Injector Driver Module

[Greg Hermann]

>Mike, thanks for the response.
>
>  Yes, there is the method of doing PWM to hold the injector current. And
>that solves the power dissipation problem.
>
>  But it creates a HUGE problem with electromagnetic radiation (EMI),
>since the injector coil's current will be switched on/off, creating 6A
>peaks and big inductor spikes.
>
>  In order to reduce the EMI, shielded cables, etc etc would be needed,
>that could make the module very expensive or not practical.
>
>  Regards,
>
>Jose
>
>>>> Mike Morrin <mikem at southern.co.nz> 08/10/98 01:20PM >>>
>At 11:17 AM 8/10/98 -0700, Jose  Rodriguez wrote:
>>  There are mainly two types of injector coils: the saturated and the
>peak-and-hold ones.
>>  The saturated coils typically have 16 ohms resistance and yes, if I used
>them the power dissipation would be greatly reduced on the injector driver
>side, specially since I could saturate my drivers, so the power would be
>negligible.
>>
>>  If most OEM coils are saturated, that would explain why they can do it
>with such small heatsinks. I wish I could put the module in the interior,
>but that is not an option.
>>
>>  Unfortunately, I am currently being forced to use the peak and hold
>coils (1.5..2.3 ohms), because the injectors are to deliver a very high
>mass flow, and work over -40..125C. I am told the saturated coils can not
>reliably work under those conditions.
>
>I thought I saw somewhere a circuit which generated the hold current by PWM
>of the injector current, so the drivers were always saturated or off.  I
>cannot remeber where I saw it, or whether it just used the internal
>iductance of the injector, or an extra external inductor.
>
>regards,
>
>Mike
I don't care at all if it's more expensive, or harder to design (to avoid
EMI problems). The PWM approach is really the only way to fly on drivers
meant to run peak and hold injectors!
>From an old racer's point of view:
1. I would want the @#$% thing designed to take a 100% duty cycle, plus
about a 25 or 50% service factor. The injector drivers just shouldn't be
one of the parts which has the potential to bring you to a stop. Period.
2. A PWM type driver (on a board which is designed to quiet things down and
ground DC from the injector) could be completely programmable as to BOTH
opening current and holding current, as well as to duration of the opening
current. It could be programmed to let a jerk like me who wants to run a
straight 24 volt electrical system simply program the thing to deal with
it, rather than worry about the size of the heat sinks with a non-standard
voltage. It would also let a body do some development work to see what
opening and holding currents the injectors could really live with without
coming apart or frying.
3. Now that I'm getting warmed up, the really cool approach would be to set
the board up so that it feeds back to the CPU what the opening and holding
currents are for control and correction. Of course, this would be a more
stable control loop if the CPU were also looking at supply voltage and
pre-adjusting injector PWM signals accordingly. Anyone want to make book on
how many cold start problems would go away with this kind of an approach?
4. Why in @#$%^ would we have chosen to design a unit using a -332 based
CPU if the objective was to be like Detroit and only be elegant at saving
pennies?
5. On the duty cycle thing: there is only one way to see where you are.
Multiply WOT horsepower per cylinder by expected BSFC (for the purposes of
this calculation, use .5 lbs./BHP-hour for a naturally aspirated motor, .55
for a turbo motor, and .6 for a supercharged motor.) This will give you the
required flow, in lbs. per hour required for each injector. Dividing this
number by the rated flow of the injectors to be used will give you the duty
cycle (% of real time that the injector must be open in order to flow
enough fuel to make the power required. Period.
6. Since a high performance engine has a wider dynamic power range than a
stocker,(bigger ratio of maximum power to minimum power) it necessarily
needs a bigger turndown ratio on its injectors than a stocker (or it will
have a soggy, lumpy, messy rich idle, (or fried pistons) among other
things.
7. If anyone out there can make really decent power out of a motor, and
also get it to idle decently, without playing any of the games I was
talking about earlier with staging injectors, and run less than a 75 to 85%
duty cycle on the injectors, you've either got to have an astronomical
amount of parasitic drag at idle, or i'll kiss your #$%!
8. The reason that open and hold injectors are often necessary for hypo
motors is because they are capable of giving a consistent, reproducible
flow quantity at about a 1.2 to 1.5 ms open time, wheras saturated coil
type injectors can only get down to about 2 or 2.3 ms before their flow
repeatability falls apart. Thus, open and hold injectors will improve the
available turndown ratio by a factor of at least 1.33. Don't let the tail
wag the dog here. The ECU and injectors, working together,need to provide
as much turndown ratio as we can figure out how to get out of them in order
to serve their intended purpose.
9. Assume an idle speed of 600 RPM-that means that one four stroke cycle
(two revolutions) takes 200 ms. If your horsepower requirement at idle is 4
HP  (.5HP/cylinder) and your idling BSFC is .40 #/BHP-hr. (low, but that's
what you want to be safe with this calc.) that means that each of your
injectors needs to flow .2 lbs/hour of fuel at idle. If our engine is going
to make 320HP at WOT that means that each injector (go back to 5 above to
see how I get this) needs to flow 20 lbs/hr at WOT. So select 24 lb rated
injectors, and they wil run at an 83% duty cycle at WOT. At idle, these
injectors need to flow .2 lbs/hr, or a .8333% duty cycle. This is also a
100:1 turndown ratio. .008333 times 299 ms equals 1.6 ms open time for the
injectors at idle. Therefore, open and hold type injectors will do the job,
but saturated coil types WILL NOT (at least not without a lousy,
inconsistent, likely to be over-rich idle)!
10. Run a few calculations of this type, and you will begin to see why I
want to play games with things like staged injectors and feedback control
of variable fuel pressure for a really seriously high output motor. These
things are simply necessary to get a turndown ratio better than 100:1.
        Sorry--this got longer than I meant for it to. Hope it helps if you
plow through it.:)
                                   Regards, Greg