Wiring and configuration SERVO DRIVES

Moderators: TomKerekes, dynomotion

User avatar
TomKerekes
Posts: 2527
Joined: Mon Dec 04, 2017 1:49 am

Re: Wiring and configuration SERVO DRIVES

Post by TomKerekes » Wed Nov 25, 2020 12:28 am

Hi Gabriel,

I think if INB and SGB were confused/swapped it would cause what you see. That is the DAC is driving SGB and you have INB connected to Kanalog GND.
Regards,

Tom Kerekes
Dynomotion, Inc.

GabrielR922
Posts: 79
Joined: Thu Aug 13, 2020 10:07 pm

Re: Wiring and configuration SERVO DRIVES

Post by GabrielR922 » Wed Nov 25, 2020 12:52 am

Hi Tom.
Surely I swapped them.

But it's a little bit strange for me, it has white and red cables, for me the red is +. (this is why i just tested 0V and DAC GND and not DAC OUTPUT)
It's reversed in this situation.

Can you can help me again with tuning?

Tomorrow I will make more tests with different settings.

Before wiring the 0V encoder, I changed the trajectory planer settings and it got a little better.
Tomorrow will test different values and put all info here. It continues overcorrecting and then going bumping (only with X and Y interpolation).

Maybe the errors were because the non 0V connection ?

Also, I have some questions about C program, should I create another topic in the correct board?

Thanks for your awesome support Tom.

And sorry if it looked I was going against you.

GabrielR922
Posts: 79
Joined: Thu Aug 13, 2020 10:07 pm

Re: Wiring and configuration SERVO DRIVES

Post by GabrielR922 » Thu Nov 26, 2020 1:59 pm

Hi Tom. How are you doing ?

Some evolution but not as planned.

I will post all info for you analise if you can.

Please if you need anymore info, tell me.

While moving a single axis it works nice.

While interpolating X and Y in some places it make corners wrong.
While engraving it's a disaster.

Tried many and many and many configurations. (more D gain, more I gain, more P gain) also less gain in all configs.
I'm losing hope :(


Thanks Tom.
Attachments
engrave.jpg
corner.jpg
YaxisSTEP.PNG
YaxisBLOD.PNG
Yaxis.rar
(493.22 KiB) Downloaded 46 times
XaxisSTEP.PNG
XaxisBLOD.PNG
Xaxis.rar
(479.67 KiB) Downloaded 47 times
TracjectoryPlaner.PNG
INIT.c
(4.32 KiB) Downloaded 45 times

User avatar
TomKerekes
Posts: 2527
Joined: Mon Dec 04, 2017 1:49 am

Re: Wiring and configuration SERVO DRIVES

Post by TomKerekes » Thu Nov 26, 2020 5:20 pm

Hi Gabriel,

One issue I see is your tuning is mostly relying on Velocity Feed Forward to create the 1200 DAC counts of output. The Integrator gain is low and the limit of 100 DAC counts prevents it from doing much. Feed forward is like driving blind folded. If the road is perfect and you anticipate corners perfectly then you can stay on the road. But if conditions change, or there are external disturbances to push you off the road, you won't do very well (cutting forces).

On the other hand, the Max Following error is set at 100 encoder counts. This means the servo errors never exceed 100 encoder counts. Which is:

100/15880 = 0.0063 inches = 0.16mm

Which is about the thickness of a piece of paper. Your errors seem larger than this. So that would indicate something mechanical such as lead screw backlash or other mechanical compliance or resonance.

The Step Response Plots only show errors of ~15 encoder counts. You might reduce the Max Following error to find the lowest value where you can still run GCode and cut wood without faulting.

The example CaptureXYZPosDestToFile.c can be used to capture data while running GCode to see exactly what trajectory is being commanded and how well the servo is performing. It can be assigned to an MCode (execute only without waiting) to start capturing data just before an area of interest. Then you can analyze the captured data with something like Excel.

But I suspect you have mechanical problems. Do you have a dial gauge? If so, measure the table relative to the Spindle and make small moves while changing direction to check for backlash or compliance. Also push on the Spindle and measure any deflection.

Please post a GCode fragment that shows the problem.

This is a large machine?
Regards,

Tom Kerekes
Dynomotion, Inc.

GabrielR922
Posts: 79
Joined: Thu Aug 13, 2020 10:07 pm

Re: Wiring and configuration SERVO DRIVES

Post by GabrielR922 » Tue Dec 01, 2020 7:14 pm

Hi Tom.

Sorry for the delay.

Yeah, it's a large machine, a SCM RECORD 2

I found a ballscrew backslash in the X axis.
It seems to have a "preload", not always the backslash occour, only with fast changing direction movement. (if changed direction with slow speed, no backslash)

I configured the Y axis properly and got around 30m/min with max 50counts error.
The machine is rated to 45m/min.
30m/min is really fast and a blink can cause a crash.

At the moment i'm using 12m/min in both X and Y because the backslash.
For the engraving problem I'm using a really slow configuration, so I can use the machine. (lot of work)

In january I will be able to take off the X axis ballscrew and send to repair.

Tom, your support is excellent, always helping and solving problems.
At first I was going to buy CSMIO boards, but thank god i choose Dynomotion.
Not just by the support, but the freedom we have with the board and C programs.

I have some questions with C programming, do I post here or in the correct board?

Also, this machine has a 10 position rotary tool changer, I have the logic done, just need some more knowledge in programming. After making and refining this program, I will post it to wiki. If you want, obviously.

Thanks Tom!

Also, are you coming to Brasil soon? Interested in bringing a combo, KFLOP + KANALOG + 2 KONNECT to sell ?

Below a photo with the tips that you posted above about I and D. (It works great, no plot because machine is off, soon will post plot)
Attachments
eixoYultrafast30m.PNG

User avatar
TomKerekes
Posts: 2527
Joined: Mon Dec 04, 2017 1:49 am

Re: Wiring and configuration SERVO DRIVES

Post by TomKerekes » Tue Dec 01, 2020 7:31 pm

Oi Gabriel,

Obrigado pelas palavras agradáveis!

C questions would probably be best in the C Program forum.

And yes, posting to the wiki would be appreciated.
Regards,

Tom Kerekes
Dynomotion, Inc.

GabrielR922
Posts: 79
Joined: Thu Aug 13, 2020 10:07 pm

Re: Wiring and configuration SERVO DRIVES

Post by GabrielR922 » Wed Mar 10, 2021 6:24 pm

Hi Tom!

after your recommendation to check my mechanical system, did a full overhaul.

now the mechanical part is 100%.

I tuned the Y servo, but if i use more than D10 it gives some unstability at a speed of ~5m/min
no unstability in the plot screen.
attached are the info.

Thanks Tom.
Attachments
plots.rar
(474.71 KiB) Downloaded 43 times
Capturar2.PNG
Capturar.PNG

User avatar
TomKerekes
Posts: 2527
Joined: Mon Dec 04, 2017 1:49 am

Re: Wiring and configuration SERVO DRIVES

Post by TomKerekes » Wed Mar 10, 2021 7:43 pm

Hi Gabriel,

Its not clear what you are asking.

You didn't include your Step Response Screen settings.

Have you tried tuning? What have you tried?
Regards,

Tom Kerekes
Dynomotion, Inc.

GabrielR922
Posts: 79
Joined: Thu Aug 13, 2020 10:07 pm

Re: Wiring and configuration SERVO DRIVES

Post by GabrielR922 » Thu Mar 11, 2021 6:24 pm

Hi tom.

Yeah, i tried a lot of different tunings, following your guide.
today i'm in a bit of a hurry, as soon as i get more time i will post all configs for you.

Upon further tests, my conclusion was that:

If I'm moving the machine in X or Y the machine works great.
If i make an ARC or move it with 2 axes, they start to become unstable and jerky.


Thanks
Attachments
jerky.PNG

Post Reply