Tom,

I do not currently have the step polarity set low, which is seems it should be for Geckos. I am using channels 8-11, and the KFlop outputs go into a breakout board (same one that's been in use for the past 3 years), which drives the Geckos. I'll give that a try. But I am losing position ONLY when jogging with the pendant, NEVER when running G-code, so seems unlikely to me that will make a difference.

I have gotten MoveExp() working with no apparent position loss, by allowing the MoveExp() to run until it gets within 0.005" of the target position, then doing clearing pcoeff and doing a final Move(). But I have another mode that uses only Jog(), with the velocity determined by the MPG spin rate, and that mode is still losing position, and I can hear the servo faulting occassionally.

Certainly part of my problem is a noisy, low-res MPG. I NEED to get a better one, but this is not the time. But I'm doing enough filtering and masking that what I'm passing to the Move commands is pretty smooth.

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Ray,
>  
> It should not be a problem issuing Jogs or Moves before the previous has completed.  They should blend together.
>  
> Actually even if the motion stops suddenly or jerks there should not be a loss of position so I don't really understand that.  I think you have Step/Dir driven servos so there should only be two possible ways the position can get off.  Loss of Step pulses going to the drive or loss of encoder counts.
>  
> A common problem is to not have enough setup time on the direction signal.  That can cause a step to be interpreted in the wrong direction.  The step pulse should be inverted if necessary so that the active edge of the pulse is at the end of the pulse rather than the beginning.  See:
>  
> http://www.dynomotion.com/Help/StepAndDirection/StepAndDir.htm
>  
> Regards
> TK
>  
>  
>
>
> ________________________________
> From: himykabibble <jagboy@...>
> To: DynoMotion@yahoogroups.com
> Sent: Wednesday, February 15, 2012 11:21 AM
> Subject: [DynoMotion] Jogging Etiquette
>
>
>  
> Tom,
>
> Sadly, I'm still fighting jogging problems - I'm getting small, random position losses. Is it perhaps not kosher to issue a new Jog() when a previous one is already in progress? How about doing a Jog() before a previous Move() has completed? Individual moves seem to work OK, but it looks to me like it's overlapping commands that's causing the problem.
>
> Regards,
> Ray L.
>