Re: Higher performance systems

Jon,

Indulging in a little "calculator fun" shows 746W (1 HP) accelerating
a 396 lb load to 1,000 IPM in 43 milliseconds; acceleration distance
is 0.36".

Identities used:

1) 1 HP = 550 ft-lbs / sec.
2) 1G acceleration (a) = 32 ft / sec^2
3) v (ft / sec) = at
3) s (ft)= at^2 / 2

Mariss

--- In CAD_CAM_EDM_DRO@yahoogroups.com, Jon Elson <elson@p...> wrote:
>
>
> Yesamazza@a... wrote:
>
> >HI
> > I have built a couple of simple low power stepper systems
and this has
> >inspired me to do something higher performance. I like steppers
in that they
> >are brushless high torque and simple but want closed loop high
speed
> >performance. The servo system I was thinking of was the Geckos
for there low cost
> >easy setup. I wonder about there accuracy and still they have
lower voltage than
> >what I would think. I would like to achieve 500-1000 IPM rapids
with a 5 lead
> >screw. A direct board connection may be the way to go to open up
the choice
> >of amplifiers. What have some of you done for higher performance
machines and
> >what kind of results have you achieved?.. I have machinery and
machining,
> >g-code ect back ground some electronics and test equipment with
little computer
> >knowledge. Any suggestions? Thanks Josh
> >
> >
> 1000 IPM? What kind of machine? For a light weight gantry router,
this
> might
> be doable, but quite difficult. For a Bridgeport-sized machine, I
don't
> think
> this is very practical. It will take HUGE force to accelerate the
> machine to
> that velocity before it reaches the end of travel.
>
> Unless the machine is very light, the Gecko drives don't deliver
enough
> power.
> The peak rating is under 2 HP. You may need more than that for
even a light
> router.
>
> Why do you feel you need these sorts of speeds? Just to prove you
can
> do it?
>
> As for the accuracy of Gecko drives, they are as accurate as any
servo
> system,
> if properly set up. If there is interference on the encoder
signals
> causing it to
> lose steps, then you need to solve that. Some users have run into
this sort
> of trouble, but it is solvable.
>
> Jon