Re: Chopper Transient Suppression

.2uS is way to short. The motor winding capacitance results in
ringing (multiple cycles, underdamped envelope) for several
microseconds. We use a 5uS blanking period (comparator ignore time)
at the start of each switching cycle.

Don't depend on the I-sense RC low-pass filter; blank it digitally.

Mariss


--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Wayne C. Gramlich"
<Yahoo@G...> wrote:

>
> All:
>
> My Axis4 4 axis chopper board is making excellent progress
> but I think I still have a problem with turn on transients.
>
> Here's the Axis4 URL for some context:
>
> <http://gramlich.net/projects/cnc/axis4/rev_a/index.html>
>
> In particular, the schematic of interest is:
>
>
>

<http://gramlich.net/projects/cnc/axis4/rev_a/schematics.html#X_Y_Z_A_
Axes>

>
> Basically, what have have is a microcontroller that
> strobes the L298 on at 25kHz. When the voltage across
> the current sense resistor exceeds the set value,
> the comparator turns the H-bridge off.
>
> I deliberately designed the circuit so that pulse that
> turns on the H-bridge takes precedence over the comparator
> that turns it off. This is because there is supposed to
> be transient spike when the H-bridge is turned on that
> will cause the voltage to temporarily exceed comparator
> set value.
>
> Currently, I have the microcontroller holding the H-bridge
> on for the minimal amount of .2uS. The behavior I get is
> that my stepper motors vibrate even when they are not being
> commanded to move (not good!) What I think is happening is
> that when I am close to the coil correct current, the turn-on
> spike causes me to turn off prematurely. The current decays
> further, until it is so far below what the turn-on transient
> is not large enough to cause problems. Then H-bridge is left
> on until the current gets back to the desired current level.
> In short, I suspect I am getting an asymetric pulse stream in
> the steady state, rather than a steady stream of symetric
> pulses all of the same width.
>
> I suspect that the transient spike that is occuring on my
> system is a bit longer than .2uS, but I lack the tools to
> actually measure the transient. Does anybody know that
> the all-in-one products like the Allegro use for transient
> suppression times? I've dug around in the spec. sheets but
> I have not quite stumbled on a specific number. Is it 1uS?
> 2uS? or more? Obviously, I can experiment and just keep
> increasing the number until things stablize, but I thought
> I would ask here first, since this group has been so helpful
> to me in the past.
>
> Any help on this issue would be greatly appreciated.
>
> -Wayne
>