Kollmorgen AKD drives

[jfavre]

Hi I am trying to connect to a Kollmorgen AKD-B00306-NBAN-0000 With a AKM52H-ACCN2-01 Servo motor via Flying Lead 13133. I have enable and encoder signals working but cannot get the motor to move via OAK controller. Not sure if i should use precision or velocity mode. Does not seem to be any voltage on pins 25 and 26 while trying to jog. Below is a link to the manual.

https://www.kollmorgen.com/sites/defaul ... v%20AG.pdf

[centroid467]

Hello, please post pictures and a diagram of your wiring with the flying lead cable. It seems that these drives can be configured to work in either mode whether you should use precision (position) or velocity mode depends on the wiring and the settings in the servo drive.

Did you wire the Oak to the analog inputs on X8, the digital inputs on X9 or X7, or to both? X9 seems like it would be best used as encoder output to the Oak.

What mode is the drive set to in the Kollmorgen software?

[jfavre]

Hi,

I did use X9 to be used as encoder output to the Oak. I did wire the Oak to the analog inputs on X8, pins 3 and 4 for enable and 9 analog in- to pin25 logic ground on Oak flying lead and pin 10 analog in+ to pin 26 AN REQ on Oak flying lead. I have tried both precision (position) and velocity mode with the oak and drive set for the same mode, the drive is now in velocity mode. I think that i am not getting output on Oak pins 25 and 26. How can i verify this? I will have to try to make some sort of drawing to post. Right now I'm bench testing setup, not much of a picture. I did go through Kollmorgen software and verify operation of motor and drive via Kollmorgen Work bench software. Thanks for the reply.

[centroid467]

Alright, thank you for breaking down your wiring.

Precision mode is position mode using A/B phase quadrature step signals. If you were using that mode, you would need to wire it to X9 which then eliminates the encoder feedback to the Oak and would not work. So, you wired it right and must use velocity mode.

It is probably easiest to measure the Oak analog output on the drive's X8 header. Black lead of multimeter in 20V DC range on pin 9, red lead on pin 10.

If you open up CNC12 and press ALT + I on the keyboard, are inputs 1-6 and 8-11 green? If not, navigate the box cursor to each of the inputs mentioned with the arrow keys and press CTRL + ALT + I to invert them. This is a temporary measure for testing so please do not forget if they have been inverted. You may have already done this as it is part of the bench test procedure.

Also on that screen, if output 1 is not green then you will not get analog output. You may have to temporarily force the output by pressing CTRL + ALT + F on OUT1.

[jfavre]

Oak pins 1 through 7 are wired to X9, encoder feed back looks to be good as I have it set for 1rev=1inch.
Got 3.267 vdc on pins 9 and 10
Inputs 1-6 and 8-11 are forced green per test procedure.
Output1 is green, not forced.
Should i be using Oak pins 25 and 26 for analog signal? I did not see any voltage on these pins.
Will be posting sketch soon.

[cncsnw]

Got 3.267 vdc on pins 9 and 10
...
Should i be using Oak pins 25 and 26 for analog signal? I did not see any voltage on these pins.

Didn't you say you had Oak pins 25 and 26 wired to drive pins 9 and 10? If you have 3.267V on drive pins 9 and 10, where is it coming from, if not from the Oak?

You can tell, pretty reliably, whether Oak/CNC12 has enabled the axis and what velocity command it is trying to send, by watching the PID Configuration screen. The "PID Out" column on that screen will show "OFF" if the axis is not enabled; or a number between -100 and +100 (representing the percent velocity command) if the axis is enabled.

What do you see on the PID Configuration screen when you select Continuous, Fast (rabbit) jog mode and hold down an X axis jog button?

[jfavre]

I misunderstood and was measuring Oak pins 9+10 vs Drive pins 9+10. There is no voltage on Oak pins 25 + 26. The SUM on PID screen ramps up to 2560000 and stops then I press jog and Oak faults out. Drive wont reenable until i cycle power to Oak. Oak will replace existing Galil ICM/AMP 1900 controller.

[jfavre]

cabinet picture

[centroid467]

Do you have anything set for the PID values now? Follow TB234 to get velocity mode tuned up right.

What is the encoder counts per rev? If it really is the 2048 from the report in the first post then you should try starting with Kp = 3.125, Ki = 0.03906, Kd = 0, Limit = 2560000, Kg = 0, Kv1 = 80, and Ka = 0.

If not, follow TB234 to get a good starting point.

[cncsnw]

There is no voltage on Oak pins 25 + 26. The SUM on PID screen ramps up to 2560000 and stops then I press jog and Oak faults out. Drive wont reenable until i cycle power to Oak.

There is no need to cycle power. Just press and then release your emergency stop button to clear the stall/fault condition.

If your Ki value is zero, then the Sum will not contribute any PID output, no matter how large it gets.

You need to watch the "PID Out" column to see what voltage you might expect on pins 25 & 26. If PID Out = 0, then the correct analog output voltage on those pins is zero.

Kp = 3.125 seems a little aggressive for an unknown velocity-mode drive. I usually start with Kp = 0.01 or 0.02, and work up from there. Also, start with Ki = 0, or at most 0.0005.

It is rare to end up with Kp much over 0.12 in velocity mode, with reasonable encoder counts. If you really have just 2048 counts/rev (equivalent of a 512-line encoder) then your gains probably will be higher; but you are likely going to have controllability problems with such low resolution.