Re: Fet controller question? and others

["Paul A. Cianciolo" <paulc@snet.net>]

Hello Richard,

Thanks for the reply.   I enjoyed looking at your site!!

Thanks also for the ideas on driving the 2 controllers.  I have never driven
such large arrays of fets before.

I was interested in the info on the freewheeling diodes,  I will need to
read more
 about that.
I gues I will scratch the SCR idea nd look for some good diodes

Thank   you again

Paulc
W1VLF
Cloudbounce Webpage  http://www.qsl.net/w1vlf/

1986 Vanagon Gas
1982 Vanagon Diesel  Turbo Diesel 1.9
GE Electrak E20 and E15  electric tractors

----- Original Message -----
From: Richard Bebbington <richard@rbebb.demon.co.uk>
To: <ev@listproc.sjsu.edu>
Sent: Wednesday, September 29, 1999 6:20 AM
Subject: Re: Fet controller question? and others


> In article <005901bf0a24$b2b574c0$89273ccc@paulc>, Paul A. Cianciolo
> <paulc@snet.net> writes
> >It consists of 2 banks of IXYS part# IXFH 50N20 ..  I checked thier site
and
> >I think they are 50 amps at 200VDC'
> >
> >The fets were used as choppers for a ferro transformer.  The minus supply
> >went to the common of the 2 banks of FETS and then each bank has a wire
that
> >then went to 1 side of a center tapped ferro transformer with the center
tap
> >at 36 VDC.  I think it was a sine wave output but may have been stepped
> >squarewave.
>
> > I have 2 of these units...  I wonder if I could parallel them?  As I
look
> >at them know I don't think there is enough capacity for one of my E20
> >electric tractors controller.
>
> Hi there!
>
> I also have 2 banks of transistors, that came out of a UPS that
> we had here in work. Your description of the circuit they were
> hooked up to sounds identical to how mine were connected.
> I've used mine in my electric barstool project
> (  look at http://www.rbebb.demon.co.uk/barstool2.html  )
>
> The trannie sets I've got are good for up to 160 A max, at 250V.
> They seem quite happy switching 70 amps into each motor
> ( the motors are not so happy with this! ;)
>
> You should be able to parallel them OK, just check that
> they're the same part number, and be careful about how you
> drive their gates. With all those parallel FETs, you will
> have a lot of gate capacitance to drive.
>
> I'd try using a dedicated high-speed MOSFET driver IC for each
> group, and feed the driver ICs the same signal.
>
> If your drive signal can't charge the gate capacitance quickly
> enough, the FETs will take longer to switch ON or OFF
> and will develop MUCH more heat, possibly even damaging them.
>
> >
> >Freewheeling diodes?   I am not sure what to use but I think that was the
> >cause of the demise of my homebrew controller.
> >
>
> >From the description of the circuit you took them from, I'd
> guess there's a freewheel diode built in somewhere in the array.
> Mine had 2 diodes, one 70amp freewheel that is across the load,
> and another 50 amp that is across the transistors.
>
> It's the diodes across the motor that'll be important.
> Make sure they can carry 2x or 3x the FET's current rating.
>
> This is where my barstool is running right on the edge -
> the freewheel diode's ratings are not big enough, eventually
> it'll smoke! ( But hey, do I care?  nahhh..... ;)
>
> >Can I use SCR's?  for the diodes....   I came across a bunch devices that
> >were used as surge suppression
>
> Hmmm.... I don't know this one....
> I believe you could, but you'd need to trigger the SCR just as you're
> turning your FETs off. If you trigger too late, the motor's
> inductance will have already built up a large voltage and killed
> your FETs.
> If you trigger too soon, the SCR turns on while your FETs are on,
> and you short the pack.... BOOM !
>
> I think I'd use diodes.
>
> Anyone know more on this?
>
> >I am also looking for a websites that talk about motor controllers and
what
> >things like freewheeling diodes do
> >beside just asorb spikes
>
> The freewheel diodes aren't there to absorb spikes, they're there to
> allow the motor current to continue to circulate smoothly.
> ( that's why they're called freewheel diodes, the current
>   freewheels round in a loop from motor to diode, and diode to motor,
>   until it slowly dies away )
>
> If you interrupt the current ( i.e. if you have no diodes )
> suddenly, the motor's inductance will generate a huge voltage.
> This voltage spike will rise up until something breaks down and
> conducts, usually your nice expensive FETs ! This is how you let the
> smoke out!
>
> Hope this helps,
>
> --
> Richard Bebbington
>