Initial Setup Troubles

[ejg3855]

I've finished building my mechanics and now I am working on getting the ABXX-E setup with Teknic Clearpath Servos that use step and direction signals. I can set my resolutions to anything I want within reason with the Teknics. I am using 200steps/rev currently. I have tuned the gains on the servos so they are running very well.

When I run the motions using the Teknic software everyting moves like it should, Video 1. https://youtu.be/-b05c4El9ZY

When I connect the axbb-e and UCCNC control the motor runs like complete garbage, Video 2. https://youtu.be/rrbqihcPoa4


I feel like I need to be missing something very simple here, or I have a bad board (abxx-e). Anyone shed some light on this for me?

Thanks

[ger21]

Try running them at the native resolution, rather than 200 steps/rev.

[ejg3855]

Try running them at the native resolution, rather than 200 steps/rev.

I'm not sure they have a native resolution. I tried 800 step
/Rev and got the same results.

[cncdrive]

I think your comparion is not OK. You running on low speed with the UCCNC and running on high speed with the servo config software.
Try to also run low speed with the servo config software, around the same speed as you running this with the UCCNC.
I'm asking this to see if this is a PID tuning issue or not, I guess it is.

Also if you have an oscilloscope or logic analyser then check the step signals to see if their frequency is smooth or not.

And yes, try to use higher resolution. I'm not familiar with the config software, but maybe it runs on high resolution encoder feedback, so it could be a difference.

[ejg3855]

I am not sure but I am reading the natural encoder resolution is 6,400 on this model. I can switch to that.

When I run the motor direct (usb teknic) at a slow speed it also runs just fine.

I do get a "step input timing error" on the servo when I try to move faster via the UCCNC. I think that means its feeding steps much faster than the servo can process.

[ger21]

What frequency do you have UCCNC set to? Try using both a faster and slower frequency, to see if it makes a difference?

[ejg3855]

There was a noticeable difference when I switched from low hertz to high hertz and there was a noticeable decrease in the studder of the system, it seemed like it wanted to go faster than the 400mhz

I am more puzzled as to why it won't operate in the (-) direction.

I am going to bring a scope home from work and try to evaluate the signals this weekend.

[Robertspark]

...400mhz...

wow

[ejg3855]

it was a typo lol I wish it went that fast.

[ejg3855]

It was a logic grounding issue. Solved!