Re: Would Snubbers Kill Ignition Coil Output? (fwd)

[Steven Roys <sroys@xxxxxxxx>]

---------- Forwarded message ----------
Date: Tue, 16 Dec 2003 21:48:59 -0500
From: Bruce Bowling <bbowling@xxxxxxxxxxxxx>
To: High Voltage list <hvlist@xxxxxxxxxx>
Subject: Re: Would Snubbers Kill Ignition Coil Output? (fwd)

When switching an inductive load, the flyback voltage will take out the FET
unless you provide a active clamp or snubber circuit.

There are many ways to do this. One easy method would be to use a ignition
IGBT. Many manufacturers make these - for instance, here is one from on-semi:

http://www.onsemi.com/pub/Collateral/NGD8201N-D.PDF

Looking at the first page of the datasheet, notice the back-to-back zeners
between the collector and gate. This is the active flyback clamp - when the
IGBT turns off the inductive load the flyback voltage increases to the
point where the clamp zener between collector and gate reaches avalanche.
This in turn activates the gate, turning on the IGBT. So, in essence, the
IGBT acts as an active flyback clamp. Note that the zeners between the gate
and emitter are there to prevent the gate voltage to exceed a certain value
(~ 15 volts or so). Be sure to have enough current source/sink for the gate.

I have been designing automotive EFI systems for DIY use, and flyback from
turning off the fuel injectors has been one of the areas of research. For
instance, look at:

http://www.bgsoflex.com/megasquirt.html

Look at the FET driver page schematic, and you will see an external PNP
transistor configured as a power zener diode. In addition, since the
controller uses PWM for injector current limit when operating (i.e. peak
and hold mode), each PWM turn-off cycle causes a new flyback spike, which
really jacks up the power dissipation for the power zener. The solution was
to use a recirculation scheme which recirculates the flyback when in PWM
current limit, but on final injector close the power zener takes. For fuel
injector applications, you want the injector to close as quickly as
possible, otherwise tuning at low injector pulsewidths is more difficult.
So, the circuit, coupled with the processor (and firmware) offers the best
of both worlds.

- Bruce




>---------- Forwarded message ----------
>Date: Sun, 14 Dec 2003 17:10:58 +1030
>From: Matthew Smith <matt@xxxxxxxxxx>
>To: High Voltage list <hvlist@xxxxxxxxxx>
>Subject: Would Snubbers Kill Ignition Coil Output?
>
>Hi All
>
>Thinking of putting electronic ignition on my aged Massey Ferguson tractor -
>discretely of course, don't want to spoil the look of the thing.
>
>Just fiddling around with the components that I've got, I have some IRF748V
>HEXFETS, which look ideal for the job of switching, having a TINY R{DS(on)}.
>
>Only problem is that they are only rated up to 60V.  This is fine for the sort
>of voltages that I would expect to find around the tractor, but I'm rather
>concerned about the magnitude of the back-EMF from the ignition coil.
>
>My thought would be to put in a snubber with RC in parallel in series with a
>diode (I've got some nice hefty TO220-style SMPS diodes).
>
>However, would putting the snubber in kill the output of the coil and not
>leave
>me enough to fire the plug?  My documentation is all SMPS stuff and doesn't
>cover this type of core arrangement, so I'm not sure what's going on in there.
>Does it act like a flyback with the mother and father of all airgaps?
>
>The idea was to have the points pulse a 555 in monostable mode, putting out a
>900ns pulse, driving the MOSFET via a totem-pole arrangement to speed up the
>switching.  (The idea being to cut down losses and stop the thing getting too
>darned hot.)
>
>Cheers
>
>M
>
>
>--
>Matthew Smith
>Kadina Business Consultancy
>South Australia
>http://www.kbc.net.au