[mcardoso]

Hi All,

I bought an old (ca 1999) Seiko SCARA robot just to mess around with as a hobby, It was very inexpensive but didn't come with a controller. I have a lot of servo equipment to play around with and my plan was to hook it up to some Allen Bradley servo drives and run the motors that way. The motors are Yaskawa SGM series and have incremental encoders. More (possibly irrelevant) details are in my project log here: https://www.hobby-machinist.com/thre...ra-robot.84001

What I am stuck on is the commutation signals (or more correctly the lack there of). I am used to incremental encoders having outputs for UVW alongside the normal ABZ signals. These encoders don't seem to have that. Any idea how Yaskawa handles commutation on these motors? I can replace the encoders to work with my drives if I need to, however I'd rather understand how their system works before I do that.

Thanks!

Mike

[mactec54]

Hi All,

I bought an old (ca 1999) Seiko SCARA robot just to mess around with as a hobby, It was very inexpensive but didn't come with a controller. I have a lot of servo equipment to play around with and my plan was to hook it up to some Allen Bradley servo drives and run the motors that way. The motors are Yaskawa SGM series and have incremental encoders. More (possibly irrelevant) details are in my project log here: https://www.hobby-machinist.com/thre...ra-robot.84001

What I am stuck on is the commutation signals (or more correctly the lack there of). I am used to incremental encoders having outputs for UVW alongside the normal ABZ signals. These encoders don't seem to have that. Any idea how Yaskawa handles commutation on these motors? I can replace the encoders to work with my drives if I need to, however I'd rather understand how their system works before I do that.

Thanks!

Mike

It will depend on the part number on the motor they came with incremental and absolute, it was a choice of the user as to what they have, if it is for the robot then they are most likely are absolute

[mcardoso]

It will depend on the part number on the motor they came with incremental and absolute, it was a choice of the user as to what they have, if it is for the robot then they are most likely are absolute

All four motor encoders are 2048 P/R incremental encoders. The encoder model is a Yaskawa TRD-Y2048.

The motor models vary by size but the catalog coding is consistent with incremental encoders.

The robot must be homed to optical switches every time it is turned on.

[mactec54]

All four motor encoders are 2048 P/R incremental encoders. The encoder model is a Yaskawa TRD-Y2048.

The motor models vary by size but the catalog coding is consistent with incremental encoders.

The robot must be homed to optical switches every time it is turned on.

So what happened to the Yaskawa servo drives ??? I did have information on these encoders but can not find it right now, they are not easy to connect to another servo brand but can be done

I found this post from some years ago that may give you some input on how to get them running

https://www.cnczone.com/forums/servo...awa-motor.html

[mcardoso]

So what happened to the Yaskawa servo drives ??? I did have information on these encoders but can not find it right now, they are not easy to connect to another servo brand but can be done

I found this post from some years ago that may give you some input on how to get them running

https://www.cnczone.com/forums/servo...awa-motor.html

The original drives were included inside the SRC320 robot controller sold by Seiko. They were open packaged and may or may not be Yaskawa drives without the plastic case around them. It would cost me $800-1000 to buy a used and potentially faulty control, cables, teach pendant, etc. and even after all that only have some basic robot functionality. This far exceeds the $200 I put into the robot. Not saying it can't be done, but my hope is to use the kinematic transforms within an Allen Bradley ControlLogix PLC. This gives me a lot more functionality than the original system has and I already have all the controls hardware so no cost added.

If you could find any info on these encoders I would really appreciate it. I do have the option to replace the encoders with aftermarket commutation encoders but again that is a fair added expensive and a lot of work to align them and rewire the robot.

Short of understanding the commutation on these encoders, the next thing I am looking into is enabling self-sense commutation on my Kinetix 2000 drives. The manual indicates that this feature is available, however I have never set it up with this particular drive. It might be coded into the motor database file.

That writeup was a great read and sounds very similar to what I am trying to accomplish. The key difference is that @laserted had an encoder that had wires for the commutation signals while mine does not. I think I'd be in business if I had the signals he did.

Thanks for the response and the help. I feel confident that I'll get this running and this is the major barrier in my way.