Re: [CAD_CAM_EDM_DRO] Gecko ERR/RES Handling

Dan,

I think you are onto something regarding the positive going step pulse. Dad
and I fought the geckos for a couple of days a couple of weeks ago, faulting
all over the place, trying to get acceleration ramping right etc. It was so
frustrating. Dad is growing to have a strong dislike for the Geckos, after
having so many problems in the past with the earlier bug ridden drives, and
another bad experience he had with the Gecko company.

Finally Dad did just what you did and flipped the direction of the step
pulses so that they were positive going and then things seemed to clear up.
We were both so exhausted, I don't think we thought it all through, since
Dad was focused on another problem relating to this in which his control
was letting the step line float with only a 10 k pull up resistor on its
input. Earlier drives he used in his system did not load down the step
input line. He suddenly realized that the 10 k was not enough to pull up
the input to the Gecko and so he reversed his whole scheeme and then
reversed the step pulses so that they were positive going. Things then
started to work better, and we finally got things working.

Is there anything documented on this? Our Gecko experience has been very
frustrating all around, and I don't know if Dad will go back to them again
or not.

I am glad Marris fixed the error/reset line problem.

We were trying to get real fast rapids (not machining) and we did not care
much about the error then. Of course .01 " of error would have been
unacceptable machining.


Dave

----- Original Message -----
From: Dan Mauch <dmauch@...>
To: <CAD_CAM_EDM_DRO@yahoogroups.com>
Sent: Monday, December 17, 2001 3:23 PM
Subject: RE: [CAD_CAM_EDM_DRO] Gecko ERR/RES Handling


> There is another way to look at the 8 bit position error
> question. True there is +/- 128 counts either side. On a system using
> a 250 CPR encoder with a 2/1 ratio timing belt and a 5 TPI lead/ball
> screw you would have 250X4X2X5 or 10000 steps per inch. Each step would
> be .0001" and if the max error was 128 you could have a .0128" max error
> before faulting which for machining would likely be way out of spec.
> Thus a +/- 128 count error would be unacceptable and you would want the
> axis to fault out. Even worse are some servos that would accept 32 bit
> error which would only increase the amount of error with the same
> encoder. The trade off would be that you could run a much finer encoder
> say 1000 CPR which would be 40000 steps per inch and a 256 count error
> would only be .000025 X 256 which would be .0064" which would still be
> unacceptable for machining. But then you would also need a faster pc to
> run the program.
> If your Gecko G320/G340 faults out to easily then there could be one of
> several problems. The first is that the G320 need a signal that is
> positive going. I recently tested Maxnc S&D and found that the Geckos
> would fault very easily. So I loaded Desknc for Dos and set it up with
> the signals also going negative(I may have this backwards but you'll get
> the idea) and the Gecko G320 replicated what I saw with MAXnc. Then the
> only thing I changed was the signal type from negative to positive in
> Desknc. WHAT A DIFFERENCE!!! Whereas with Maxnc step and direction, I
> really had a big acceleration ramp and could not get anywhere near the
> speed that I wanted. With the change of signal I could shorten the
> acceleration curve and get awesome speeds. So check your software. Let
> me know what you are using and I might be able to tell if that is the
> problem.
> The other problem is not setting the acceleration curve properly. I was
> messing with desknc for windows and cncpro on another day and ran into a
> time consuming problem of setting the configuration files so that the
> acceleration was quick yet would not fault out the geckos and still
> maintain top speed.
> Last you may have an inertia problem depending on the machine and you
> may have to increase the ramping of the acceleration curve to over come
> this.
> But I suspect the problem lies with the software and not with the
> Gecko's.
> With regard to the err/res question I have a little PCB that I use on
> some of my Gecko systems that give me a front panel display of the axis
> that fault out. But you don't need them once you set up your machine
> properly. You can follow Mariss's previous email that uses a momentary
> switch and daisy chains the err/res terminals together and uses the
> encoder + and - or operation. I have mine set up for lift to engage (the
> motors), release to run position and lower the switch to disengage the
> motors. If any axis fault out all axis fault out.
> Hope this helps
> Dan
>
> -----Original Message-----
> From: David Eldredge [mailto:techdave@...]
> Sent: Monday, December 17, 2001 10:19 AM
> To: CAD_CAM_EDM_DRO@yahoogroups.com
> Subject: Re: [CAD_CAM_EDM_DRO] Gecko ERR/RES Handling
>
> My 2 cents here:
>
> My Dad and I are not very pleased with the Gecko error/reset handling
> either.
>
> I think there should be a separate output for the fault signal, and one
> for
> the reset line. The fact that you have to build another whole circuit
> board
> to handle these conditions is not very good design, in my opinion.
>
> I also do not like the 128 step limitation on the fault condition. In
> my
> little world of messing with this motion control machinery stuff, ( a
> little world admittedly) the +_ 128 steps is marginal. I know Marris
> says
> it is plenty, but I remain unconvinced. It is a pain in the neck
> getting
> faults all the time while I am developing stuff, like my CNC plasma
> cutter.
>
> I think there is an alternative, and I have ordered some in to try, but
> I
> realize that this list has a lot of loyalty to Marris, (GECKO) and
> recognize
> that he has helped many folks and has taught us all a lot too, so I
> offer
> these 1cents worth of comments in hope that Marris would reconsider for
> others who may feel the same.
>
> Dave
>
>
>
> Addresses:
> FAQ: http://www.ktmarketing.com/faq.html
> FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
>
> Post messages: CAD_CAM_EDM_DRO@yahoogroups.com
> Subscribe: CAD_CAM_EDM_DRO-subscribe@yahoogroups.com
> Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@yahoogroups.com
> List owner: CAD_CAM_EDM_DRO-owner@yahoogroups.com, wanliker@...
> Moderator: jmelson@... timg@... [Moderator]
> URL to this page: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO
> bill,
> List Manager
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>