Re: Gecko burned out (not about tinned leads)

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Steve Stallings"
<stevesng@n...> wrote:

>
> A loose wire can kill a driver that does not
> have protection from the high voltages produced
> by interrupting the current through an inductor.
> If you break the circuit while current is
> flowing through the stepper motor coil, it is
> possible to produce several thousand volts
> of inductive "kick". If it is not clamped,
> it is enough to eat the MOSFETS alive. The
> Gecko products are engineered to operate well
> when used properly, but they do not have the
> extra circuitry to protect against this type
> of abuse/fault.
>
> Regards,
> Steve Stallings
>

So when you pulse the motor with square waves you don't get this
inductive kick? That's the purpose of the recirculating diodes in
the opposite FET? I don't think the DV/DT of a loose connection
will produce thousands of volts of inductive kickback (maybe
hundreds and that would possibly overvoltage the FET's). Most
designs driving inductive loads have snubber networks that dampen
the fast rising FET killing spikes.

Driving large inductive loads (like big motors) can get to be a
challenge. You spend half your time designing the circuit and drive
and the other half figuring out how to keep it from destroying
itself.

Things happen pretty fast when you punch a load with a squarewave
with a rise time of 40 ns, then turn it off while current is still
flowing as in a PWM drive. While pulse by pulse limiting is nice it
gets complex to do. With about twice the components (of a Gecko)and
the resulting price, you could design a drive that will withstand a
direct short and just smile....Oh wait, they have those and their
about 350 to 600 bucks a shot!<g>

> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "caudlet" <thom@t...>

wrote:

> <snip>
> >
> > So, I'm curious. How does a loose wire (not shorting to

something)

> > burn up a Gecko? Too much kick back voltage from the inductive
> > load? This sounds like another "legend" that has been told so

many

> > times it becomes fact.
> >
> > I would vote for simple component failure (it happens) provided

the

> > winding is not shorted.