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

The SW2 bank of switches turn on/off a series of capacitors that control the integration time constant of the velocity loop. Pot 1 (I think or 3) is a variable resistor that controls the Proportional gain for the velocity loop.

SW3 is two banks of switches that control the gain constants for the current (torque) loop. Switches 1-5 are for an array of resistors, measured in ohms (k omega is short for 1000 ohms), that are analogous to Pot1. Switches 6-9 are an array of capacitors analogous to SW2, SW3-10 is a shunt that bypasses all of the capacitors in the current loop, and would be the comparable to setting either I=0 or infinity (not sure which.)

Todd, thanks for all your help, especially the SW2 vs SW3 clarification. I have many pics of the ways I have tuned Pot1 & Pot3 to each SW2 setting, but only attached 1. The file name is the settings, example: s.047in7p1 10.0p2 3.0p3 9.0 fd20.png is sw2=.047,ini(file) set to 7, pot1=10.0, pot2=3.0, pot3= 9.0, feed 20"/min. Be careful of units, units in each pic, uses same units unless 1 trace needed different scale, then it's white, so you can read scale from pic. The motor has not jerked. I am not seeing the pattern.

Many thanks!

Those are pretty much universally awful. Did you switch the drive from velocity command to current?

Todd, Good morning, thank you for all the help. I have ponder what I may have set up or wired improper. The A-M-C power supply was a flying lead style, so the capacitor has 2 screws (marked +-) with short length of wire that I think would go to a buss bar type connector, so you can supply power to many drives. I have crimped & soldered 2 (1 for x & 1 for y axes) sold copper 14 wire using a crimp on style of terminal and screwed under each screw in the capacitor. Does this sound ok?
When I started tiring to tune, I had a ground wire off the drives & if seemed that if you hold one motor shaft (each motor spins at varying slow speed) the other motor would spin faster if applied friction to first motor.
This morning, I have rechecked all wires going in the back of plugs to be shore all is secure.
I am planning to start at the beginning, all factory settings. It seemed Pot1 wanted to be around 10 to 10.5 turn, no matter what SW2 setting. I would of thought that a better rated power supply would bring this down to 7 or lower, because of the better match. Pot3 seemed to like 7 to 10 turns, this pot3 CCW brings the purple trace down.

Today I will try with FF1 =1.0, and work back through the settings again.
Many thanks
Edit: No I have not switched to current command, all SW1=off (1 to6).

Actually current mode would be SW1-1 off, and SW1-3 on. For velocity mode SW1-1 and SW1-3 should both be in the off positions.

Todd, thanks! I have the proper settings for velocity mode, SW1- 1 & 3=off. This morning starting with sw2 & sw3 all factory setting, Pot 1 going CW until motor makes faint noise, then CCW 3 = 9turns. Pot3 started low, went CW until purple trace was close to the red trace, this is 6.25 turns. In the pic, 6.5 turns has the purple trace higher in relation to to red trace. This is what is required to set pot3, correct? Get purple trace to track red trace as close as possible, while FF1 is 1.0?
When I go higher on pot3 the Y axes motor (is off the machine) wants to turn faster, I use pot4 (y axes) to slow it, but you always over shoot & motor goes opposite direction, so I just shoot for slow. It seems if you change pot3 very much, you need to re-visit pot4 to slow y axes motor each time.

Many thanks

Current loop is tuned with the motor off of the machine. Since we aren't trying to tune the current loop (Because you don't have an o-scope.) We're hoping the default current loop settings are sufficient.

Pretty sure to get any meaningful tuning of the velocity loop done the motor needs to be installed on the machine.

At this point the velocity rise of the motor in the traces you've shown since changing the power supply is much much too slow to be remotely close to useful. But it is possible that if you have the motor off the machine you have to have pot 1 set so low to keep it quiet that the response becomes that slow.

Todd, thanks! If I raise pot1 (the pic is 10.5) any more, when I push arrow key to start x axes moving, the Y axes (motor on the machine now) will start to oscillate (like .5 turn of machine handle) until F2 (joint error out) is up.
If Y axes oscillates very much (I will use pot4 on Y axes) to slow or stop motion, it usually will just get slow oscillations. To get best trace y axes, needs to be stopped when I push the arrow key to trigger trace.

thanks

Todd, I have re-read the top bullet sentence, on page 6 of 9 (this is the page with all the SW2 & SW3 settings. It says
"A larger resistor value will increase the proportional gain, and therefore create a faster response time"
This is what we are after, is it? Like in SW3-1 to 5 go from 20, to 120 k omega?

Many thanks!

Movements of one axis should have no effect on the other. I don't know what's going on there, but it doesn't sound good.

At this point you should have all the SW3 switches all in their default positions.

Since you changed power supplies you you have to start over from the beginning. Trying to retune the velocity loop in the drives that involves Pots 1, 3 and the SW2 switches only.

If the default current loop settings on the drive are insufficient for you to tune the velocity loops. I think you are going to have to track down someone local to you with an o-scope (and the knowlege to use it.) to help you tune the current loops.

You can't tune the current loop with the velocity loop enabled. Just like you can't tune the velocity loop with the position loop enabled. You have to have a constant steady input to tune to and if the outer loops are varying the input you cant tune the loop you're working on. And since you can't really observe the current loop without a scope you're probably best off leaving it with the default settings.

You were making good progress, back about pages 17-18 of the thread, when you ran into the "jerk" caused by I'm not sure what. The AMC tech seemed to think it was a limit caused by an insufficient power supply. (I wasn't fully convinced that was the problem.) You've since changed the power supply. You need to go back to the tuning strategies you used to get the good velocity loop tuning you were seeing back there, and find out if the new power supply actually improved how fast you can command the axis to move without running into the jerk problem you were having then.