3.3V Logic for 5V Servomotor Optocouplers

[NitroFreak]

Hi,

I am using a 180W Servo for my Y axis.
Today i tested the setup and ran into a small problem, the Y axis seems to layer shift. It looks like missed steps, but that is impossible. The pulleys are also very solidly mounted.

This should be a benchy.

Then i printed a 20x20x60mm tower to avoid "curves" in the gcode, it turned out like this

Now i am strongly suspecting the optocouplers in the servo are not receiving clean signals.
I checked the datasheet and they want 5v or 24v, with a High of 3.5v.
At the moment i am running the servo directly from the board at 3.3v without the expansion board.

The weird thing is that when i move 100mm in the menu, it moves pretty much bang on 100mm.

Now, would i need to get the expansion header breakout board which outputs it to 5v?
Or can i also get one of those level converters?
https://www.exp-tech.de/module/stromspannung/5410/sparkfun-pegelwandler-bidirektional?gclid=EAIaIQobChMI49eEvOKi4AIVFM53Ch2hSwjDEAQYBCABEgIMM_D_BwE&fbclid=IwAR1Tsu0Aza-6VgJB0Lqm2HQxC1V-anbKJmK7ibih1uIFOxwAeWAseFykFOc

Thank you
Max

[NitroFreak]

I have now ordered the expansion board.
But is it theoretically also possible to hook the servo up to the 5V pin? Or, what would happen if i did that?

Thank you
Max

[dc42]

No you mustn't use the 5V pin, you might damage the outputs if you do.

Did you use M569 with the T parameter to extend the step pulse timing for your servo drives?

[NitroFreak]

@dc42
Yes i did not use the 5V pin.
Yes i set the pulse timing to a min of 3 microseconds, the manufacturer recommends 2.5 microseconds.
I have ordered the breakout board just in case the 3.3v are not enough.
Do you recommend shielded cables or are they not necessary?

Thank you
Max

[dc42]

You don't need to use shielded cables to connect the servo drives.

Check the DIR to STEP setup time of your servo drives. It may be longer than 3us. That would account for an occasional layer shift.

[NitroFreak]

I am pleased to let you know,
that the Breakout board solved the problem!
So it seems to have been the optocouplers.


The 3 microseconds seems to be enough for my speeds at the moment, i will let you know how it turns out when i push it.

Thank you
Max