Re: INCONSISTENT STEPPING - WIERD SITUATION

Jon,

You have to change direction at some point. If the drive takes
a "snapshot" of the direction input at the active step edge, you
would have a direction timing "setup" requirement then.

Pretend it exists and act if it does. Have the direction true at the
active step edge and hold it for the duration of the step. Doing
otherwise takes unneeded effort in a state machine.

If I may brag a little bit, since we started 2 years ago, we have
gone from zero to passing the 1,000 drives per month shipped
threshold last December. The majority go to ISO 9002 certified OEMs.

They seem to find the "funny" direction setup requirements well
within their capacity to deal with satisfactorily, as do most people
here.

Sorry for being a little snippy. Its been a long day.

Mariss

--- In CAD_CAM_EDM_DRO@y..., Jon Elson <elson@p...> wrote:
>
> Note that the Gecko drives have funny requirements for the direction
> signal not changing for up to 500 uS after a step pulse.
Apparently,
> for Gecko drives to work correctly, the direction signal should
only change
> at the beginning of a step pulse. Most software does not generate
> the signals like this. Most other drives sample the direction
signal
> during the step pulse and then don't mind if it changes later. Many
> software CNC programs let the direction signal return to '0'
> between step pulses, which will drive Geckos crazy in one direction.
> That is why I'm guessing this might be the problem. Se if master5
> has any info on setting up for Gecko drives.
>
>
> Jon