some good advice (pid)

Peter,

<No, something is wrong if FF1 is not independent of P>


FF1 change a bit, and I got this result.


thanks

Yes, there may be a small interaction
(because the drives are not perfect velocity mode drives)
but this was a huge change like using a wildly wrong FF1 value,
perhaps the hal file was not updated or somehow out of sync

It also looks like the motor drive velocity loop is not tuned too well
(too low velocity feedback gain?),
you can see at the beginning error spike the slope of the PID out
being ahead of the velocity ramp to compensate for the drive's
lagging but the drive does not respond for maybe 20 mS or so

this is best achieved (ignoring the peaks)

How do I remove the peaks?
looks good?
notice something wrong with my PID?

thanks,

It looks like you should be able to have more P gain before instability (the oscillation) sets in
You can try slowly adding some D to see if you can make it stable enough to add more P
but the (fairly large) spikes may also be because your drive(s) are out of adjustment
(looks like too low velocity gain) or perhaps some other drive setting.

Do you have the drive manuals available?

I have the manual but printed.

I've been testing looks but failed to reduce the peaks, on the contrary increase.

think I have to adjust my drives? but that is more difficult.

check my manual to scan and upload the most important thing.

but the truth somewhat disappointed at not being able to achieve anything. but i keep trying.

increase my max_acceleration. and boot speed was faster.

Sure, this it expected, if you increase the acceleration, the peak errors will be larger

Have you tried slowly adding D as I suggested? this should reduce the oscillation

friend

You really need to adjust the driver?, Really gives me a little scared to go to damage the drivers, is that the procedure is simple. but remember my limitation of English (English correctly interpreted). and my lack of knowledge on the subject.

I try tomorrow with D. to see if the error is reduced.

I appreciate all your help, God bless.

Peter,

adjust the drivers in <tacho>. and zero pot.

Watch the new plot. I think it is much better in acceleration, were the peaks.
start and end.

the tacho was maladjusted.

good friend hope now if can be successfully completed.

Actually it looks like you went a little too far with the adjustment

At the beginning of the velocity ramp where position once lagged. now it leads

Whatever you adjusted on the drive, you should adjust for minimum spike at the beginning of motion

Also there seems to be a dropout of encoder counts during the slew. This needs to be looked into. I would check the encoder wiring and encoder power
(and make sure that the encoder shield wires are only grounded at one end of the encoder cable)

This may also be a symptom of a dirty encoder