|
|
Post by paulg100 on Jan 31, 2016 16:25:45 GMT
HI Gadawg
MM not sure if they will cause a similar problem, the brackets ive made to hold the pots have some flex in them which I think will be enough to allow for some missalingment, with that in mind I'm going
to try a non flexible coupler. Ive order one and will see how that works.
I can make something myself if I really have to but trying to buy something off the shelf just to save time and effort.
|
|
|
|
Post by paulg100 on Jan 31, 2016 16:32:23 GMT
Annnnnd... Another big milestone First test with platform just loosely connected and controlled manually to get idea if any major problems. This with 15% motor speed and its looking good  |
|
|
|
Post by vicpopo on Jan 31, 2016 20:11:29 GMT
Looks great ! It seems to be ok for the amc sensor position setup !
|
|
|
|
Post by mhensen on Jan 31, 2016 21:10:56 GMT
Wow.. now you are king!!! Top work and it really looks great!!! Keep it up and start tuning!!! If you have the equipment, can you also measure some info on amperage usage?! This is still my main worry :-)
|
|
|
|
Post by paulg100 on Jan 31, 2016 21:32:43 GMT
thanks guys, really feels like a good achievement today,
you can set the wj200's to show amperage usage when running, once i have things set up a bit better i can run some tests and let you know.
im running delta, not star with bigger motors than you, i think you will be fine.
still some work to do before tuning things much, i have to adjust the length and re-thread the connecting rods once i have their length dialed in and i have to get the frame off and weld the secondary support frame to it, but i can clearly see where to go with things now the electrics is mainly done.
only have weekends to work on things now as work is busy but ill push on as quick as i can..
|
|
|
|
Post by paulg100 on Feb 7, 2016 14:31:07 GMT
Today i had to replace my pot mounting assembly as using the flexible couplers wasnt working (not accurate enough), the pots are very sensitive. This was assembly Mk1, made some nice brackets so its a shame to scrap it   Unlike some of you other people i dont have access to a 3d printer although i might have to change that in the future! so i went with the old analog methods i know and made some M12 threaded couplers from acrylic by duplicating a bolt thread.  The pot slots into the hole:  |
|
|
|
Post by paulg100 on Feb 7, 2016 14:35:56 GMT
Then i injected silicone rubber, its actually dental impression material good way to make use of material i already have.  The pot now pops in and out of the silicone but its enough to hold it firm when inserted. Seems to be working much better so far. So now i can start leveling the torque arms properly:  The white spacers on the lever arms will be replaced with metal later. I had the platform running off a joystick through BFF earlier in the week but it was a bit patchy as the pots were not centering properly. Should be able to give it another go later today and will get a vid.... |
|
|
|
Post by paulg100 on Feb 7, 2016 16:25:45 GMT
damn no luck!
im starting to think there is something else happening thats causing my re centering issues as even with the new couplers the motor is not returning to complete center after moving
and im pretty sure the pots are secure now. This can be seen by the feedback being caused and also the spirit level is off after rotation and return to home.
Is it possible that EMI could be causing a problem with the pots tracking? how would EMI manifest its self in our builds?
4 out of the 6 pots i have i used screw terminals as opposed to soldering so im as sure as can be they are not damaged.
EDIT: or maybe its a software/setting issue?
|
|
|
|
Post by paulg100 on Feb 7, 2016 17:24:39 GMT
|
|
|
|
Post by paulg100 on Feb 7, 2016 20:02:50 GMT
ok i ran autotune with rotation (H001 - 2), set H002 - 2 and A044 to 3 (sensorless vector).
I also increased sensor oversampling to 2 and tried 3 as well on AMC.
Nothing really seamed to help.
options left:
either the pots are still slipping in the couplers or its a noise/EMI problem, anyone have any thoughts? thanks
|
|
|
|
Post by tronicgr on Feb 7, 2016 20:38:08 GMT
damn no luck! im starting to think there is something else happening thats causing my re centering issues as even with the new couplers the motor is not returning to complete center after moving and im pretty sure the pots are secure now. This can be seen by the feedback being caused and also the spirit level is off after rotation and return to home. Is it possible that EMI could be causing a problem with the pots tracking? how would EMI manifest its self in our builds? 4 out of the 6 pots i have i used screw terminals as opposed to soldering so im as sure as can be they are not damaged. EDIT: or maybe its a software/setting issue? If your PID value (P) is low, it won't give the motor enough speed near the target position that could result in "undershoot" like in your case. So I would suggest increasing the P setting on thr amc1280usb until it positions correctly, but without oscillation ("overshoot "). Also its better to have this setting done AFTER the autotune of the WJ200 is done first. Thanks Thanos |
|
|
|
Post by paulg100 on Feb 7, 2016 21:49:05 GMT
Thanks Thanos,
I was running a PID of 18,0,0. I tried raising the P gradually and tested each time until 24 which started causing notable oscillation.
At 18 im not getting any oscillation, its just the motor is not returning to exact home causing sensor feedback.
I tested this after successfully running auto tune.
Its late now but ill try and make a video tomorrow to show what is happening.
Thanks.
|
|
|
|
Post by tronicgr on Feb 7, 2016 22:55:07 GMT
Well, set sensor oversampling to 1 for starters, then in the wj200 parameters, decrease ramp times to 0.5 or 0.1. Then you can increase the P without getting as much oscillation. What is the max RPM of your gearboxes (or what is the motor max RPM and reduction rate of greabox??)? You should have max RPM output of 10-25 RPM on the output of gearboxes, if its faster, reduce the MAX Motor speed from the AMC1280USB LCD menu (will also stop oscillation).
Thanks
Thanos
|
|
|
|
Post by paulg100 on Feb 8, 2016 12:02:06 GMT
OK thanks,
Motors are 1400rpm with 60:1 gearbox so about 24rpm max.
Max speed on AMC has been set at 25% so far for testing so about 6 rpm.
one thing the autotune did do is help with one motor which was stuttering bad in sensorless vector mode, after running autotune its smooth now.
Infact all the motors are running nice and smooth apart from this jittering when they stop or return to home.
Tonight ill reset oversampling to 1, ramp time ive been running at 0.5 but ill try a combination of lower ramp times and higher P to see if i can improve things.
and maybe increase AMC max speed to 45% = 10rpm.
thanks!
|
|
|
|
Post by paulg100 on Feb 8, 2016 21:56:16 GMT
OK didnt have to much time to play tonight but reset oversampling back to 1,
Finished autotuning all motors
Increased motor speed on AMC to 45% which should be about 10rpm on my motors.
Increased P to 20
Decreased Motor forward/reverse speed to 0.05
No luck yet unfortunately, this is still the result:
I tuned the pot before running the test through xsim offline 6dof software and all is quite with no frequency going to VFD.
Once finished (and even during movement) you can hear the feedback and signal is being sent to VFD when it should be 0
have same problem with all 6 motors.
Not sure what else to try at the minute, maybe increasing feedback limit/dead zone on amc, its at 10% at the moment.
|
|
