Weird behavior of servo motor drive

This is about putting a gear motor on the axis.
I am very puzzled.
I have this servo gear motor mounted on the knee. It is a 65V motor,
4.4 amp continuous and 26 amp maximum. 15:1 gear reducer.
http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/34-Adding-Servo-Control-to-Knee/
It works fine and if I apply power to it from a large battery charger, it goes up and down very well.
The next thing I did is I connected it to a A-M-C 30A8 drive, the same kind as what I have on all other axes.
http://igor.chudov.com/manuals/A-M-C/30A8-Datasheet.pdf
This servo drive gets 72V DC power from the same DC power supply as all other drives.
I have NOT connected this drive to EMC2 at all, and I send it a analog signal from just a little DC power supply.
I tried torque mode and velocity mode.
It behaves VERY weirdly. If I tell it to move the knee down, it moves it a little (think 0.5 inch), and then stops and the knee actually comes back up (like by 0.1 inch)!
It does not move it up by more than 0.1 inch and does not come back, as in the opposite example.
I am struggling to find an explanation.
The only one that I can come up with about this going down behavior, is that perhaps the pneumatics of the pneumatic knee support is messing me up.
Or, perhaps, this is a bad servo drive.
I am at a big loss, any ideas would be most welcomed.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Mon, 28 Feb 2011 09:09:48 -0600, Ignoramus28206

Is the current limit set to an appropriate value?
Have you measured the actual output current via the current monitor pin?
--
Ned Simmons

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Ned, I will try to measure the current, indeed. I am not sure what exactly is wrong, but it is a great idea. I have another 30A8 and I can easily swap it.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

Suggest use an automotive battery charger that has an ammeter on it....connect directly to motor leads and note current draw going up and also when going down--adjust your counterweight so as to obtain an appx equal reading in both directions.
After that, your tuning should be a fairly textbook example unless you've locked the knee or weighted the table down with an engine block or other heavy item...
Suggest if you have a dc ammeter off of an old welder you can then place it in series with the amp motor leads for adjusting your offset and curr pots--after that you finally want to adjust gain for a critical or slight underdamp.
AMC has a much better .pdf someplace in their tech litereature--explains servo setup / tuning theory from a generic viewpoint sorry I don't have time to look for it at the moment maybe later.
--



Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

http://www.a-m-c.com/download/manual/AMC_AnalogDrives_InstallManual.pdf
--



Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Yes. Thank you. Note though, that I am not even trying to tune it. I was just trying to get maximum output out of it.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ignoramus28206 wrote:

Have you tried running the motor from the drive with the motor disconnected from the table?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

No, but this is a great question. I may do it.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Umm, yes you are....

Sounds to me like it's freewheeling when going down--to the point where it actually temporarily reverses as soon as you remove sig...but when going up, your not giving it nearly enough current.....
HORSE------->CARRIAGE
--you MIGHT be able get away to running a heavily imbalanced mechanical load in IR mode IF there is enough OFFSET adjustment available at the amplifier pot...but in my experience, probably not...
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

PT, All the AMC distributors I've looked up (so far) say, "Call us for a quote".
How much do those drives cost? (yes, I know there are varying capabilities in the line, but any thumbnail prices?)
Say, a 3-ph analog drive capable of 20A...
Thanks, LLoyd
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Feb 28, 2:14pm, "Lloyd E. Sponenburgh" <lloydspinsidemindspring.com> wrote:

From the discussion:
The motor works with a dc source. This implies the motor is working and is not overly loaded.
Problem has somthing to do with the drive amplifier. I have used older versions of these devices and they are almost bullet proof. I expect it is somthing to do with how it is hooked up or the jumper/ switch settings.
Make sure the motor and power supply is hooked up to the correct terminals. Don't ask me why I know to check this one.
Make sure the amplifer is set for current/torque or Voltage mode
Make sure the control signal is grounded correctly. ie positive to input possitve negitve to negitve and GROUND. Don't ask me why I know to check this one too.
Put a DVM on the current moitor signal and see if it is showing current
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Mon, 28 Feb 2011 16:14:12 -0600, "Lloyd E. Sponenburgh"

It's been a while since I've purchased an AMC drive, but around $400 for a DC powered 20A brushless drive, $500 for AC line powered, is in the ballpark.
--
Ned Simmons

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
$30 on ebay

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

AND

SO... we have a middle ground! Thanks.
LLoyd
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
OK.
I have fixed all servo drive issues.
1. Some bolt was too tight and there was way too much friction. 2. I set the drive to provide only 25% of peak (25% of 30 amps = 7.5 amps) and that was not enough. I changed it to provide 50% of peak amps, which is 15%, and now it works great. 3. Like I said, the drive is not yet connected to EMC. Partly this is because I have not had time, and partly it is because I need to first get a D/A converter card from Jon, which I forgot to order in time.
At least, what I have now, can be considered a motorized knee without the CNC feedback/positioning. I can move it up and down, by providing a signal from a little power supply, with a toy knife switch.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I mean 15 amps

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
"Lloyd E. Sponenburgh" <lloydspinsidemindspring.com> wrote in message

Dunno I usually buy them off a ebay on as-needed basis for automating dedicated second op machines
--AMC, Servo Dynamics and Copley come up there quite frequently and since they are basically interchangable.....
As to surplus 3 phase IMO you might just as well make up your own power supply--either buy one that's setup for DC or buy a single phase one and jerk off the existing power supply since they are usually just a stock dc unit with the addition of a bridge rectifier, cap and bleeder mounted onto the heat sink.
--



Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Feb 28, 7:09am, Ignoramus28206 <ignoramus28...@NOSPAM. 28206.invalid> wrote:

This sevo amplifer requires analog signals for both command and velocity feedback. I assume the motor is brush dc motor not a bushless dc motor. Two drive wires vs three.
To test the amp/sevo motor I would set the amp for current mode. Make sure all the jumpers are set correctly. Also make sure the ground and common connections done correctly. IIRC the manual was a pain to figger out jumper settings.
If you want to implement a velocity control mode in the amplifier you are going to need a analog tac on the motor.
I have used an early version of this amp for a dc brush motor drive and it worked well.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/34-Adding-Servo-Control-to-Knee/
You need to adjust your counterweight or connect active feedback before it can be properly tuned.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Well, I thought, that with a torque mode, I can just move it at ful torque (full speed really) when I send 10v to the input. And that is NOT what is happening.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.