Re: Axis Stall Problems

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Irby Jones" <irbyrat@c...> wrote:

>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Wayne C. Gramlich"
> <Yahoo@G...> wrote:
> > The chopper is implemented using a dedicated Microchip
> > PIC16F676 microcontroller (one for each axis) running
> > at 20MHz where a couple of A/D lines are used to measure
> > the current flowing through each L298 H-Bridge. Each
> > axis is being sampled at approximately 20KHz. If the
> > current is too high, I turn the H-Bridge off and if
> > the current is too low, I turn it on.
>
>
> Wayne,
>
> I like what you're doing with this concept, and have been following
> this thread. I have been thinking of a similar concept, where I want
> to be able to microstep at 100 microsteps per regular step. I have a
> bunch of steppers that have a 500 line encoder attached - 2000 lines
> with quadrature. It would seem possible to have a control loop that
> microsteps the motor until an edge is reached on the encoder inputs,
> closing the loop around the encoder (at 2000 steps per rev.) similar
> to a servo. So I'd like to help you with some info you may not have.
> Right now I have a system using the L297/L298 combo that works fine.
> I don't know if you have seen a actual scope trace of what the
> current-limiting of this pair looks like or not, but here is one:
>
> http://members.cox.net/tararatski/L298_current_sense.jpg

Thanks for the two traces. My scope is kind of dead, and
the one I am borrowing is dying as well.

> The top trace is the voltage across the .5 ohm sense resistor. The
> bottom is the voltage at the sense input of the L298 (pin 14 in this
> case). The L298 pin 14 signal is filtered by a 10K series resistor
> from the sense resistor and a 680pf cap to ground. The scope is set
> at 10 microseconds per division and .5 volts per division. The L298
> oscillator is about 18KHz (in fact with the 56 microsecond period,
> it's 17,857 Hz). The stepper is a Powermax II M21NRXA, 8 wire,
> running the coils in series. The specs for this configuration are
> phase resistance .92 ohm, inductance 2.8 millihenry. I took these
> with my digital camera, so it's hard to see the slope of the top
> trace exactly, but there is a little rise after the ringing stops.
> The reference voltage into the L298 is .240 volts, and it trips at
> just about that. This circuit has a "holding current" reference
> voltage that kicks in after a small delay without a step pulse, and
> the traces show the "holding current" configuration (with about .48
> amps current into the still motor).
>
> Hope these traces help you figure out what to do next.

My gut feeling is that my strategy would have worked if
I had used a DSP (Digital Signal Processor) with 2 channels
of A/D that sample at 200KHz+. Alas, the PIC16F series
does not fit the bill. There are plenty of outfits pushing
DSP's out there, but at this point in time I do not want
to expend the effort figuring the intracacies of a DSP.
I'm leaning towards a plain vanilla L297/L298 pair.
Yes, this has been done before.

With regards to using a 2000 edge encoder to control
your stepper, that sounds quite interesting. I think
it can be done with the right DSP or FPGA. Note that
when a your motor is cranking at 20 rev per second
(= 60 inch/min rapid with a 20 thread per inch leadscrew)
the edges are coming at 40,000 transitions per second.
It is going to be very desirable to have a hardware
quadrature counter at that speed. Again, this says
DSP/FPGA to me.

Best wishes,

-Wayne