[Diy_efi] Flyback Circuit: was new EFI controller

[David Hooke]

Marcel,

I beg to differ.

In a perfect world, Vbatt and Ground to the controller have low source 
impedance, so dumping large currents into either of them will be fine. I 
reality, we generally have much lower impedance on the ground line than 
on the Vbatt, because we've designed and wired our systems that way.

There's generally a low impedance path from the injector + to the 
battery (maybe via a relay).
There's a low impedance path from the injectors to the switching device 
and clamp. Ditto from ECU ground to battery. We know this 'coz we used 
big wires, lots of connector pins and a ground plane to ensure it.

There's a much higher impedance from the ECU supply to the battery: 
smaller wires, a fuse, not many connector pins, board traces instead of 
a ground plane. So the large flyback currents will take the path of 
least resistance which is not back to the injector+ via the battery 
wire, but through the power supply decoupling to ground (and hence back 
to the injector+), and those caps will not handle the currents involved: 
two things may happen, the caps'll blow up as you exceed their peak 
current rating, and the voltage at this point will rise considerably, 
(V=I*Zesr), which will stress the regulator etc. Use Kirkhoff's current 
law, with the injector as the current source..

Your circuit is best deployed not in the ECU, but directly across the 
injectors where the return path  is not going to affect other things. 
Then I'd agree with what you say.

davidh


>------------------------------
>
>Date: Wed, 12 Mar 2003 17:27:08 +0100
From: Marcell Gal <cell at x-dsl.hu>
>To: List for general do-it-yourself EFI talk <diy_efi at diy-efi.org>
>Subject: [Diy_efi] Re: new EFI controller
>Message-ID: <E18t94O-0004Dr-00 at al.x-dsl.hu>
>In-Reply-To: 
>	<412354F6931D994CBA71D0D76BF62639012AE348 at caebe001.europe.nokia.com> 
>References: <412354F6931D994CBA71D0D76BF62639012AE348 at caebe001.europe.nokia.com>
>Content-Type: text/plain; format=flowed; charset="iso-8859-1"
>MIME-Version: 1.0
>Content-Transfer-Encoding: 7bit
>Precedence: list
>Reply-To: List for general do-it-yourself EFI talk <diy_efi at diy-efi.org>
>Message: 6
>
>Hi Bill, 
>
>  
>
>>Define inefficient. It is all a trade off to do with steering the flyback current.
>>    
>>
>
>Did you look at the circuit carefully?
>( http://caffrey.dk/megasquirt/ schematic, last page, T3 )
>To connect (the zener-emulator) to GND instead of 12V is
>simply a mistake. It does not make the injector switch off faster.
>The additional 12 V just heats the transistor more
>instead of putting back the current to the supply for
>some useful purpose. There's no tradeoff there. Bad in every
>ways (heat,cooling,space,component selection, energy)
>without any  positive effects
>(the injector sees the same voltage, since you have to use
>a higher voltage zener that way). It even worsens predictibility,
>since switchoff will depend on supply voltage.
>That's what I mean by inefficient. 
>
>If you think of dissipating the injector current faster
>(zener instead of a schottky or diode), that's really
>a tradoff, that can be justified, (the MS discussion is about
>this topic, right?) but I'm NOT talking about that
>one this time. 
>
>Marcell 
>
>  
>



_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi