Z+ homing switch during a job (this threw an estop condition in UCCNC)
- Corrected issue with switch (material in switch most likely, cleared it off) and home limit input verified by seeing the expected output on diagnostic page
- When I re-enable machine Z starts jumping around and smashing into my part
- Hitting cycle stop (despite no file loaded, no MDI commands, and no movements requested by the UCR201) does nothing as far as I can tell
- Regardless of number of times UCCNC is disabled/enabled it continues to make the same phantom Z movements

The only way to recover from this was to restart the software. No combination of cycle stop, estop/resets in UCCNC or the UCR201 stopped the movements.

https://photos.app.goo.gl/VTtP18geFo32yspE9

[Matt]

Yet another instance of this bug here today, except this time nobody was anywhere near the UCR201 but it was still active and screen display lit up. Without notice while installing a work piece Y rapidly moved an estimated 12 inches in the Y+ direction, and while I was rushing to grab the UCR201 and camera to take another video for you the Y immediately moved back an estimated 12 inches again at maximum speed and there it stopped. The only interesting thing to note this time is that the UCCNC screen displays "Softlimit reached;" and it did NOT disable the UCR201 (normally when tripping a softlimit I will need to manually cycle the estop on the UCR201. Nor did it trigger an estop condition in UCCNC. During this phantom movement the UCR201 displayed STOP above MPG mode.

So now I'm left to think this is maybe a problem somewhere in the software and not so much the UCR201 hardware.

This very much reminds me of the bug I linked a video of where Z kept jumping around. Only difference is this time it wasn't somehow stuck on repeat.