Centroid Community CNC Support Forum

I am having trouble clearing up the Y axis lagging behind on other axis's. I am getting as much as 20,000 encoder error on the paired Y axis.
I have tried adjusting Parameter 341 and 344 (axis 2 and axis 5, Pair axis) with minimal change.
I have tuned each drive in Sigmawin 7 separately by disconnecting opposite drive cable but leaving drive coupled. I averaged the "moments of inertia" of 389 and 408 then set both at 60% of that set now at around 239%.
Following Error is fairly identical for both paired drives.
I assume I need to make a change in the drive parameters but not sure what any help would be appreciated.
X and Y are running Yasakawa .85 kw servo motors thru planetary drive to rack.
X axis has minimal error of around 125 at 400 in/minute while paired Y is 16,000 at 400in/min
I have servo parameters and drag test for all drives but can't get them to load on forum.
I tried PID setting for velocity mode TB 252 with no significant change.

i wouldn't use velocity mode.

do you have the position mode delay set properly?

how are you measuring the following error?

for odd file types. you can zip them up and post the .zip file here or better yet host the files on your google drive and post link here (they don't have to be uploaded to this site.)

https://www.centroidcnc.com/dealersuppo ... ds/267.pdf

I like I stated have tried adjusting Position mode delay from 0 to max of 60 with minimal change.

The error shows on PID screen error and tops out at 20,000 counts, I can also see it in the cuts that its lagging a huge amount.
I tried Position mode settings PID settings which didn't bring error count down much. nor did velocity mode PID settings.
I made some adjustments today in drives setting. in the custom settings I bump the "feed forward" up from 14 to 220 and "feed Back" from 81 to 120.
error on PID screen has drop from 16,000 to around 10 when moving at 300"/min.
But now I have a bad whine/ grind sound that I can't get rid of.

I added additional files including servo parameters files.

I would reset the drives to factory the default and configure according to the TB 267

At first glance, Keith's suggestion here looks like what JWM used to call a "Labrador retriever assignment". Go do something time-consuming, which might or might not have any useful purpose, but which will keep you out of my hair for a couple days.

Comparing, "y1a.usrs" with a similar file from a different system, only one thing stands out as unusual: Pn002 appears to be set to use an absolute encoder, even though Oak does not support absolute encoders. However, your encoder feedback seems to be working, so maybe that does not really make any difference.

There are a dozen or so other parameters with non-default values, all of which could be expected from running the advanced autotuning with reference input. None of those values stand out as being particularly out in left field.

It sounds to me like you are on the right track. I would probably back off on the custom settings -- especially the "feed back" gain -- to see if the whining/grinding noise can be reduced to a tolerable level, without letting the following error grow to more than maybe 200-300.

i have helped many customers with drive parameters issues for many years and have burnt many hours trying to help users dig themselves out of huge drive setup holes that they have dug for themselves and learned that with ANY drive setup it is absolutely the best approach to reset to factory parameters and follow the centroid setup guidance for that drive. which typically takes under five minutes to do. just seems like a good approach to eliminate any unknowns and get to a known/understood baseline in which to start to work on things.

and to clarify OAK does support Absolute encoders with an optic direct board for use with yaskawa.
https://www.centroidcnc.com/dealersuppo ... ds/301.pdf

Fair enough.

Carl: did you happen to notice whether your tuning, response, and following error issues with the slaved pair were present between when you got all the axes working in early December, and when the Y (or was it V) pinion slipped in mid-January?

It does seem a little odd that the following error issue seems to have cropped up only after the pinion-slip issue.

It is also very odd that you did not see a change in following error as you made changes to parameters 341 and 344. Depending on the change, errors might get worse instead of better; and depending on a lot of other conditions the amount of error may never be very good; but it seems to me like those parameters should always make a difference.

So, based on the assumption that things were working better in December and early January, maybe it is worth going back to square one, at least with your Y and V servo drives.

In the mean time, Centroid can work on making the delay-time autotune work with paired axes.

The drives in question are not on Opticdirect board. I will reset to incremental encoder not sure how that got changed.

I believe following error was there before encoder slipped. There was some change with parameters 341 and 344 but the error is so large it didn't clear it up significantly.

I noticed motor speed was defaulted to 10000 so set to 3000 to match motors
I tried starting a new tune with defaults, (not factory reset) first run I got an error the pn522 "Completed move width" was too small. It was set to 7 I had never change it, I set it to 2000.
Ran tune mode 2 again and it never finished after 45 minutes.
Restarted and it gave change mode and pn522 to larger value. set pn522 to 5000.
tune keeps failing.
Think I will start by resetting drive.

I am using Sigmawin+ 7 to match drives. It doesn't follow along TB 267 very well parameter labels seem to have changed and also choices. Does anyone have the correct settings for these drives that I can just enter in?

I have power disconnected from slave drive, but it is still coupled. I plan to just copy parameters over after tuning primary drive.