Re: S.s. MOSFET-driving

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "by way of Terry Fritz <twftesla-at-qwest-dot-net>" <Kchdlh-at-aol-dot-com>

Thanks for the tips, Dave.  I concur completely.

Ken Herrick


In a message dated 12/31/2001 5:29:30 PM Pacific Standard Time,
tesla-at-pupman-dot-com writes:


>
> Subj:Re: S.s. MOSFET-driving 
> Date:12/31/2001 5:29:30 PM Pacific Standard Time
> From:<mailto:tesla-at-pupman-dot-com>tesla-at-pupman-dot-com
> To:<mailto:tesla-at-pupman-dot-com>tesla-at-pupman-dot-com
> Sent from the Internet 
>
>
>
> Original poster: "David Sharpe by way of Terry Fritz <twftesla-at-qwest-dot-net>"
> <sccr4us-at-erols-dot-com>
>
> Ken, Terry, All
>
> Check International Rectifier website (www.irf-dot-com) on FET
> driving info. On Semiconductor (ex. Motorola Semiconductor
> Group-www.onsemi-dot-com) and IXYS (www.ixys-dot-com) also
> have app. information on FET and IGBT driving techniques.
>
> I first started working with 1st generation power FET's in
> early 1980's, and I have built power switches up to 500A
> (1400A pulse at 30V) using IRL3803 30V 140A FET's.
> Keys for reliable driving without oscillations is you MUST
> put a slow down resistor in series with the gate from the
> driving circuit, and no larger then a 5K pulldown to the
> source.   The gate speed control resistor according to
> On-Semi / IRF should be a minimum of 5 ohms, and I
> know at work we have used as large as 100 ohms
> successfully.  Resistor size controls dV/dT speed of
> switching during turn on and turn off.
>
> Additionally, across gate to source, place a bidirectional
> zener clamp (as Digikey part no. P6KE16-CACCCT-ND,
> 16V).  Max Vgs on Power FET's is generally +/-20V, if you
> overvolt a Vgs gate, you can have reduced gain, or FET will
> turn on in active region (may fail shorted).  Another trick,
> especially with multiple FET's in parallel is to pass gate
> lead through a small ferrite bead to prevent high e-field
> impulses from stimulating FET into self oscillation.  I have
> not seen that problem on circuits I have built (either DC
> switching, or max power frequency switching of 400 Hz).
> In the later condition, typical 90% on time was 20usecs,
> and off time was 5usecs, with 100 ohm series resistor and
> maximum power frequency of ~125Hz.  Another important
> point in high power paralleled switches is to make the circuit
> physically symmetrical.  Asymmetric circuit layout may lead
> to circuit impedance parasitics that will stimulate very
> ugly mulitple megahertz self oscillations between adjacent
> devices (a squeal, some smoke, a snap, and lots of
> pergatory comments...  :^C   )
>
> Also be aware that if you try to make the FET switch
> "FAST", the demons of circuit Ldi/dt will rear their heads.
> I have seen overshoots as much as 30V on a circuit that
> was predominately resistance (no intentional inductance)
> that was powered by 5V, and the total PCB power circuit
> loop length was considerably less then 6".
>
> Regards
> Dave Sharpe, TCBOR
> Chesterfield, VA. USA
>
> Tesla list wrote:
>
> > Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>
> >
> > Hi Ken,
> >
> > Richie has a bunch of tips for protecting TC FETs too at:
> >
> > http://www.richieburnett.co.uk/mosfail.html
> >
> > Cheers,
> >
> >         Terry
> >
> > At 12:43 PM 12/31/2001 -0500, you wrote:
> >
> > >
> > >>
> > >> Hi Ken,
> > >>
> > >> You may want to see Marco's discussion of IGBT drive problems and
> fixes at:
> > >>
> > >> http://personal.inet.fi/atk/dncmrc/
> > >>
> > >> I found that I needed series resistance (100 ohms) to slow down the IC's
> > >> switching speed or they would blow the IGBTs.  The IC designers wanted
> to
> > >> force
> > >> the gate capacitance to the voltage they want instantly and they did a
> bit
> > >> "too" good of a job ;-))  That is the only trouble I ran into but I
> don't
> > >> have
> > >> transformers and such (fiber optic).
> > >>
> > >> Cheers,
> > >>
> > >>         Terry
> > >>
> > >>
> >
>
> ==========SNIPOLA
>
> >
> > > Terry-
> > >
> > > Thanks for the tip; I'll check it out.  It's likely that the 100 ohm
> resistor
> > > damps out destructive oscillation, which will occur when the low-Z and
> very
> > > fast drive signal meets up with the MOSFET's gate capacitance.
> > >
> > > Ken
>
>