[Diy_efi] Re: injector current-feedback driver

[Marcell Gal]

Hi Mark, 

> The Qfeedback needs to be a PNP transistor to work with the Rcurrent
> connected as shown. 
> The Schottky diode is not connected in the right place

yes, you are absolutely right. I fixed those typos. It must have
been very late last night ;-)
http://x-dsl.hu/genboard/injector/ 

> that could be made better, and the dependence of frequency on a lot of
> variables in the circuit components would make this circuit pretty hard to
> predict in exactly how it would perform,

the inductance and supply voltage determines di/dt
through the injector. Rhisteresis/R2 determines
the histeresis voltage. This together determines frequency.
Is it that bad or unpredictable? Or are you afraid of
the temperature dependance of Qfeedback? 

>> Can 4 paralleled HC7404 inverters drive an injector driving
>> NPN transistor?
> ..
> the injector, then 4 paralleled gates do not have nearly enough output to
> drive Q1 as shown. (I assume you want to saturate Q1, and use the pulsing
> feature to provide the current limiting.) 

yes, your assumption is correct. 

> Since FET's do not need actual
> current to keep them on, they are easier to drive with logic gates; that
> would be a better choice for this design.

What if Q1 is a darlington?
What FET do you recommend for max 6A? Do I need a FET-drive
than (like megasquirt) or 5 Volt HC output is fine for
driving FET gate? 

thanx:
 Marcell 

Maybe one other change would be right:
To connect Rhisteresis directly to Qfeedback base,
but the common point of R2 and C2 via a resistor.
That way C2 could not unwantedly slow the positive
feedback effect of Rhisteresis...

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