Re: Bullet-Proof G201 / G210

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "jmkasunich
<jmkasunich@y...>" <jmkasunich@y...> wrote:

> Here's another message to look at. Sorry
> it's so late, I was out of town for a week.
>
> Direction input latch. I'm not sure I like the
> multiplexed direction line concept, but I certainly
> do like the idea of getting rid of the long hold
> time requirement on the direction input. Latch
> the direction signal on the active edge of step
> inside the drive, and you will remove one source
> of confusion.

Taken care of; direction latching comes with the integrated pulse
multiplier circuit.

> Internal DC capacitor. Adding an external 470uF cap
> seems like a band-aid, and you never know what the
> ESR will be, or how long the leads will be, etc.
> The drive should be able to operate regardless of
> power supply lead length, especially if you are
> selling it as "bullet-proof". It will probably be
> impossible to keep the same height with a big old
> cap sticking up. I wouldn't mind if the drive was
> 2" tall, what about your OEM customers?

This is a bit of a quandry. A goal is to keep the same form factor
and pinout. Several of our large OEMs (> 1,000 units a year) mount
the drives upside down on a PCB using their own PCB mounted 5mm
sockets. The cap is also on the PCB.

> I disagree with those who want more voltage - it
> seems at least some of those guys are talking about
> servos, not steppers. For steppers, 80V should be
> plenty.

I agree. There is also the issue of liability should someone run a
160VDC rated drive directly from a rectified 115VAC line (sans
xformer) and become metaphysically challanged as a consequence.

> Should be protected against shorts from any output
> to any other output, or to ground. Protection for
> shorts from output to positive supply would be nice,
> but not if it adds cost.

It will use top-side current sensing. An interesting option I have
running senses the top MOSFET "on" Vds using small-signal diodes and
transistors. No wasteful "in-line" components. If the MOSFET "on"
voltage exceeds a set value (1V or so), a fault condition is
presumed. This would also neatly detect over-temperature in the
MOSFET directly because of the positive temperature coefficient for
Rds.

> How are you going to report a fault (overtemp,
> output short, etc.) to the computer? It would be
> nice if this and the DISABLE function were opto-
> isolated like the step and direction inputs, to
> allow complete separation between power and
> control.

Don't know yet. Probably thru a .1" center header out the back and an
onboard "fault" LED.

> John Kasunich

Mariss