Trinamic PANdrive closed loop steppers.

dc42

External drivers with opto isolated inputs always require extended step pulse timing. See the T parameter of the M569 command on the GCodes wiki page.

marnog

@dc42 Thank you for the reply! Can you advise on a value for the step pulse timing that might work? If not, I will contact Trinamic support.

T3P3Tony

@marnog The trinamic datasheet should specify this.

marnog

@t3p3tony Thanks!

marnog

@dc42 & @T3P3Tony , I received feedback from Trinamic regarding the step pulse timing. They provided me with step pulse width and interval values which I set as per the GCodes wiki page (M569 P5 R1 T0.5:0.5:0.125:0), but still no movement.
Trinamic also advised that the drives have RS485 transceivers, and not a optocouplers. Could this make a difference?

NitroFreak

Try much higher values, i had a similar problem with my servo and the values needed for step pulse timing needed to be around 200-300%. if it works for like T30 then try to dial it down.

marnog

@nitrofreak I'm going to try that. 200-300% for the step pulse width and interval values only, or also for the direction setup time?

NitroFreak

@marnog said in Trinamic PANdrive closed loop steppers.:

@nitrofreak I'm going to try that. 200-300% for the step pulse width and interval values only, or also for the direction setup time?

I just increased them all, and it started working.
Specifically, they wanted 2.5 microseconds for step pulse timing and 5 microseconds for the direction setup time. When i set the duet to T3:3:6:6 it worked fine for slower speeds, but after 250mm/s i saw a small uniform layershift every step. So i just increased everything to T8:8:10:10 and it worked.

marnog

@nitrofreak said in Trinamic PANdrive closed loop steppers.:

@marnog said in Trinamic PANdrive closed loop steppers.:

@nitrofreak I'm going to try that. 200-300% for the step pulse width and interval values only, or also for the direction setup time?

I just increased them all, and it started working.
Specifically, they wanted 2.5 microseconds for step pulse timing and 5 microseconds for the direction setup time. When i set the duet to T3:3:6:6 it worked fine for slower speeds, but after 250mm/s i saw a small uniform layershift every step. So i just increased everything to T8:8:10:10 and it worked.

It didn't work. Trinamic advised 0.5µs minimum. I tried 1.0µs, 2.5µs, up to 5.0µs - nothing.

marnog

@dc42 & @T3P3Tony , I have been communicating with Dirk Heuer of Trinamic about this issue, but it still persists. Will it be possible for Dirk to contact you directly via e-mail to try to resolve this issue? If so, I will appreciate it you can send me your contact details via DM which I will forward to Dirk for him to contact you. I will truly appreciate it.

dc42

Sure, he can contact us. Ask him to get my email address from Mario or Timo at Trinamic.

The driver chips used by the expansion breakout board are similar to RS485, so the signal levels should be compatible.

If you are just seeing 1 lost step at each change of direction, then it is probably the step-to-direction hold time that is critical. Shortly after sending the leading edge of a step pulse, RRF updates the direction signal ready for the next step, taking care to satisfy the configured step-to-direction hold time.

marnog

@dc42 said in Trinamic PANdrive closed loop steppers.:

Sure, he can contact us. Ask him to get my email address from Mario or Timo at Trinamic.

The driver chips used by the expansion breakout board are similar to RS485, so the signal levels should be compatible.

If you are just seeing 1 lost step at each change of direction, then it is probably the step-to-direction hold time that is critical. Shortly after sending the leading edge of a step pulse, RRF updates the direction signal ready for the next step, taking care to satisfy the configured step-to-direction hold time.

Thank you very much David, I will relay the message and ask him to get in touch with you. I appreciate your help!