[Diy_efi] injecotr flow bench plans needed.

[Brian Dessent]

> injectors. I will want to be able to adjust the pulse width for the
> injectors so I can test many different setting. I have plans and such
> for the mechanical end of the project, I just need some help ont he
> electronic end. I have searched the archives with no luck. Thanks for
> the help!

I would use something like the the CA1524, which is designed for doing
closed-loop PWM controls, but you can just as easily attach a
potentiometer from V_ref to ground to create a voltage divider which is
used as the input control voltage.  This would allow a duty cycle of 0%
to about 95% with the turn of the knob (you'll need both A and B outputs
in parallel to get the DC >50%.)  However, the chip is designed for
switching power supplies and stuff, so its oscillator is usually run in
the tens to hundreds of kHz range, which is probably a lot faster than
what you want.  Let's see, the rated max values of the timing R and C
are 120k and 0.1uF, that comes out to a period of 12ms.  That would be
the longest you could expect, which seems pretty short for injectors, so
never mind.  Data sheet here if you're curious:
http://www.intersil.com/data/fn/fn1/fn1239/FN1239.pdf

Well, I'm sure you could do it with a 555 as well, look for some "555
cookbook" sites or books and I'm sure you'll find one.  The problem with
555s is that usually you don't get control of the duty cycle and the
period independently.  What might also work is running one 555 as an
astable multivibrator (i.e. a pulse stream) which triggers a second 555
(or other) that's configured as a one-shot, whose duration is varied by
your potentiometer.  This lets you control both the frequency and the
pulse width separately.  The down side is that you'd be controlling 
pulse width and not duty cycle.  The difference only really matters if
you intend to vary the frequency, which I don't see as being necessary. 
A 556 is two 555s in one package, BTW, so undoubtedly there's a circuit
to do just this floating around out there.

And of course you'll need to select a driver transistor of some kind. 
These days you can almost just pick any power MOSFET with sufficient
specs and not worry too much about the gate drive circuit.  Just make
sure that there's a protection diode in anti-parallel (anode to ground,
cathode to drain -- assuming low-side switching.)  Most power FETs have
such a diode integrated into the package.  You might also want to
consider one that's descibed as "logic level" which usually means that
5V is enough to fully turn it on.  Suggest copying whatever the
Megasquirt folks did for their drivers if any of this is unclear.

Brian

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