On Friday 13 December 2013 13:50:44 Marius Alksnys did opine:

> 6 axis articulated robot. Its 5th and 6th joint motion is transferred
> from servo motor to the end by belts. Controlling 1-5 joints is trivial,
> but 6th one has to be offset by 5th position.
> 
> This means you have to rotate both 5th and 6th motors equally to move
> only 5th joint.
> 
> ( BTW, this is the reason why I started to think that there could be
> three different objects / levels in linuxcnc: motor - joint - axis )
> 
> There are resolvers on the motors and I use 7i49 card to read them.
> 
> I am using genserkins for world mode and an offset component for
> offsetting 6th motor position and feedback signals in relation from 5th
> motor position signal. And it works correctly, except for homing, but I
> can't figure out how it _normally_ works.
> 
> My homing uses resolver index. The problem is that I get following
> errors even when MIN_FERROR is 60 degrees. And I see that resolver
> position feedback signal makes a step when homing.

Yikes.  I know little to zip about resolvers, only that if its lying to 
you, I would call it a resolver problem and would be all over the maker 
asking for a solution from them until it was proven that it was indeed my 
setup generating the error.  I'd move one of them that is doing this to a 
stepper just laying on the table, and log what it says for at least a turn 
going both ways when the stepper is driven in full step mode, one step at a 
time, where it generally will rest at very close to 1.8 degrees per step.  
Probably closer than .05 degrees per step from ideal.  If there is a repeat 
error cycle every 4 steps, thats the motor itself, so use the average of 
the last 4 steps which should remove that error, but anything else has to 
be a funkity resolver.  IMO of course.  Over 200 steps, a full revolution 
of the motor, it has to be 1.8 degrees per step for most steppers.
 
> I do see that big steps happen in resolver feedback signals, when they
> approach index. But the same is with normal joints.
> 
> Any ideas how to solve this?

See above as a "I have got the tools" way to isolate resolver error to the 
resolver.
 
> I know I could do this in kins, but I would like to be able to use
> standard genserkins.
> 
> 
> 
> Some lines from my config (joint / motor components and signals are
> numbered from 0):
> 
> .hal file:
> -----------------------------------------------------------
> loadrt offset names=offset-j56
> 
> addf offset-j56.update-feedback servo-thread
> addf offset-j56.update-output servo-thread
> 
> net mot.4.pos-cmd joint.4.motor-pos-cmd => pid.4.command
> offset-j56.offset
> 
> net mot.4.index-enable pid.4.index-enable
> net mot.4.index-enable joint.4.index-enable <=>
> [HM2](BRD)resolver.04.index-enable
> 
> net mot.5.pos-cmd offset-j56.out => pid.5.command
> net mot.5.pos-fb => offset-j56.fb-in pid.5.feedback
> [HM2](BRD)resolver.05.position
> net jnt.5.pos-cmd joint.5.motor-pos-cmd => offset-j56.in
> net jnt.5.pos-fb offset-j56.fb-out joint.5.motor-pos-fb
> 
> net mot.5.index-enable pid.5.index-enable
> net mot.5.index-enable joint.5.index-enable <=>
> [HM2](BRD)resolver.05.index-enable
> 

Cheers, Gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>

* Knghtbrd notes he has mashed potatoes for brains tonight
<Valkyrie> yum, can I have some?
<Knghtbrd> um ...
* Knghtbrd hides from Valkyrie
A pen in the hand of this president is far more
dangerous than 200 million guns in the hands of
         law-abiding citizens.

------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT 
organizations don't have a clear picture of how application performance 
affects their revenue. With AppDynamics, you get 100% visibility into your 
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to