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

Todd, Thank you!! I am reviewing the SW2 settings in the AB15A100 manual on page 6 of 9. I have only tested 4 options out of total of 60, so far, is that correct? I should be thinking about a way to test the 1st, the middle options and then the far end of option and review traces, then lock in on the area that look most promising & then go 1 by 1 setting at a time to see how that goes.

Am I reading the chart correctly? You reference the SW2-2,3 are on; this is column .737 under the SW2, it shows (1 & 4=off) in the (velocity loop integrator cap options), correct?

Should I pick my group of settings to try first by picking column .147 (use the setting on=1; & 2,3,4 = off) the column 1.837 has been tested.

I need to copy the page and use it as check off list, I will have usb erased and use file name that describes the test.

I am wondering if I have a handle on how to approach the testing.
Many thanks

The chart moves from left to right increasing capacitance. The higher the capacitance the slower the responce (less overshoot). You want to work your way across the chart and find the lowest capacitance setting that has almost no overshoot. Right now you're at 1,2 so skip ahead a few columns and try 1,2,3. If that has no overshoot move back to the left, if it still has significant overshoot move a couple columns to the right.

You don't have to test every setting, the technique is called bracketing the response. You move up in larger steps, until you go to far, then move back and forth in smaller and smaller steps to find the ideal setting.

PS: I think you may want to fine tune pot1 so the motor is quiet at rest (no humming.) If that still gives acceptable response.

Todd, thank you! I have turn down Pot1 from 11 11/16 to 11 5/8, was able to get feed to only 25.5 (from 20), if I use next feed rate 27"/min it will jerk motor, see trace. The file name starts with the mf 1.517 (from the chart heading) vs. switch settings.

Many thanks!!

Interesting, you may want to show that last trace to the AMC engineer and ask his opinion. The drive eventually does get the motor up to speed. I wonder if it is hitting some kind of current limit? Would it be possible for you to check your power-supply and the motor's voltage while the "jerk" is occurring?

I'm guessing you don't have any good way to look at the amps being applied to the motor then either. If you have access to a real o-scope, you might be able to monitor the current output of the drive using pin 8 on the drive's plug instead of buying an expensive current probe.

If it is a current limit issue it's possible that lowering your acceleration rate might fix it.

But I think would suggest trying to tune the PID loop in Linuxcnc first limiting speeds to 20ipm while doing it.

Todd, I have tweaked Pot1 (from 11.687 to 11.562; pot3 is still at 3.312 (I tried tweaked this a well, but good where it was) and feed is 30.8"/min. This is not consistence it seems, if I start late in the trigger window it seems to have better chance of working. Attached pic has this trace, both are same settings.

Many thanks

I'm not sure I understand what you mean by "This is not consistence it seems, if I start late in the trigger window it seems to have better chance of working."

If you look at the top of the Halscope just above the plot screen and below the Pos slider there is a horizontal black bar that represents the sampled data, the vertical line through it is your trigger point, and the rectangular box is the area displayed on the plot screen. Moving the horizontal Pos slider moves where the window displays. Moving the trigger Pos slider (the vertical one) changes how much data is sampled before and after your trigger event. I see now that you must be making at least two jog moves and are actually viewing the last move. Move the horizontal Pos slider at the top of the screen so that the box overlaps the vertical line through the sample bar, to actually see your trigger event on the Halscope screen.

Todd, Many thanks! I moved the trigger bar so its inside the box. This I will try with all in the future, seems better. When I talk about consistence I have the same settings on both traces, one looks good, not so much for other trace. I can appreciate if I am on edge of the tolerance that 2 out of 7 traces do not show the same. Currently, I am to work trying to raise the purple(white) trace closer to the green trace while increasing the feed rate. The 2 traces attached are the same settings, but traces do not look the same.

many thanks

Todd, Thanks you so very much for your help. The feed rate I can increase a small amount to 32.5"/min, but the distance between the traces is growing. The units are all 100m/unit. The small amount I was able to tune is 1/16 of a turn. Should I continue?
Many thanks!

Edit: To stop the drift in the Y axes, I plan to set Y drive SW1 & SW2 &pot 1 & 2 settings, use Pot4 to remove drift and save to file.

The drift will stop once you start adding P when tuning in Linuxcnc. I'm curious what the AMC tech will say about the image of jerky move.

What was the method they gave you for determining the setting for pot2 the current limit? I am curious if turning that pot up a little will help with the jerk problem.

(I think I calculate the pot2 setting for a 6.25amp continuous current limit to be approximately 11-1/3 turns from the low stop or 1-2/3 turns from the hi stop.)