AMC drive tuning

Interesting. The XY encoders are single ended linear scales. Pretty course at 10 micron scale. The z is a rotary 1000ppr encoder in quadrature and sounds rough too.
Could a previously miss configured setup have been masked by the slower old sd1525 drives?

I was hoping for this to have been a simple drive swap and be up and running by now.

This is why is so important to test with open loop (FF1 only)
First step should be to bisect the issue (drive interface or feedback)

I'm attaching the hal and ini.

Made very little progress today. I feel like I'm going backwards.
I set the pid to 0, FF1=1 and the output scale to 3.33. (10v equals to theoretical 200ipm. 3.33 for output scale is this correct?)
I assume its wrong as jogging at 1ipm took approx 16.8 sec to travel 1 inch. Reading through the ini docs, I'm not fully grasping the settings.
Oddly, adjusting the settings, at one point, the machine moved almost twice as fast in one direction than the other. This is at 1 ipm.
Drift is very little. Which is good. At fast table movement things seem relatively smooth. Acceleration may be too high, stop seems rough.
The handwheel is fighting me when I push but when I pull it continues moving in the pull direction which seem wrong. The encoder on the z feels more locked. joint.0.f-error climbs in the same direction/slope when pushing/pulling on handwheel. Should't it reverse?

Is the asymmetrical handwheel response with only FF1?
Does it behave this way with the inputs shorted or unconnected?

Is the 7I77 AOUT and GND wired to then the drives REF+ and REF- pins?

That's correct. With FF1=1 and pid all to 0 it's not symmetrical.
Unconnected its behaving correctly. Locked and fighting back as expected. A shown in my prior video clip beginning.
Yes, the 7I77 AOUT and GND are wired to drive REF+ and REF- pins

Strange. I wonder if there is enough switching noise that the common mode input range of the drive is being exceeded.

How is the drive input common connected?

Pin 4 ref in + and pin 5 ref in- on p1 signal connector to AOUT and GND from 7i77
Chassis ground on drives to buss bar. Don't have any other ground connections other than drains from shielding connected from tach shared with AOUT harness at the drives. Previous photo attached shows cabinet layout.

There is a negative tach input thing that I didn't understand. The drive manual suggest use signal ground for positive input. With motor leads connected as outlined in the maual I got a runaway and simply reversed the tach input wires. That was a suggested fault correction and confirmed by AMC tech support as should be okay. Now I'm wondering if I need to reverse the tach wiring and reversed the logic via dip switch? Was my approach wrong? My guess would be it doesn't matter as long as it doesn't run away??

My guess is that you need to connect signal ground (2) to the 7I77 GND

PCW wrote:

Strange. I wonder if there is enough switching noise that the common mode input range of the drive is being exceeded.

How is the drive input common connected?

Oh, yes, I've seen this on other systems. If the drive's analog circuitry is fully floating, it can exceed the common mode range of the input amp. So, you need to tie the CNC control's ground to the ground reference of the servo drive's
analog electronics. Your symptoms sound a LOT like this kind of problem.

Jon

Okay, I will try that tomorrow. Pin 2 or 11 on that connector P2 is already used with the tach.
Same GND from 7i77 (ex tb5 pin3) that goes to ref-? Or do you recommend another ground?