I had a working setup of my MPCNC a few weeks ago. Since then, I decided to redo the wiring to the steppers. I tested all the wiring harnesses when I was finished making them, plugged them into the controller, and verified that the gantry moves as intended.
When I went out to work on it yesterday, I suddenly encountered an issue where the gantry will try to move and instead starts making loud noise and vibrating badly. Initially it seemed like it was only the X-axis, but then it started happening on the Y-axis as well. If I give the gantry a gentle push while it is trying to move, it sometimes will catch and start moving, but often it just sits there and shakes. It looks like the stepper motors are maybe skipping steps, but I can’t for the life of my figure out what happened.
Anyone have advise on what to investigate? It is built with the mini-Rambo and serial stepper drivers.
Has something frozen up? Are all the bearings spinning freely? If you turn off the power (and maybe unplug the motors or disengage the belts), can you move the gantry easily by hand? Is the spindle buried in the spoil board? Are all the pulleys tight to the motors? Did you experience a drastic temperature change recently that may have inadvertently adjusted some of your hardware due to shrinkage? (Shut up, Jeffe…)
Marion, I’ve tried checking all the wires. I did make new serial wiring harnesses to replace the really long ones I ordered from the site. They all tested good when I made them, and the movement was fine then. I’ll double check and make sure I haven’t caused any issues.
When the power is off, I can push the gantry around without any issues. I don’t hear anything grinding, and the only resistance I can feel is from the magnets inside the steppers. Spindle isn’t an issue as it is pulled all the way up. And the pulleys seem to be properly tightened. It isn’t skipping teeth on the belt.
The weather has cooled a little bit, but I doubt it would be enough to affect the motors.
The grinding is from the steppers skipping steps. There are 4 positions, A,B,-A,-B, then it repeats. When the steppers is trying to get to A, and it gets pushed past -A, it will skip to the next A, and that quick skip sounds like gears grinding (to me, I’m not a mechanic).
That can happen easily, as Marion pointed out, if one of the coils isn’t making a good connection. The connection can get worse in shifting temperatures, because even the metal in the connectors will get shrinkage (shut up kv…). That is pretty rare though. The mini rambo controls the current by software, so that shouldn’t have changed.
You should be able to check the motors and the wiring by testing the resistance of each coil at the mini rambo. Make sure you power off the MR before removing those wires. Be sure to also jiggle the connectors a little bit. I know you said it wasn’t a wiring problem, but it really sounds like it. Gremlins like that are hard to find.
If it definitely isn’t the wires, then maybe it is something mechanical, but you said it moves well enough, and it is just air movements, so it should have a lot of headroom for friction.
Got some time to troubleshoot last night and it’s looking like the Y-axis wiring harness has an issue of some kind. X was working great last night when connected to either the X or Y drivers, so the drivers are good. Looks like I’m going to have to re do the ends on the wiring harness to make sure they are nice and solid.
It’s maddening. My first setup the axes were not moving right and it was so strange what was going on. Everything seemed good but finally found one of the connectors to the endstops was shorting out occasionally because of some bad soldering on my part and that created havoc with everything.
I have been wanting to post this a few times, and I thought you could use a laugh. I think it applies many times:
One of the things I really like about the tmc drivers is that you can get a report if you enable TMC debug. It will print out if the driver thinks it has an open loop or a short circuit, and even which loop.
On my previous controller, with dumb drivers, I would sometimes just hold the gantry back while it was moving around and I could feel the difference between a good amount of torque and when something was wrong and it would skip steps. I found a wiring issue from that, and it was only when the wire was stretched out across the whole table. I also get in the habit of tugging a little on the back of dupont connectors. If it’s a bad crimp, the cable will come right off. Just a tiny tug.
That’s my art/magic. No science there, but those are my superstitions.
I’m not sure how many connectors I’ve been through trying to get things working. One trick I did figure out on them not fitting in the shroud was to give them a little squeeze with the pliers. Crimping them seems to flatten them out just a little to much. The squeeze closes them back up so that they fit.
I bought the wiring kit from Ryan, but decided I didn’t like or trust the connectors. Also, I used 18ga wire (I think the kit was about 22ga?) just for added mechanical strength (and because I had some). I’m sure the 22ga is adequate electrically. I ended up cutting off the connectors and soldering most of the electrical connections. Soldering creates a great electrical connection, but shouldn’t be done on any portion of the wiring subject to flexing, as the solder creates a “hard spot” and eventually the wire right next to the solder joint will break.
Yeah, I’m working with 22ga 4-connector ribbon (sold for LED strips). It’s too big for the little crimp (.25mm), but too little for the middle crimp (.5mm). At least for a dead-simple crimp. Plus, those damned wings never want to do anything useful. If I squish them enough to nicely go into the crimper, they’re too small for the insulation.
Thanks for everyone’s suggestions and support. I tracked it all down to wiring issues. Had to replace the connector that plugs into the board, resolder a loose wire in the plug, and replace the connector in one of the steppers due to intermittent connection issues.
With all those fixed, it is running beautifully again. I have it pretending to cut out some stuff so I can verify the full operation.
Reprinted the rollers for the x-axis, put new bearings in them, and installed them on the machine. No dice still skipping. After standing there running tests, my dad noticed that it seemed to skip whenever the connector bounced against the bearing. Turns out that was the problem. Fastened the connector out of the way and it works great!
Thanks for all the help everyone. Was finally able to get my job run to setup the spoil board.