From: "Bill Bowden" <**email@example.com*>
Subject: Re: MOSFET overvoltage protection
X-Newsreader: Microsoft Outlook Express 5.00.2314.1300
Date: Tue, 08 Oct 2002 00:42:09 GMT
NNTP-Posting-Date: Tue, 08 Oct 2002 00:42:09 GMT
Organization: AT&T Worldnet
Bob Wilson wrote in message
> In article ,
> **firstname.lastname@example.org* says...
> >I have a question about using a MOSFET (IRF640 or 740)
> >to drive an automotive ignition coil (about 6 millihenrys,
> >1.5 ohms, and charged to 50 mJ). Worked for awhile using
> >only a MOV across the drain/source terminals, however the
> >device failed using an RC snubber (0.5uF, 8 ohms) in place
> >of the MOV. The gate to source resistance of the failed
> >device reads about 60 ohms. I'm guessing the device failed
> >from a high negative voltage to the gate as there was no clamp
> >diode in place to protect the gate. However there is an integral
> >diode inside the device from drain to source which should have
> >limited the drain voltage to less than a volt negative, so I'm not
> >sure why it failed. What is the usual practice to protect a
> >MOSFET from overvoltages on the gate and drain?
> >If a clamp diode is used on the gate, should it be a zener diode,
> >or fast acting schottky diode?
> Even MOVs are a lousy way to protect against such spikes. They have an
> extremely soft knee. They are mainly used on AC applications since they
> polarity insensitive. Zeners (or their close relatives, Transzorbs, or
> are the way to go. They are extremely fast, and have a hard, well defined
> clamp voltage. MOVs require the voltage to rise to more than double the
> point where they start conducting, before they are conducting significant
> clamping current.
> The integral FET body diode is useless as a means of protecting against
> positive overvoltage, and is rather slow to turn on with reverse voltage.
> Use a fast diode in parallel for this. You may also be seeing Miller
> that cause the gate to blow, when very fast drain voltage risetimes are
> involved. Hang a 12V zener across the gate.
Ok, I added a 12 volt zener across the gate (NTE5127A) and a 200 volt,
1 watt zener across the drain/source (NTE5105A) but I'm not sure of the
power rating for the drain zener. The peak coil current is 4 amps, and when
MOSFET turns off, the same current will be diverted into the zener with
about 4*200 = 800 watts peak dissipation. The rep rate is 200 per second
and DC input power is about 10 watts but I don't know what the efficiency
Do you have an idea how much power the zener will need to dissipate?