CL Step servo issues

[bsjoelund]

Hi Tom,

As there is severe problems with the Yahoo group I will 'continue' my questions over here assuming all messages comes thru.

Hi Bengt,

What kind of motors/drives do you have? How are they configured?

Regards
TK

Did answer but my message just vanished. So here we go....

AC Servos discussed late March with analog problems and 'we' decided to go for CL Step, output via rj45 connector on Kflop.



0/20 for X
0/21 for Y
0/22 for Z

What amazes me is that Z with the highest possible load works very well.



But both X and Y are not 'acting' so where do I go from here?




Axis show that both X and Y are creeping with no commanded move, what is the cure for that?




Hoping for some advice to get things back on track.

Cheers
Bengt

Similar Threads:

• Need Help!- Step file issues
• Need Help!- Spindle Interference on Kanalog DAC ground causing Kflop Step issues?
• High speed issues, 2 Geckos, 1 car battery, 1 step up converter
• Servo step multiplier
• Units Per Step For Servo

[TomKerekes]

Hi Bengt,

Before jumping ahead to using closed loop Step/Direction you should verify that open loop Step/Dir is working first. It's less confusing if you go step-by-step and verify each step is working correctly before proceeding to the next step. The basic sequence is:

#1 - verify the Step/Dir signals to the Drive are working open loop. Commanding a motion to the drive should move the correct distance. Returning back should repeat to the original position. There should be no drift in physical position after repeated moves.

#2 - verify the Encoder feedback is working correctly. Commanding a motion the Encoder position (red) should follow the Command (blue) with only small errors. There should be no drift in measured Encoder position after repeated moves. Observe the amount of following error.

#3 - enable closed loop control.


It would be simpler to work with one axis at a time.

Instead of photographing the screens take a screen shot using alt-print screen Then paste into a program like paint.exe and save to a *.png file.

Regards

[bsjoelund]

Hi Tom,

#1 So I change to No Input and Step Dir and output 4

Starting with X axis only.

Cheers
Bengt

[bsjoelund]

Hi Tom,

Now I did have a memory stick with me

Both X and Y axis move as commanded, several repetitions of size 100000 = 25mm and no loss of position.





So next up is #2

can you kindly be detailed in what steps to do please.

1. Set driver back to Quadrature input
2. Change Config to Encoder and CL Step and change output to 20 (X and 21 on Y)

As Z is already working in quadrature with settings in earlier picture, but not X/Y with similar settings.

What next?

Cheers
Bengt

[TomKerekes]

Hi Bengt,

Why did you change the driver to Quadrature input? Maybe you are thinking you need to use Quadrature input to get quadrature encoder output? The input and output should have nothing to do with one another. The whole idea of step #1 is to verify you can command the drive to move reliably using either Step/Dir or Quadrature or any other mode. After it is verified to move correctly do not change anything regarding how the drive is driven.

Also to verify the Encoder feedback is working you do not need to change to CL Step. You only need to configure Encoder Input Mode and observe the Encoder. You can observe the Enoder on the KMotion.exe Axis Screen. Or the red plot in the Step Response Screen. You plots show no encoder motion. So either the encoder is not moving, or it isn't wired correctly, or it isn't configured properly.

HTH
Regards

[bsjoelund]

Hi Tom,
Well I do what ask me to do, I verified that S/D is working with no Enc feedback. As regards Quadrature Output, you wanted me to go for this during testing in March as it would be much better as regards pulse width or something in those lines.

So now I will set Encoder Input Mode and see what happens. ( it very strange that Z axis is Enc input and Quad output and it seems to work well. Any explanation to this Tom? )
Have double checked the encoder wiring and all 3 axis are connected as they should be.

So next up will be screendumps with Encoder input.

Cheers
Bengt

[TomKerekes]

Hi Bengt,

Sorry for the confusion. I just meant to verify the Drive will "Step" properly in both directions in what ever mode you intend to use. If we decided that Quadrature would be the best mode then that should be the mode you use and test.

I don't have an explanation why your axes behave differently. Maybe X and Y are not configured properly.

Regards

[bsjoelund]

Hi Tom,
Been down in the workshop doing tests with S/D + Encoder and Y is OK but X is not OK. Some days ago I had a mishap with Z and brake that resulted in Z going down fast and unfortunately my tool cart was in its way and one sharp corner managed to damage in going cables to the motor.
Power cables OK but obviously encoder cables was damaged. Now I have to find how many of the 15 is broken.

Y axis


X axis


Will report my findings with the cable as soon as I get it repaired,

Cheers
Bengt

[bsjoelund]

Hi Tom,

Re X axis, the encoder did get severe damage. I was in the hopes that the cable was damaged but it is OK.
So I have now ordered new encoder from manufacturer in China.

I will continue to get all in place with what I have.

Question regarding the Z axis brake. Is there some C code that can monitor BIT 156 so it can not be switched OFF before brake is activated.
BIT 156 power enable to drivers, SON is onboard that releases the brake as it is now, 24VDC externally to control relay. So here is the problem
if BIT 156 is turned off the brake is still non active. So I need a watchdog of some sort to be safe when drivers are powered of.
Have had several 'run-aways' with knee plunging down uncontrolled. One of these 'trips' did damage the encoder.

Cheers
Bengt

[TomKerekes]

Hi Bengt,

I don't fully understand your description. How is the Brake controlled? By an output?

If your code is turning off Bit 156 to disable the drivers can't it also simply activate the Brake?

Regards