Page 1 of 1

OAK Spindle speed issues <<Answered>>

Posted: Sun Dec 10, 2017 9:09 am
by tblough
Getting my new machine configured and am slowly working my way through the startup issues. First one is a spindle speed mismatch at lower rpm. My machine has an encoder on the spindle and the displayed spindle RPM matches exactly with a Shimpo digital tachometer so I'm sure the encoder is reading correctly.

If I issue a M3S500 MDI command, the resulting spindle speed is 475RPM. The offset seems to reduce linearly as speed increases and when I reach 3000rpm, commanded and actual speeds match. This remains the case up to the max speed of 4000rpm.

Any ideas on what could be causing this and where to fix it?

Cheers,

Tom

Re: Spindle speed issues

Posted: Sun Dec 10, 2017 9:46 am
by Gary Campbell
Tom...
In the AC drives I am familiar with there is a parameter that uses a percentage +/- 100% to allow the user to adjust the internal workings of the drive to sync with the output of the control board. These are called by different names, but "Analog input Offset" and "Analog Input Gain" are what I am familiar with. With your proportional issue "Gain" is the one I would use as Gain applies a percentage (higher or lower) to the incoming analog volts across the board. Offset would bump (or decrease) more or less by a fixed amount.

That said, I was under the impression that when an encoder was in place, the analog output was adjusted to desired RPM by increase/decrease of analog volts until matched. Maybe someone with more Acorn experience than I will chime in and clarify this... for both of us.

Re: Spindle speed issues

Posted: Sun Dec 10, 2017 12:30 pm
by tblough
Thanks, Gary. The VFD is a ABB ACS355 which only has upper and lower bound settings for the analog input. There a no linearization coefficients to adjust. There are PID adjustments for the motor control but not for the analog input.

My puzzlement stems from the reason you stated. I though if I commanded 500 rpm, the Oak would adjust the analog output voltage until the encoder reported 500 rpm. I do have a 1.5:1 belt drive in the system with the Oak set to 4000rpm max spindle and 0 min. The VFD is set correspondingly at 0-6000rpm. Both are configured for 0-10V being the full range, and in the upper range, the motor is 1.5X the spindle speed as expected.

Tom

Re: Spindle speed issues

Posted: Sun Dec 10, 2017 1:02 pm
by frijoli
Acorn only sends a 0-10v signal. It does not compensate.
Currently the encoder is only used for displaying RPM, and monitoring spindle position.

I'm not sure why they chose to do it that way.
Clay

Re: Spindle speed issues

Posted: Sun Dec 10, 2017 3:04 pm
by DICKEYBIRD
frijoli wrote: Sun Dec 10, 2017 1:02 pm Acorn only sends a 0-10v signal. It does not compensate.
He's using an Oak Clay. Rich guy. ;)

Re: Spindle speed issues

Posted: Sun Dec 10, 2017 3:51 pm
by frijoli
DICKEYBIRD wrote: Sun Dec 10, 2017 3:04 pm
frijoli wrote: Sun Dec 10, 2017 1:02 pm Acorn only sends a 0-10v signal. It does not compensate.
He's using an Oak Clay. Rich guy. ;)
I'm blaming this on Gary. He said Acorn. lol

Re: Spindle speed issues

Posted: Sun Dec 10, 2017 6:35 pm
by cncsnw
Standard PLC programs (Oak, AllIn1DC, etc.) do not attempt to chase spindle speed, even when encoder feedback is available.

Since it is usually possible to dial in speeds within 1% or 2% with open-loop control, adding compensation to the PLC program would add a lot complexity for little gain.
Any ideas on what could be causing this and where to fix it?
It could be that the analog trim on the Oak board is a little out.
It could be that the analog response of the VFD is non-linear.
It could be that the analog input of the VFD puts excessive load on the PLC's analog output, thus pulling the voltage down at low levels.
If you want to isolate it further, you could try this:

Verify that spindle control is in auto mode and the override is 100%
At the MDI prompt, enter "M3 S400"
Note the actual RPM
Measure the DC voltage on the analog terminals
Enter "M3 S800"
Repeat the RPM and voltage measurements
Continue up through M3 S3600

Now you can see whether the analog voltage varies linearly with commanded RPM,
and also whether the actual RPM is directly proportional to the analog voltage.

If the analog voltage does not vary linearly with commanded RPM, but is instead distorted down when the speed command is low, then try temporarily disconnecting the analog wires at the VFD and repeating the series of voltage measurements.

Re: OAK Spindle speed issues

Posted: Sun Dec 10, 2017 9:26 pm
by tblough
You are the MAN, CNCSNW! Voltages were right on at each step running at most 0.01V high. I went ahead and recorded the actual RPM at each step and ran a regression on the points and they were incredibly linear. The drive has a couple of parameters that I though might correct for the skewed slope.

First I tried adjusting the minimum allowable speed to 31rpm while leaving the maximum at 6000 thinking that that might skew the analog input slope but it had no effect on the slope just the minimum rpm that could be commanded.

Then I looked again at the analog input parameters. There is a minimum AI and maximum AI setting that seem to be 0 - 100%. On a whim I tried setting the minimum below zero and it let me set it to a negative value. A setting of -0.7% completely eliminated the skew and now commanded RPM values correspond within +-1rpm of measured.

Thanks again CNCSNW for pointing me in the right direction.

Tom