Nema 23 problems



  • Hey guys. About a month ago i was finally able to get my printer to a good working state - or so i thought. I started printing smaller things without issue, now I’m beginning to push the boundaries a bit and print bigger. The issue is, I’m starting to get later shifting. The shifts always occur in the Y axis, where I’m using a Nema 23 geared down with a 3:1 reduction, which drives a shaft and two g2 belts to slide a gantry (I can post photos if needed).
    At first, I thought this was a mechanical issue. My print head sits very close to the build plate (only a couple mm higher than the lowest tip of the nozzle), so I thought some portion may be catching, but the Nema 23 is pretty tough, and I have to apply quite a bit of force to cause the gantry to shift (and even when I do that, the belts slip before the motor skips).
    Most recently, I’ve decided that maybe the motor is getting too hot and is not behaving optimally, but I can’t imagine it’s getting that hot without causing any damage or permanent flex to the ABS mount it’s bolted to. Plus, I’m only running it at 1000 - 1200 mA (the motor is rated for 2 or 3 amp max, I believe). I would increase the current, but as I’ve said already, the belts slip before the motor stalls, so I don’t think adding MORE torque would really do anything.
    Does anyone have any ideas? I’m lost and it’s starting to get frustrating because all I want to do is print big! Haha.
    Thank you in advanced for any and all help (and I apologize for the super long write up)


  • administrators

    Have you checked whether the supply voltage is sufficient for the maximum travel speed you have configured, bearing in mind the motor specifications and the gearing you are using? The gearing will increase the motor back EMF and hence the supply voltage you need. Use the calculator at reprapfirmware.org to check.



  • I’ll double check, but I think I should be fine. I’m using a 24V 600W Meanwell power supply. I’ll probe my supply voltage during some fast travel moves and see if I get any dips in voltage.


  • administrators

    @wcj97 said in Nema 23 problems:

    I’ll double check, but I think I should be fine. I’m using a 24V 600W Meanwell power supply. I’ll probe my supply voltage during some fast travel moves and see if I get any dips in voltage.

    No, it's not dips in voltage I am concerned about, it's the rotational + inductive back EMF reaching the supply voltage.



  • This post is deleted!


  • The easiest thing is to try upping the motor current. If the motors are rated at 2 or 3 amps then you should try 1700 mA which is 85% of 2 amps. When you push the gantry manually with the motors powered up, you are testing the holding torque but this will drop significantly at speed. Anyway, increasing the torque is the quickest and easiest thing to try.


  • administrators

    Increasing the motor current is easy to try, and it will help if the problem is insufficient torque and the supply voltage is sufficient for the requested travel speed (as limited by the M203 maximum speed) and motor current. It won't help if the problem is insufficient driver supply voltage to overcome the motor back EMF.



  • @deckingman i’ve Tested against the torque at speed as well. In both situations (holding and moving at considerable speed) the belts slip before the motor stalls



  • @wcj97 said in Nema 23 problems:

    @deckingman i’ve Tested against the torque at speed as well. In both situations (holding and moving at considerable speed) the belts slip before the motor stalls

    OK, fair enough. Never seen a belt jump teeth when moving a gantry at high speed myself.



  • @dc42 so how can I use the back emf info from the calculator to help? I’m already using a 24V power supply, so I can’t up my supply voltage because I would surpass the Duet’s input range. Are there diodes I can place in series with the motors to prevent back emf? Should I swap my motor for a Nema 17? I initially thought the smaller motor couldn’t handle the weight of the gantry, but I think now that a Nema 17 would be sufficient. Would using the smaller motor make my problem worse? Or could it potentially help?



  • @deckingman the belts are stretched over a 600mm span, so even with a lot of tension there’s potential for enough slack to slip, but even so, I have to apply quite a bit of force to make the belts slip during printing - way more than what would be applied to the gantry during a typical print.


  • administrators

    @wcj97 said in Nema 23 problems:

    @dc42 so how can I use the back emf info from the calculator to help? I’m already using a 24V power supply, so I can’t up my supply voltage because I would surpass the Duet’s input range. Are there diodes I can place in series with the motors to prevent back emf? Should I swap my motor for a Nema 17? I initially thought the smaller motor couldn’t handle the weight of the gantry, but I think now that a Nema 17 would be sufficient. Would using the smaller motor make my problem worse? Or could it potentially help?

    The first step is to use the calculator to see whether motor back emf is limiting the torque at the travel speeds you are commanding. If it is, then you have a choice of lower maximum travel speed, lower motor gearing, or lower motor inductance.



  • @wcj97 said in Nema 23 problems:

    @deckingman the belts are stretched over a 600mm span, so even with a lot of tension there’s potential for enough slack to slip, but even so, I have to apply quite a bit of force to make the belts slip during printing - way more than what would be applied to the gantry during a typical print.

    Fair enough - as I said, just trying to help. Hope you find the answer.

    Oh, BTW and for info, I throw about 3kgs of mass around at 350mm/sec non-print speed on frame that is 700mm square and never seen a belt jump teeth (which it would need to do in order to slip).



  • UPDATE: for anyone who was following this thread, I believe the issue was indeed problems with back EMF. I decided to swap my NEMA 23 for a smaller NEMA 17 I had lying around, which had a much lower inductance (also a smaller rotor so less rotor inertia). The smaller motor is providing plenty of torque. I’ve been running prints now for a little over 24 hours with the same accel/jerk/speed settings for the smaller motor and so far have had no issues. I’ll post another update in a few days after I can run some more prints and see if the layer shifting is totally eliminated. On another note, I think using the smaller motors is all around more beneficial. My printer is totally silent now (those trinamic drivers are crazy!), and the motor now skips before the belt, so I can implement stall detection which I’m excited to try out.



  • 12 hour print down with no issues! Seems to have fixed the problem (knock on wood)! Thanks for all the help guys!