Re: servo vs stepper

Steve, I have been searching for an oscilloscope. I had one in a
research lab that I used to own that would have been perfect. Maybe
I can borrow it. Can't seem to find anywhere to rent one yet...

It does take the x and y axes simultaneously at 45 degrees with the
same rates but makes some noise.

Thanks,
Keith

--- In CAD_CAM_EDM_DRO@yahoogroups.com, Stephen Wille Padnos
<spadnos@s...> wrote:

> Right - OK.
>
> Do you have an oscilloscope? Access to one?
>
> You should take a look at the pulse train coming from your

controller

> board. The Geckos have some specific requirements regarding pulse

shape

> (4 uS high time, 0.5uS low time, I believe). Look at the step and
> direction inputs on the Geckodrive, and get your board to send a
> moderately high speed pulse train to the Gecko (do back and forth

rapids

> over the full travel, on one axis)
>
> There are several things to check with a scope:
> 1) Step pulse fall time. The falltime should be fast - like 0.1

uS or

> so. The Gecko clocks a step on the falling edge of the pulse, so

this

> should be as clean as possible.
>
> 2) Low level of step pulse: This should go pretty much to ground.

It

> may not, depending on the output driver on your board. Anything

below

> about 1.5 volts should mean that the optoisolator LED is being

turned

> on. The driver on your board should be able to sink 16mA (I

think).

> Any less may take longer to get the signal through the

optoisolator (or

> not work at all).
>
> 3) High level of step pulse: This should go all the way to 5V (or
> whatever the step/direction supply is). Incidentally, the control
> supply on the Geckos should be 5V. You can run it at a higher

voltage

> if you put a resistor inline, but it's not a good idea. They

aren't

> compatible with 3.3V logic (there is a current thread on the

Geckodrive

> list regarding this).
>
> 4) Step pulse width: Make sure it's 4uS on and 0.5uS off. Even

longer

> is better (you probably don't want to get steps going at 200kHz,

so

> there's no reason to push the minimum step timings).
>
> 5) Make sure the direction line isn't changing
>
> One of the boards available (Pico systems or Campbell - I don't

remember

> which) has a mode where it outputs direction changes synchronously

with

> the step pulse (on the rising edge of the pulse, I think - so

there's no

> uncertainty about which direction to step when the step line goes

low)

>
> It's possible that the output from your board isn't perfectly

compatible

> with the Geckos.
>
> Out of curiosity, if you do 45 degree diagonal moves, can you go

faster

> than 20 IPM? I'm not sure what the answer would tell us, but it

would

> be interesting to know.
>
> Keith Clark wrote:
>
> > <>Yes, I have the Geckos set for 7 amps which is the rating for

my

> > current motors' current. I have a 75 volt 16 amp powersupply that
> > is unregulated. I think the steps per inch is correct since I get
> > accurate distance travel. I have the rapid setting set at just
> > under 20 ipm since it stalls with higer settings and the acel and
> > decel parameters do not seem to help but when not correct causes
> > occasional stalling even at lower speeds.
> >
> > Steve, thanks for the consideration
> > Keith Clark
>
> Always up for an interesting question :)
>
> - Steve