Keyboard hold-to-jog stops working at Speed 4 (CM 653)

Hi all,

Noticed a minor bug with Carbide Motion 653 that I wanted to flag. When jogging with the keyboard (arrow keys, Page Up/Down), hold-to-jog works perfectly at speeds 1, 2, and 3. But at speed 4, holding a key moves the machine briefly and then stops. You have to release and press again. It doesn’t do this 100% of the time, but it does more often than not. (It’s not a usb device issue - same thing happens with laptop’s onboard keyboard and with external USB keyboard).

This happens on all axes (X, Y, and Z). It doesn’t happen when using the on-screen jog buttons with the mouse — those work fine at all speeds including 4.

Reproduction:

  1. Connect to machine, go to the Jog screen
  2. Set speed to 4
  3. Hold any arrow key or Page Up/Down
  4. Machine moves a short distance then stops
  5. Switch to speed 1, 2, or 3 — hold works as expected

Tested on CM 653 stable on macOS. Happy to provide any other details if that’s useful!

It wouldn’t hurt to give the latest beta (654) a shot. I’m not sure if the code for keyboard inputs was touched, but one of the bug fixes mentioned jogging.
https://carbide3d.com/carbidemotion/beta/

2 Likes

Thanks, Winston. I gave that a try but unfortunately, I can confirm the same bug persists in 654. Thanks for the suggestion though!

I see 654 has been promoted to stable — just noting that the Speed 4 hold-to-jog issue is still present in it as I reported here on March 2. Would love to see that addressed if possible! Thanks for considering.

This is interesting, @motorharpdesign. Your continuous jogging stops, and my jog shuttle continuous jogging pauses every inch to two inches. Of course, on the shuttle there is only HI and LOW jog speed settings. I wonder if somewhere way down in the jogging code these symptoms are related…

My assumption has always been that the software pauses so that it doesn’t get ahead of where the hardware is.

I’m not so sure about that, Will. There is significant overtravel on the continuous jog (I have buried the endmill in the work on two occasions), whereas the old Mach3 jogging with the Shuttlexpress (still my gold standard) stopped instanly upon releasing the jog control. The jog shuttle also seems to buffer clicks of the dial and effectively overtravels when I’ve been turning the dial too fast (there should be a short timeout and then motion cease upon stopping turning the dial).

But I don’t want to hijack Kevin’s thread too much so I’ll shut up now. :slight_smile:

Will, I did some more thinking on this while I was picking up my laundry from the laundromat.

The software should never be in a position to get ahead of the hardware. Jogging should not be at such a rate that acceleration/deceleration (and thus advance motion planning) is needed. Stay within the pull-in speed capability of the steppers. That way, mash the button and the mechanism moves. Let up the button and the mechanism instantly stops. No opportunity for overtravel. Now this concept is probably germane to Kevin’s issue too…

…only on HI speed. LO continuous jogging with the shuttle is pause/stutter free, I realized this afternoon while setting up a job. I’m wondering afresh if our symptoms are related, @motorharpdesign

I went back and tested on CM 640 (the last build before the jog rework in 641) — hold-to-jog at Speed 4 works perfectly. Smooth, continuous movement on all axes, no pausing or stopping.

Same test on CM 654 — stutters and stops after a short distance, though not every time, exactly as described in the original post.

Here are two videos showing the difference:

CM 640 — Speed 4 hold-to-jog (working):
https://i.imgur.com/gZ1F4AB.mp4

CM 654 — Speed 4 hold-to-jog (broken):
https://i.imgur.com/MsZV9Dk.mp4

@WillAdams I don’t think this can be the software intentionally pausing to stay ahead of the hardware — build 640 didn’t have this behavior. Something in the jog rework (641+) introduced it.

@Randy it seems to me that your jog shuttle stuttering at high speed is almost certainly the same bug — the fact that it only happens at high speed and low speed is fine matches exactly what I’m seeing with keyboard input.

Build 640 is still available for download if anyone wants to verify independently. Speeds 1-3 work fine on both versions — it’s specifically the high speed / Speed 4 continuous jog that regressed.

Thank you for your writeup, @motorharpdesign, but I can’t try it because

If you have a Carbide 3D Shapeoko or Nomad running Carbide Motion Build 651 or later, then it will work with the Jog Pendant.

1 Like

Funny that the update that modified CM to allow the new jog pendant Carbide sells simultaneously broke the smooth functioning of every other jog option like wireless controllers setup to use the keyboard controls. A suspicious man would see this is shady.

It’s been broke for me as well at the same update, interestingly if you use the software buttons in CM it doesn’t do it, only when using keyboard commands or controllers that use the keyboard commands, oddly specific way to break it. Hopefully it was accidental and will be corrected.

Just as a note, jogging from the keyboard works entirely as expected on Windows 10. Entirely smooth, no stutters,

CM 655, Windows 10.

I think the term is “unintended consequences”… :slight_smile:

On my Windows 11 CNC PC, the screen jogging works perfectly also, but there I do not use continuous jogging, just incremental (I have not tried keyboard jogging yet). My problem is specifically with the jog shuttle continuous jogging (and the continuous rates being way too high).

@Randy yeah, agreed — probably an unintended side effect of the pendant rework that hasn’t cycled back. Supporting keyboard jogging is still a no-brainer feature and having the machine intermittently stop jogging is definitely a bug! Just tested keyboard hold-to-jog at speed 4 on 656 and it’s still broken — same behavior as 654/655.

@wmoy — any word on whether this one’s on the list?

@motorharpdesign, as I pointed out in another thread, the latest CM (build 656) totally solved my jog shuttle glitches and it is now performing as I had hoped (admirably), so I frankly haven’t gone back to check out the keyboard/screen jogging on my installation.

1 Like

Oh, and if that seems like a subtle plug for the jog pendant, I want to blow away the subtlety and say right out that the jog pendant is the best thing for job setup since ZigZag (or Rizla if you prefer) paper. Actually used in combination for touchoff…

1 Like

I won’t speak for the software dev team, but we made a recent change in how we filter user inputs to eliminate the “double jog” bug when using touchscreen and mouse inputs. It may have introduced a bug elsewhere in the input pipeline. Two steps forward, one step back kind of thing. We have our best guy looking into it.

One thing I’ll ask, does the same thing happen when you use W, A, S, D instead of arrow keys? Because the software platform for Motion does treat them slightly different.

2 Likes

@wmoy thanks for looking at this. I tested it and can confirm WASD has the same behavior as arrow keys on CM 656 (still intermittent, happens roughly half of the time when I attempt to press and hold to jog at Speed 4). When it fails, it jogs a few inches then stops even though the key (arrow or WASD) is still held down. Happy to capture a video showing the reproduction if helpful.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.