Set up using AMC AB15A100 drives, brushed DC motor Prototrak Plus with Encoders

More
01 Jul 2022 17:54 - 01 Jul 2022 17:56 #246387 by new2linux
Todd, thank you for all your guidance!! The attached trace is with P=24.4745, this is as high as it will go with feed set to 20 ipm. I tried to set feed to next increment 22 ipm, but only motor jerk resulted. File name seta=pot1 11.562; pot2 11.5; pot3 3.312; uf 1.517 Units: all50M/div except f error as shown in trace = 5m/div.
Many thanks!!
Attachments:
Last edit: 01 Jul 2022 17:56 by new2linux.

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

More
01 Jul 2022 18:12 #246388 by new2linux
Todd, Thanks! I have not checked brushes, after looking just now it looks like a regular/general brush style of motor. Has small slot for screw driver in plastic housing, 90* apart. I will look at them over the weekend & be fussy to re-assemble just like taken apart.
As for the weekend, I will check the brushes, try to tune this set up, better. If no luck I will go back to the 1.617uF settings & work through the pot1 & 3 (leave pot2 alone) again, working for a higher feed rate, but prepared to settle for 20ipm. Unless I find a power supply that will work with the rest of the parts.

Todd, I hope you and your family the very best over the holidays!! Many thanks

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

More
01 Jul 2022 19:05 #246391 by Todd Zuercher
Try changing the max acceleration setting in your ini file from10 to 7 and see what happens. You will have to shut down and restart Linuxcnc after making such a change to your ini file for the change to take effect.
The following user(s) said Thank You: new2linux

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

More
05 Jul 2022 14:02 - 06 Jul 2022 15:38 #246648 by new2linux
Todd, thanks for all your help. I changed the .ini file after cleaning the brushes from the "X" axes motor. The pic shows what I believe to be left over coolant (pic to big) that got thick and sticky. Used contact cleaner to clean brushes and holders. Trying to tune clean motor with ,ini file edit, starting low and with factory settings, may swap motors from x to y.
A-M-C suggested a power supply, I am looking at them now.

Many thanks

Edit: The excellent support from A-M-C had this suggestion:

"My previous suggestion still stands as there needs to be sufficient voltage to supply overcome the back emf of the motor.
A recommended AMC power supply for this drive would be the PS16L72 orPS16L72-L.
It is a 72VDC power supply capable of providing sufficient current to your motor though the volage supplied the motor would be above the motors voltage rating, the reference command will need to be reduced to accommodate for the extra voltage."

After re-reading, it almost sounds like the "back emf" type of design, has a way to store the energy & uses it to assist the drive & motor when changing direction of motor. I am expecting delivery of a PS16L72-L tomorrow, late in the day. I plan to mount it in the electrical cabinet & get it wired, so tuning can continue. The plan is to leave POT2 at, I believe 11 turns (as per this post #246301 ). All others (sw2 & 3) back to factory settings, starting with POT1 until motor squeals, back off 3 turns, move on to POT3.

Many thanks

 
Last edit: 06 Jul 2022 15:38 by new2linux. Reason: Up-date & why

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

More
06 Jul 2022 19:39 #246752 by Todd Zuercher
I'm still doubtful changing your power-supply will help much, (but it won't hurt anything either.)

Back EMF is a property of every electric motor, and is what cause the motor to stop spinning faster and faster for a given voltage applied to it. The back EMF of a motor does not change. If the old power-supply had sufficient voltage before, it still should be able to spin the motors at the same speeds as before. Changing the drives like you did would have no influence on either the supplied voltage of the power-supply or the back EMF of the motor.
The following user(s) said Thank You: new2linux

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

More
08 Jul 2022 15:24 - 08 Jul 2022 16:41 #246869 by new2linux
Todd, thanks for your help. The new power supply has been installed (needed to strip the electrical cabinet so I could welded up old holes) and started the tuning process.
Please give a second opinion, I am questioning my process, completed are the Pot1 turn CW until motor squeals, back up 3 turns; Pot2 is set to 3 turns CCW, from full CW (as per calculation); Pot 3 turn CW until motor comes around, if I turn this too much the motor will start to drift & seems to pickup speed.

Now with FF1 set to 1.0 (all others set to 0.0) feed set to 20"/min, I need to get hm2_5i25.encoder.00.velocity (purple trace) to match trace axes.0.joint-vel-cmd (red trace), using Pot3 & SW3-6, 7, etc...
If this seems, ok I will continue.

Many thanks

Edit: After checking & re-check, it appears something is going on with "X" axes, at the drive, I have swapped motors, same thing happens, the motor will oscillate just as soon as F2 is released, I have feed rate at lowest and still motor turns quickly about 5 turns (hard to tell) & reverses direction does the same, the joint error will appear or use F2 to stop. I measured voltage at the power supply 74.9 VDC, checked the small transformers, reads 5 & 24 VDC. I am so disappointed with myself, the tuning was going ok & then could not figure out what has occurred. Only thing I can come up with is as you study SW3-1 TO 5 options (every option is used) you see 2x as many as the SW3-6 to 10 (some options are not used), and maybe I used settings that were not available. This is only thing I can come up with as why & what was I doing.

many thanks
 
Last edit: 08 Jul 2022 16:41 by new2linux.

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

More
08 Jul 2022 16:26 - 08 Jul 2022 16:28 #246881 by Todd Zuercher
Sounds about right. Is that what the AMC tech recomends?

I'm not entirely sure what you mean by turning pot3 till the motor comes around.

If you still run into the jerk problem as before, I think it will probably be a sign that proper current loop tuning may be required. But we might be able to try blindly turning on additional current loop integrator switches (SW3 6-10) to see if we can see improvement. (In other words test flip a switch, test, re-tune see if there is improvement or if it is worse or no change, then repeat testing and tuning with another switch.)
Last edit: 08 Jul 2022 16:28 by Todd Zuercher.
The following user(s) said Thank You: new2linux

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

More
08 Jul 2022 16:45 - 08 Jul 2022 18:38 #246882 by new2linux
Todd, thanks you! When I talk about motor comes around, if low Pot3 setting, motor does not seem to want to run at all when arrow key is used & too much & motor drifts quickly, Pot 4 will change direction of motor but usually over shoot adjustment & motor wants to run other direction.
many thanks

Edit: Please review my prior post about oscillation, upon push F2 down.

For example SW3-1=ON;2=OFF;3=ON;4=OFF;5=ON; is not available. See attached pic of table.

Edit: I have re-edit the .ini file back to 10.0 from 7.0.

Edit: I have swap the drives, I plan to re-tune to motor to this drive. This is from my earlier post:
"After checking & re-check, it appears something is going on with "X" axes, at the drive, I have swapped motors, same thing happens, the motor will oscillate just as soon as F2 is released, I have feed rate at lowest and still motor turns quickly about 5 turns (hard to tell) & reverses direction does the same, the joint error will appear or use F2 to stop. I measured voltage at the power supply 74.9 VDC, checked the small transformers, reads 5 & 24 VDC. I am so disappointed with myself, the tuning was going ok & then could not figure out what has occurred. Only thing I can come up with is as you study SW3-1 TO 5 options (every option is used) you see 2x as many as the SW3-6 to 10 (some options are not used), and maybe I used settings that were not available. This is only thing I can come up with as why & what was I doing.
I believe Pot3 by going from 5or6 to 4.0ccw has reduced the oscillation some. I pray this is just a tuning issue.

Many, many thanks
Attachments:
Last edit: 08 Jul 2022 18:38 by new2linux.

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

More
08 Jul 2022 18:35 #246892 by Todd Zuercher
You may have to put all of the switches back to their factory defaults, then do the base tuning for pot1 and pot3, and start over from the beginning tuning the velocity loop in the drive (Linuxcnc open loop with FF1=1 and all other PID settings =0.) The settings you found that worked before may need to be very different since the voltage and currents the drive sends the motor will be different than before.

I don't understand what do you mean by "For example SW3-1=ON;2=OFF;3=ON;4=OFF;5=ON; is not available."? Why wouldn't it be available? It's the 11th column 110kOhms.

At this point I think I'd leave the SW3 switches in their factory default positions 1=off, 2-5 on and 7=on, 8-10 off.

Then after tuning the SW2 switches and the pots, if you still can't get good results, we could try some things with SW3-7-10 (leave 1-6 in their default setting). I would not recommend blindly trying things with SW3 switches 1-6 without looking at the motor current feedback with an o-scope.
The following user(s) said Thank You: new2linux

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

More
08 Jul 2022 18:53 - 08 Jul 2022 19:26 #246896 by new2linux
Todd, Many thanks!! Am I not reading it correctly; the SW3-6 to 10 (the very bottom does not allow the opposite of .1047 uF (this is: off,on,off,on,off,on). It does not have option (on,off,on,off,on,off). The SW3- 6 TO 10 table is 1/2 the size of SW3- 1 TO 5. Also the units in the 3 tables have 2 different units types, SW2 & SW3-6 TO 10 have uF; table SW3-1 to 5 uses units k (omega) (this is the resistance measurement unit?) as the units to describe what the SW3 does.
I plan to pray tonight & start fresh in the morning, re-set everything to factory settings with the drive and motor swap. Tune pot 1&3 each time to any change made to drive dip settings, with power off.

Many thanks

Edit: Todd, I think your karma counter is stuck, FYI. Many thanks for all your help.
Last edit: 08 Jul 2022 19:26 by new2linux.

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

Moderators: piasdom
Time to create page: 0.128 seconds
Powered by Kunena Forum