A-M-C A12A100 Servo Drive with SouthWestern Ind ProtoTrak Plus

More
25 Feb 2021 16:37 - 25 Feb 2021 16:59 #200110 by new2linux
Todd, You have been more than generous with your time and knowledge. I am ever so grateful (more than very grateful)! PCW, I am just as grateful for your help as well. These drivers are new. If I were to move in to a model of driver that has (P3 connector) "encoder in put" it is a shore thing, correct. I may find $ticker $hock, but need to ask the question. Many thanks!! Please comment

Edit, Should I also think about the amp rating? The original 12a8 drives had all power I was in need of. It may help if I had a accurate description that was short in length to describe the primary issue: This is from earlier post by "PCW" it as I know it as to use the following to insure full 10 volt is going to drive:

"You can check that the analog +-10V outputs function properly by first disabling the drives
then launching LinuxCNC and running these halcmd commands in a terminal:

halcmd -kf
halcmd: delsig x-output
halcmd: delsig x-enable
halcmd: setp hm2_5i25.0.7i77.0.1.analogena true
halcmd: setp hm2_5i25.0.7i77.0.1.analogout0 10
(check for +10V on 7I77 analog out 0)
halcmd: setp hm2_5i25.0.7i77.0.1.analogout0 -10
(check for +10V on 7I77 analog out 0)

note that the "halcmd:" is a prompt printed by halcmd, not something you should type

Do you have your motor specs? 6.25 A peak sounds quite low "

The full 10 volt was delivered to drive, and insufficient voltage was delivered to motor?
Last edit: 25 Feb 2021 16:59 by new2linux. Reason: Ask question

Please Log in or Create an account to join the conversation.

More
26 Feb 2021 13:11 - 26 Feb 2021 14:48 #200191 by new2linux
Todd & PCW, Many thanks for the help. This component in the pic (the file is too large to attach) is the 1 part that has not been accounted for, during the encoder replacement. It is located inside with the encoder wheel on end of motor. This quote is what was said at the time. I have email asking the question, if this could cause the problems & if stocked item. Described the problem as: Why is the drive so sluggish with a full current command?

This is the quote:
"All that to say, I think you could use a single-ended encoder
directly, and bypass the circuit board with the DS8830N chip on it, it
doesn't sound like you need it at all. We do have a similar device,
that converts single-ended to differential, but it looks like it may
be unnecessary. I can provide a list of part numbers to do this either
way, completely single-ended, or with the differential line driver
added, it's up to you."

If there is any possibility this will fix and is a stocked item, I will buy them. I will try to get a picture attached. Many, many thanks for the help.
Attachments:
Last edit: 26 Feb 2021 14:48 by new2linux. Reason: add pic

Please Log in or Create an account to join the conversation.

More
26 Feb 2021 16:10 #200201 by PCW
Unless you are having physical position errors, that is , the machine accumulates
position errors after many moves as measured by a dial gauge. there is no reason
to be concerned about the encoder connections.
The following user(s) said Thank You: new2linux

Please Log in or Create an account to join the conversation.

More
26 Feb 2021 17:10 - 26 Feb 2021 20:13 #200207 by new2linux
PCW, Many thanks!!I had to be completely through in my effort to fix this, that was the only part that was not replaced or accounted for. thanks!
Edit: I have reached out to the AMC and was assured that this drive will work as advertised. Is there any chance that I have just not noticed the preferred tuning adjustments or wrong sequences in the method? many thanks!

Edit01: Is this a formal name or is there a different name for: "back EMF as a form of feedback for velocity control"? I have refereed to this "back EMF" as encoder input. Is this encoder input or is there another name or term that is used? many thanks, just trying to understand better.

Todd, If your offer is still available I would be very grateful for any help, with the understanding, at your convenience. Attached are the .ini & .hal files that I use. many thanks!
Attachments:
Last edit: 26 Feb 2021 20:13 by new2linux. Reason: Ask question, attach files

Please Log in or Create an account to join the conversation.

More
27 Feb 2021 16:28 #200308 by new2linux
Todd & PCW, Thanks for your help. I have come to the realization that the diagrams that I have been referring too, from the very 1st post at this location: forum.linuxcnc.org/12-milling/32294-prot...7-5i25-cards?start=0
I did not get diagram correct as to 7i77 card (TB5) pin 3 GND needs to go to location 4 +REF IN on the A-M-C drives (P1) signal connector and (TB5) pin 4 AOUT0 needs to go to location 5 -REF IN. It is my belief this is the issue that has always been an issue, only recently was resolved with the a12a100 diagram (this current post).

The rest of the story is that the drives ab15a100 at this post: forum.linuxcnc.org/12-milling/39963-repl...t-amplifier?start=60
The ab15a100 drives have a place on P1 (signal connector) on the drive (no separate P3 connector) for encoder input, pin 2 Signal GND, pin 6 Encoder-B IN & pin 7 Encoder-A IN. I am thinking to ask for exchange, to go back to AB15A100 for the encoder input.

I believe I should go back to the prior posts and edit the 1st page (at the start) and attach a good diagram and say to use this new diagram.

I deeply regret my error and will try to edit other post so others will have a chance with success. Suggestions welcomed.

Many thanks!

Please Log in or Create an account to join the conversation.

More
27 Feb 2021 17:21 #200312 by arvidb

new2linux wrote: Edit01: Is this a formal name or is there a different name for: "back EMF as a form of feedback for velocity control"? I have refereed to this "back EMF" as encoder input. Is this encoder input or is there another name or term that is used? many thanks, just trying to understand better.

An encoder is a physical device that senses position. Here's a description and some images: en.wikipedia.org/wiki/Incremental_encoder - The second image shows the insides of an optical encoder, with a thin steel disk with slots etched into it and an optical reader that detects the slots.

Back-EMF is a property of electrical motors which can be used to sense motor speed. See e.g. en.wikipedia.org/wiki/Counter-electromotive_force "In motor control and robotics, back-EMF often refers most specifically to actually using the voltage generated by a spinning motor to infer the speed of the motor's rotation, for use in better controlling the motor in specific ways."
The following user(s) said Thank You: new2linux

Please Log in or Create an account to join the conversation.

More
28 Feb 2021 13:26 - 28 Feb 2021 13:28 #200408 by new2linux
arvidb, thanks for your reply! My understanding is much clearer now. Recently I was talking to a sales representative about drives and I used this (back-EMF) thinking it as the encoder input, but this measures the motor current inside the drive, w/o using input from encoder, to monitor velocity of motor. The encoder is counting down as to when to turn "power off", so all moving parts can coast to a stop. The tuning of the drive to the motor is adjusting or allowing for all (motor, table, ball screw , etc) the friction and moving parts. many thanks!
Last edit: 28 Feb 2021 13:28 by new2linux. Reason: spelling

Please Log in or Create an account to join the conversation.

More
02 Mar 2021 16:00 - 02 Mar 2021 16:03 #200796 by Todd Zuercher
Ok, I've attached the double pid loop configs. (Hopefully without any typos or other mistakes.)

Tuning this, first you must tune the inner velocity loop. Do this by sending an open loop step command to it. The easiest way is to just set all of the position loop constants (P, I, D...) to zero except for FF1=1. then command some short moves. (A better way might be to set up a square wave siggen connected directly to the pid.x-vel.command but that is a lot of temporary hal configuring to do.) Increase the VEL_P to get the velocity feed back to most closely follow the commanded velocity without overshoot or oscillation (minimize the rize time). adding VEL_D if it helps with any oscillations. It is important that this is smooth without oscillating. After you are mostly happy with that. Then tune the position loop PID and FF terms following the instructions in the normal servo tuning tutorials.

File Attachment:

File Name: 2pid.hal
File Size:11 KB

File Attachment:

File Name: 2pid.ini
File Size:3 KB
Attachments:
Last edit: 02 Mar 2021 16:03 by Todd Zuercher.
The following user(s) said Thank You: new2linux

Please Log in or Create an account to join the conversation.

More
02 Mar 2021 16:09 #200797 by Todd Zuercher
However if you are going to revisit trying to use the other drives with encoder feedback to close the velocity loop in the drive. I'm sure it would give better results than doing it in Linuxcnc at the servo-thread speed. (Provided you hook it up and tune the velocity loop in the drive correctly.)
The following user(s) said Thank You: new2linux

Please Log in or Create an account to join the conversation.

More
02 Mar 2021 16:18 #200799 by new2linux
Todd, You are a total & complete gentleman all the way through!! Many thanks for hanging in there with me! For the additional cost of the AB15A100 drive that allow encoder (at P1) input, I am going to ask for exchange. This will take several weeks. I will edit other post as to be aware of my mistake. many thanks

Please Log in or Create an account to join the conversation.

Moderators: piasdom
Time to create page: 0.164 seconds
Powered by Kunena Forum