Re: RTLinux vs RTAI and EMC

[Ray Henry <rehenry-at-up.net>]

This is why I advise folk to do a bit of up front design.  It is also the  
reason I recommend the Gecko 340 rather than 320.  Not a lot of machines even 
work to a quarter thou.  Make that a 2x pulse multiplier or a 1 to 1 ratio 
and a machine is running about as fast as your Bridgett.

Ray


On Monday 10 June 2002 00:32, you wrote:
> Dale,
>
> If you take a 500 line encoder (not a particularly high resolution) that
> gives you 2000 steps per motor rev. Now is you have 2:1 belt reduction
> (very common. But on the low side for most servos motor for best
> performance) that puts you at 4000 steps per screw turn. Now if you are
> using 10 tpi acme screws (again a common setup) you now have 40000 steps
> per inch of travel. So if you want only 60 ipm you need 40 kHz.
> Certainly not a very impresive rapid rate, but sure will push freqmod.
>
> Tim
> [Denver, CO]
>
> > -----Original Message-----
> > From: emc-at-nist.gov [emc-at-nist.gov] On Behalf Of Dam
> > Sent: Sunday, June 09, 2002 10:44 PM
> > To: Multiple recipients of list
> > Subject: Re: RTLinux vs RTAI and EMC
> >
> >
> >
> > Tim,
> >
> > I've done some testing of freqmod on an AMD K6-2  450 Mhz.
> > With period
> > set at 0.000016 the pulses on my scope sure look like 16us
> > high and 16us
> > low, that figures to be 31.25k Hz. How fast do you need?  I'm still
> > trying to figure out the ini parameters for using my own step &
> > direction servo drive.
> >
> > Dale
> >
> > Tim Goldstein wrote:
> > >Ray,
> > >
> > >Unless the speed of freqmod can be drastically increased then
> > >steppermod is needed for step and direction servos (Gecko, Mr Bill,
> > >Rutex).
> > >
> > >Tim
> > >[Denver CO]
> > >
> > >>Do you really need steppermod.  Freqmod will work nearly as
> > >>well but not
> > >>quite as fast.