[Gmecm] "Peak and Hold" injectors on a saturated driver

[bcroe at juno.com]

OK, finally tried my idea.  This is for you guys driving 
P&H injectors with a saturated driver and a series 
resistor.  Another variation is mine (AC*EL VII), a 
current limited driver driving the P&H injectors in 
my 79 Cadillac, but the current limit is way too 
high, and the power gets too high at max power.  

This is not a fancy schmancy design with buffers 
and new drivers, digital access, or programability.  
Your drivers still do the work, with a little pulse 
shaping added.  You manually adjust a resistor 
to set peak  time, and you select a power resistor 
to limit the hold current.  

So you are using the power resistor (between the 
injector and the driver), selected to allow the injector 
rated hold current.  For each driver.  I am adding a 
MOSFET across the resistor to short it out.  This gives 
the peak current the injector wants.  After the desired 
peak time interval is passed, an RC ckt driving the 
MOSFET times out.  The MOSFET turns off, more 
gradually than that sharp edge from the standard 
P&H, but fast enough.  The injector is already open, 
so the exact time doesn't matter.  

When your driver turns off, the injector closes and 
kicks back through the hold current limiting resistor 
to your driver kickback clipper, probably set around 
30 V to speed up injector closing.  The injector will 
go another dozen volts higher because of the drop 
through the hold current limiting resistor, but this 
won't damage anything.  The infinitesimal decrease 
in injector closing time is nothing compared to the 
faster opening.  Those who already have series 
resistors were already living with this extra kick.  The 
RC circuit resets itself for the next injection cycle, and 
we are ready to go again.  Those who want to get the 
kickback pulse back to the original value (why bother?) 
could add your own diode network around the injector 
coil.  

You are using your original driver here, which is still 
in control of the pulse width.  It must have the ability 
to drive the peak current, for at least the 2 ms or so.  
And you will need a circuit for every driver, that's 8 for 
my sequential injection.  Each driver needs that 
power resistor you were using before, and about 
half a dozen other parts.  That's about 50 for me.  
Now that I have that AC*EL tamed, I can pay 
attention to tuning.  

Bruce Roe





30 Jan 2005 "Jeff Stevens" <9secz at comcast.net> writes:
> > no problem .......just need to use resistor pack to match 
> > the impedance for the delco to be happy.  [snip]

> Using a resistor pack denies the low impedance injector 
> the "peak" current it was designed to use to open quickly.  
> Yes, the injectors will run and the ECU drivers won't burn 
> out, but the injector opening times will be *much*
> slower than if you used a proper peak and hold driver.
> 
> What I have seen on my injector flow bench is a wide 
> variation in minimum pulsewidth the injectors need to 
> open when you limit the current.