Comments on: ATOMI – the ultimate 4 axis CNC controller

By: Bernhard

Bernhard — Mon, 04 Dec 2017 13:42:02 +0000

I am very curious about your success with using back EMF feedback for sensorless stepper motors.

I found this page searching the web for “sensorless stepper closed loop”. The only commercially available drive seems to be from Nanotec (https://youtu.be/pRljKJFpcvw, although it is a little unclear which of their controllers actually support this mode of operation).
I would love to see any competitors…

By: jan beck

jan beck — Mon, 21 Nov 2016 10:36:11 +0000

Tero, thanks for pointing out those CUI units. I was not aware of them, and their pricing is great!

By: jan beck

jan beck — Mon, 21 Nov 2016 10:22:51 +0000

I have done that recently. If you are interested in the things I found, you can see here:
https://sites.google.com/site/janbeck/optical-table-cnc-lathe—motor-selection

In short, steppers seem to need higher resolution encoders than regular servos, because the poles are much closer together. Once that was done, I had a nice 4 phase servo running.

Sorry, Tero, I am using a SnapAmp for this one. But I felt the info is relevant to the question anyway…

By: Marco

Marco — Thu, 12 May 2016 21:13:56 +0000

Hi Tero,

Do you have any news about EtherCat board?
Keep Up the Great Work.

By: walter

walter — Sun, 31 Jan 2016 03:34:12 +0000

Great product. And thank you for the EtherCAT!

By: Tero

Tero — Sat, 30 Jan 2016 12:22:30 +0000

Well, using closed loop mode doesn’t increase stepper torque a bit (unless you configure higher current, which you could do as it runs cooler). But closed loop allows you taking 100% out of motor torque without risk of stalling. In open loop mode you need some 30-50% margin to avoid stalling so motor appears lower torque.

But it is true that max speed of stepper will usually be less than servo motor even in closed loop mode. Many steppers seem to run cloesd loop ~1.5-2krpm at 48V which is sufficient for most apps.

Sensorless operations should be definitely possible. I’ve already experimented with sensorless angle sensing algorithm for stepper and it seems to work quite nicely. It takes lots of more R&D to get some real results.

By: TonyT

TonyT — Thu, 28 Jan 2016 17:29:41 +0000

Based on my study of torque curves, I would say that driving steppers as high pole count servos improves the torque curve (basically the torque curve becomes flatter, giving you more torque at higher speeds), but it does not magically give them the same characteristics as a typical low pole count servo drive.

Stepper motor torque loss at higher speeds can be a serious problem. I recently had to cut the maximum speed of a stepper motor in half because it wasn’t reliably producing enough torque at those higher speeds. Unfortunately, I couldn’t run it as a servo because it doesn’t have a motor-mounted encoder.

Sensorless control for stepper speed should be possible, since it’s become routine for servo speed control, but for position control, you still have to figure out how to handle accurately keeping track of position.

By: Tero

Tero — Wed, 27 Jan 2016 23:03:12 +0000

Thanks for feedback :) If we continue IONIZER, it would be rather based on a control module that allows reduction of external components instead of IONI drive. Time will tell.

Also, EtherCAT support is on the way! It will be a converter board between EC and SimpleMotion so all current drives will work on it.

By: João

João — Wed, 27 Jan 2016 21:35:57 +0000

x2 for the IONIZER project, and for direct protocol of communication with linuxcnc (Ethercat maybe?)

About this project, looks like a well placed concept of product, if I need steppers drives I will look for theses :)

By: Klaus Stickdorn

Klaus Stickdorn — Wed, 27 Jan 2016 21:24:06 +0000

Could it be that you continue also IONIZER project again? It would be wonderful.
Greetings and a happy rest of 2016!