Injector firing circuits

[pds at iaccess.com.au]

Hi, I'm in the process of transplanting a Ford throttle body injection unit 
from their Australian Falcon inline 6 cylinder engine onto my Peugeot V6. 
The idea is to replace the elaborate carby system (one single barrel and
one twin barrel plus levers vacuum acuators etc. etc) with efi.

I've had a manifold welded, machined and drilled to accept the TBI unit. The
computer I've designed is based on the Dallas 80C320 running at 24 Mhz. 
Essentially it's equivalent to an 8031 running at about 66 Mhz. I designed 
it for another project and I have 4 left over. I'm going to use the 
MAP sensor from the falcon along with the throttle position sensor mounted 
on the throttle body. Water temp will come from the existing sensor 
mounted on the water pump. Air temp is from a sensor mounted in a hole 
drilled in the manifold near the right rear cylinder intake.

The two injectors used in the falcon seem to be peak and hold types.
Scoping up the ECU up reveals a 700-800 uS pulse at full battery volts 
followed by a variable time current limited tail, with 50 volt spikes at
the transitions.

The question is what is the =best= circuit to use to achive this type of drive.
I've looked at IRF530 powerFETs but the drive circuitry seems too fiddly. It
seems that they need about 160 mA to overcome the gate capacitance. Bipolar
transisors all seem to have very low current gains and take a long time to
switch off. 

This has led me to the Philips BUK101-50GL device which seems to be designed
as an automotive solenoid driver. The data says they can be connected directly
to the micro. As the 80C320 doesn't have a PWM output I may try the 
following circuit. Tweaking the RC will get the 700-800 uS peak pulse 
while the other half will hold the tail. The BUKs have protection built in 
too so that's less work for me.

Incidentally, if you used a PWM output to do the current control, wouldn't
that cause lots and lots of EMI? Thinking about it, why don't efi engines
cause lots of radio interference. The injectors cause quite large (50 volt)
spikes which usually cause emi.

                                               ----------*-> To injector
               |\              |\   |\      |--|         |
Input   >------| 0-*-||-*---*--| 0--| 0-----|            |
Bank 0         |/  |    |   |  |/   |/      |--|         |
                   |    |   \                  |         |
                   |   ---  /                 ---        |
                   |   /_\  \                  -         |
                   |    |   /   2 x Philips BUK101-50GL  |
                   |    |   |                            |
                   |    ----*              current limit |
                   |        |                  ---\/\/\--|
                   |       ---                 |  resistor
                   |        -               |--|
                   -------------------------|
                                            |--|
                                               |
                                              ---
                                               -
Comments please from budding efi designers!!

As you've probably guessed by now, my background is electronics, computers and
communications. So, I don't really know what I'm doing here, which will 
probably help me, because if I did know I probably wouldn't have started this 
project.

//----------------------------------------------------------------------------
// Phillip Bertolus                                         _  |\  
// Program Development Systems                             / |_| \   
// 57 Purtell Street (P.O. Box 21)                       .'       \
// East Bentleigh, 3165                                 /          \
// Melbourne Australia.                                 \     _    /
// Phone: +61 3 563 7755                                 \_.-' \_*/
// Fax:BBS:FaxBack:Voice Prompt:Network: +61 3 563 7855          v