Loosing steps (?)

Dear All,

could you please help me troubleshooting a problem?

Inputs:
Boards: arduino uno, arduino mega, GRBL firmware installed on them
Motor driver: DM542
Steppers: NEMA23, NEMA17
Sender sw: GRBL candle
Test Gcode: a simple gcode that creates a d=3mm hole with 3.2 mm depth by 0.1mm steps, and with G1 commands only. (No G2/G3) The test takes about 1.5 mins. I run it without any load, and with the motor only.

Result:
The motor on Z axis doesn't go back to the starting position. I marked the shaft with a pen, and recognise ~20 degrees differency. I tested it with 3 NEMA23 and 1 NEMA17, the results are the same. At this point we could think the problem is the motor driver, BUT if I run the commands from the Gcode file that contains only the Z movement, it's perfectly accurate! If it contains the X-Y movements as well, it fails! I tried to run the test with only one driver and one motor only, but I experienced the same problem. The longer the project, the bigger the inaccuracy. I tested it on 2 types of boards, and the results are the same as well.

What the hell is going on? :) Have you experienced something similar? Any idea?
Thanks for your help!

Hi,
steppers losing steps is the never ending hazard of steppers.

All stepper motors lose torque the faster they go, it just plain physics, there is no avoiding it. One of the specifications of a stepper is its
inductance, and the lower its inductance the better. This is because the degradation in torque is least with low inductance motors.

Manufacturers of steppers try to make the highest torque motors, but they also have the highest inductance. First time buyers don't
know about inductance and get trapped with steppers of great torque at low speeds but can scarcely get to 500 rpm.

What is the inductance specs of your motors?

Ideally 23 sized motors would be 1mH-2mH, 1mH preferred and reject anything over 2mH.
17 size steppers should be 1mH-1.5mH.

The classic way to counteract the negative effects of inductance is to use the highest possible voltage drivers you can get.
What is the maximum voltage your drivers can tolerate? What voltage are you actually using?

Craig

Thanks Craig for your reply.

First of all the dry parameters:
Motors are: SMNM23S18KG31 and SMNM23S18KG13, and 42H33HM-0504A2-12
Specs for the first one is here: https://reprap.world/products/motors...p_30_59_kg_cm/
My drivers can handle 18-50V, and I use a 24V (switching mode) power supply.

I absolutely agree with your opinion and with the rules of physics.

But the most strange part of this problem is that, if I run the mentioned two tests:

1) the same gcode with only Z axis commands
2) the same gcode with Z +XY axis commands, but the Z is exactly the same as in the first example

it fails at test 2), but it's perfect at test 1).

The hardware, software, the settings are the same for each. Only the Z motor is wired, and the same motor, the same driver, the same voltage, same current, same wires, everything.
The only difference is that it sends other commands to X-Y motors between the commands for Z. I have to highlight that only the Z motor is connected for the test. (and I tested with other driver, other motor as Z as well, and the results are the same.)

So what can be an explanation for this from this aspect?

Thanks,
Peter

Hi,

Quote:

---

So what can be an explanation for this from this aspect?

---

Yes that is unusual and does rather suggest some other fault. Whether indeed there is some other reason there is no need to deliberately hobble
your machine by using a paltry 24V supply. If the drivers will handle 50V then use a 50V supply.

When you go to replace those steppers don't buy ones with 5mH inductance, that's a joke, 2mH or less is what you require. Your existing 5mH steppers
will probably start missing steps at 200 rpm with such a low driving voltage.

Craig

Thanks Craig,
I have just bought a 48V power supply, I'm going to test the system with that.
Thanks for your advice!
P