Re: INCONSISTENT STEPPING - WIERD SITUATION

Evening All,

GOOD NEWS !!! Mariss suggested that i change from using a seperate
power supply to the game port power out and it corrected the lost
stepping and intermittent stalling. when i put a voltmeter from
direction to ground i had voltages of .756 and 3.38 during
directional changes. if these voltages are supposed to be 0 and 5
volts then something is still in error.by the way i have constructed
the current dump circuitry and have also implemented this.when this
motor is running it runs very smoothly at feedrates of 10 i.p.m. or
greater in any direction , however at feedrates below this the motor
still growls and sounds unhappy but hasn't lost any steps YET.is
there any other procedures that i can do to eliminate this? i have
tried the trim but theres not enough trim to be had !!!! haha

i'd also like to thank each of you for your help and insights

jeff

--- In CAD_CAM_EDM_DRO@y..., "mariss92705" <mariss92705@y...> wrote:
> Art,
>
> Not to worry. The multipliers were originally designed for point to
> point operation; machine stops, change direction, go again.
>
> People started using them for constant contouring.
>
> That required a revision so that each sent step is "tagged" with
its
> direction. When the step executes, it goes in the direction it was
> tagged with. That was done on REV-4, 3 months ago. No more problems.
>
> The next revision will have it tagged 4 steps deep. The only way to
> spoof it then will be to change direction 4 times while
accelerating.
> Not a realistic situation.
>
> Though it was never designed to do constant contouring, with REV-4
it
> has been able to, based on reports and extensive testing here.
> Because I am paranoid about this, REV-5 will quadruple-nail this
> situation shut.
>
> You are taking care of this in software. The REV-4 (current
version)
> takes care of it in hardware as well.
>
> Mariss
>
> --- In CAD_CAM_EDM_DRO@y..., "Art Fenerty" <fenerty@h...> wrote:
> > Surprisingly enough I have had this complaint before. It is
usually
> the
> > step/dir pins being swapped or the polarity is wrong. Gecko's do
> indeed need
> > special handling of the DIR bit ( Actually, it must be held on as
> long as
> > the multiplied pulse train) but Master5 takes this into account.
> (Couldn't
> > have it NOT run Marisses drives..that would be kiss of death).
> >
> > Try swapping step/dir or polarity and see what happens.
> >
> > Good luck,
> > Art
> > Master5 Software
> > http://users.andara.com/~fenerty/master.htm