Kflop/Kanalog Wiring for Yaskawa SGDH drives

[TomKerekes]

Hi Chris,

It sounds like you are understanding things. Maybe post an updated plot. I see your Velocity setting (9000000) is way too high - but you are not achieving more than ~400,000 because of distance and acceleration limits.

You might use a 2nd order low pass filter to reduce output spikes further than a first order filter will. Possibly 1000Hz Q 1.4.

Good luck.
Regards

[upsyd0wn]

Tom,

When the system powers on the SWE is enabled and my drivers contactor is activated. I have been manually activating JP13-1 to send the servo on command then jogging from the DIO screen figuring that I would program this at a later date. When I select the Enable button in the Step Response screen the SWE disables and my drivers lose power, the only way I've found to reset it is to reset the controller by unplugging the usb and plugging it back in. Any thoughts on why this is happening?

Thanks,
Chris

[TomKerekes]

Hi Chris,

Pushing "Enable" on the Step Response Screen downloads all parameters in the Step Response, Config, and Filters Screens to KFLOP. Make sure all screen parameters are set properly before pushing "Enable". You can set the screen parameters in KMotion manually, by Uploading from KFLOP, by Importing from a C Program, or by loading a *.mot file.

If a configuration is downloaded to KFLOP that reconfigures the IO pins on KFLOP JP7 as step/dir outputs this will kill the communication between Kanalog and KFLOP and the watchdog on Kanalog will disable SWE.

Could that be what you are doing?

HTH
Regards

[upsyd0wn]

I started from scratch repeating these steps:
Since the DAC/Motor/Amplifer are working and the Encoder is working we should be now able to servo.

We have already set the default (low gain low output) servo parameters by loading the KanalogInitialPID.mot configuration.

#3 Open the KMotion.exe Step Response Screen
#4 Push the Zero Button
#5 Hold your hand over EStop in case the axis runs away
#6 Push the Enable Button
The axis should now be servoing. The servo should now hold position with a weak servo. If the motor is displaced it should return to where it was.
Step #4 (move)
On the Step Response Screen select a reasonable Move size (in encoder counts) and push "Move". The Axis should Disable the axis, download the parameters, Enable the axis, move the amount specified, then back to where it started, upload, and plot the results.

The "Command, Position, Output vs Time" Plot mode should be selected at this time. The Command (Blue) plot is the trajectory (or Motion Profile) that was created using the Move Size, Max Velocity, Max Acceleration, and Max Jerk all in units of encoder counts.

The Position (Red) Plot is the measured response.

The Output (Green using Green axis on the right) is the DAC output to the Amplifier in DAC counts.


Step #5 (tuning)

The idea of servo tuning is to get obtain the required accuracy while remaining stable.

This will be covered in more detail later, but the basic process it to:

#1 increase PD gains until the system goes unstable then reduce them
#2 add Filters - Low pass in Filter #2 reduce spikes in the output (ie 500Hz Q1.4)
#3 add Lead/Lag compensator to increase phase margin (this often works better than using D gain)
#3 increase Max values (Output, Integrator) to allow driving the motor amplifier higher to allow higher speeds
#4 increase Vel, Acceleration, and Jerk to find optimal values
#5 add I gain (typically small value of 0.0001) to reduce steady state (persistant errors).

Repeating over and over again... I cant seem to smooth out the output signal, and the position always seems to overdrive then come back... In my attempts to try different filters/PID settings I always start vibrating violently. Attached is what I keep seeing the most frequently. Silently frustrated over here... but keep trying. Wish I could get one axis good so I could move onto the others. My X and Z axis use the same driver/motor, then my Y is a smaller AC servo. (crossing my fingers I might stumble on a magical combination...)

Any pointers from the attached picture?

Thanks,
Chris

[TomKerekes]

Hi Chris,

Nice summary.

From your screen shot I can't see the filters or the Output Scale (right side for green plot). Please make separate screen shots and also save the raw data. That allows us to plot the data in various ways and zoom in (left click drag) to see more clearly.

I don't see a 2nd order low pass filter in Filter #2. Try adding that and also try increasing the D gain. Normally it should be 10X or more of the P gain value to have a damping effect.

Your max error of 200 may be too small also. The error may be clipping at 200 but I can't zoom in to see.

If you set a small Max Following Error (ie 500 counts) then if the system goes unstable it should immediately disable before becoming violent.

HTH
Regards

[bhurts]

Out of curiosity what is your encoder counts per rev and your motors rated rpm. I have not seen them mentioned in the thread but may have missed it.

Ben

Sent from my HTC6435LVW using Tapatalk

[upsyd0wn]

So how do I attache the data file? looks like it is about 1.2mb.

The 2nd order actually helped this time. before it seemed like every time I added a filter it would start vibrating.

How do you go about reducing the noise at the very beginning?




Thanks!
Chris

[TomKerekes]

Hi Chris,

Use the Manage Attachments button to attach files in cnczone.

You actually have a first order low pass filter and a 2nd order Low pass filter. I would remove the first order. Also Q might be 1.4 instead of 1.0 which make a more traditional filter type.

I wouldn't worry about the ringing at the beginning. Pushing the "Move" button disables the Axis, changes any parameters, and re-enables the axis back to where it was. Any drift may cause a jump and the ringing. This wouldn't happen in normal operation. Balancing your amplifier to reduce drift when disabled may help.

Regards

[upsyd0wn]

Ok, see the attached files. What is the max error we are trying to get under? The movement seems pretty smooth... I'll work on determining the encoder counts per inch / rotation in the meantime, tomorrow is my day off so I'm hoping to get my other Axis's moving!

Thx!
Chris

[upsyd0wn]

1 in = ~45050 counts

(This Forum is awesome.. not only for the feedback, but I find myself referring to it just so I know what I did every day! Hopefully it helps others as well, I didn't seem much information about the Yaskawa drivers, at least that was helpful for me)

[upsyd0wn]

Ben,

The X and Z axis motors are:
400W 3000 RPM motors

I'm getting 8192 counts per revolution.

1 in = ~45050 counts

About 5.499 revolutions per inch.

Chris

Out of curiosity what is your encoder counts per rev and your motors rated rpm. I have not seen them mentioned in the thread but may have missed it.

Ben

Sent from my HTC6435LVW using Tapatalk

[bhurts]

At 800000 counts per second for velocity you are commanding over 5500 rpm. 409600 counts should give you 3000 rpm for the velocity.

Ben

[upsyd0wn]

Ahh, good catch, thanks!

[bhurts]

Your 've going to get 545 ipm rapids at that setting. You'll need to decide what you really want to rapid at and that will determine what you set velocity to. I found that I had to really take time to understand the velocity, acceleration, and jerk settings in relation to the encoder counts and motor rpm to get the tuning stable. It will save a lot of headaches down the road. When you command unrealistic values for those settings tuning is a pos and you just scratch your head trying to understand why it's not getting better. It is a great learning experience though.

Ben

[TomKerekes]

Hi Chris,

How well you need to do depends on your requirements. It looks like you are achieving 150 encoder counts at rapid speed. Based on your resolution that would correspond to:

150 / 45050 = 3.3 mils = 85um

Regards

[upsyd0wn]

I think that .003" of accuracy will do! Once I get everything working perhaps I will work on cleaning the signals up more.

The issue I am having now is trying to figure out why channel 1 (JP11-1) is outputting .991v when idle. Channel 0 and 2 are .037v.

Thoughts on that one?

(Got it, just missed the loading of the KanalogInitialPID.mot on that channel.) Thanks!

Chris