Re: Axis Stall Problems

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Paul Kelly" <tech@c...> wrote:

>
> The 16F628 has comparators. User a timer to turn on a
> transistor every 1/20kHz seconds and a comparator
> interrupt to turn it off. You have a programmable
> internal voltage reference to use too.. Note though
> that you need to ignore the switching transient. It was
> very clear in the posted scope traces that a large current
> spike occurs when you switch, if your comparator trips
> off this, then you aint gonna get that motor to turn...
>
> Best check how fast those comparators are first though..
> Just checked. 400nS. That should be fast enough.. Then
> there's the interrupt code. Say 10 instructions @Fosc=20MHz..
> that's only a max interrupt rate of 400KHz assuming the
> micro is doing nothing else... Not an unfair assumption,
> as soon as you start moving at any speed and get busy
> handling step interrupts you'll stop PWMing.. That gives
> you a best case of about 20 current levels at 20kHz chop
> frequency. Not bad but not great either...

I actually tried the 'F628 for revision C of my board.
Here's the URL:

<http://gramlich.net/projects/cnc/motion/rev_c/index.html>

I used off board serial DAC's, so being able to change a
clip voltage was not particularly fast. I thought my new
design was more clever, but I was wrong. ;-)

> Alternativley Futurlec carry L297's for US$3.80 Qty 1

L297's are looking really attractive to me right now.
No, microstepping, but they should work just fine.
I realize that L297/298 boards have been done before,
so I am largely reinventing the wheel by going down
that path.

[snip]

-Wayne