A repost

[Tony Bryant]

> 
> I do not think that the above circuit will just keep the injector
> open. Most injectors take at least 1 millisecond to open. And
> what I have been told is that it takes roughly the same amount
> of time to close. The driven frequency of the circuit is 10 khz,
> or a period of 0.1 milliseconds, at ~50% duty. So, with a starting condition
> of the injector being closed, the injector will open roughly only 10
> percent before it closes again. Does the injector take longer to open this
> little mount than to close? I do not know, but instinct (only) tells
> me that it probably takes a little longer to open the injector than
> to close it (open requires overcoming spring force + inertia + charge-up
> time of injector solenoid (inductance), whereas closing is inertia +
> collapse time of solenoid). I have not thought up of all of the forces
> acting on the injector for opening and closing, so I may
> be in left field. But 50% duty is a good start, and it works
> for me.
> 

Looking at it from a purely electrical point of view:

Since the current supply to the injector when directly hooked to a 
12V supply is about 5A (assuming 2.4Ohm injectors), a fifty percent 
duty cycle gives 2.5A which is enough to open them very hard. Perhaps 
that will bounce the pintle off its stop, giving the appropriate 
vibrations.

My experience tells me you only need 300mA to hold an injector open 
(not under fuel pressure). Therefore a average 300mA current causes 
just enough force to balance the pintle spring.

So thats 6% duty cycle.

But another thing to consider that dumping current into the mostly 
inductive inductor is done at 12V, but the injector dumping it out is 
at 40V (or whatever zener voltage you pick), 

and since V=-LdI/dt, the rate of change of current is 
proportional to the voltage that the change is being done at.

so it will discharge at 3.3x (=40/12) the rate of charging - 
logic says that you need about 3x the duty cycle to hold the same 
current. Hence duty cycle of about 20% to hold injector at average of 
300mA current.

Oh well enough theorizing, I guess I'd better backup my ramblings 
with some experimental evidence :-)


'71 510 Daily driver
'72 510 Race car
'85 V12 XJ-S Garage ornament