Love This

[Chris Conlon]

At 02:01 PM 1/29/99 -0700, Greg Hermann wrote:

>>With this device a closed loop electrically controlled Fuel pressure
>>trimmer is possible, meaning big injectors, lower pressure at idle,
>>to help get reasonable pulse widths.
>>Bruce
>
>Hey--if you wanna do it right--use a rising rate regulator IN COMBINATION
>with a pwm voltage control to the fuel pump AND a calculated fuel pressure
>correction to the injector pw calc. It would hafta be a primary--secondary
>fuel pressure control loop to be stable--program what the fuel pump is
>supposed to do according to load/rpm ()with a look-up table), let the

My take on this is that the first level is a simple feedback loop,
adjuting fuel pump duty cycle to try and keep fuel pressure constant.
Also the usual pulsation damper (or whatever they're called), and
somewhere for vapor bubbles to go. A float valve sounds like the Right
Way, but it doesn't sound like they're easy to get, not ones that are
qualified for fuel rail duty anyway. How about just using a small
bleed orifice, like a nitrous jet, feeding the same old return line?
You'd end up with a *little* more fuel heating but it's cheap, easy
to do, safe, reliable, and the fuel vapor would have someplace to go.

That being said I don't think the first level is very good, because
the loop bandwidth will be low. Steady state is fine, transients, as
always, will be less good.

So next your *actual* fuel pressure becomes an input to the ECU, and
adjusts your injector pulse width. I think a somewhat larger than
usual pulsation damper might be good; fuel rail pressure probably
has a lot of the same pulation issues as intake manifold air.

The next stage is where Bruce wants to be, the desired fuel pressure
is settable, and the actual current fuel pressure is still used to
adjust injector PW. The *possible* shortfall here is when you suddenly
need a lot more fuel, but since you are still measuring FP it may not
be a big issue.

The next step from there would be to try and get better fuel pressure
response time, via a PID-controller type of setup. Use delta-MAP *
RPM or some other measure of changing load to try and put more power
thru the fuel pump *before* dropping fuel pressure tells you that
you need to. But either way you're still using the actual FP to
calc injector PW, so you're still way better off than before.

The reasons I don't think you need a mechanical regulator are
(1) You're already measuring FP to a high degree of accuracy. The
resultant injector PW is going to be a fair bit more accurate than
using a mechanical regulator and assuming it's keeping FP constant
within say +/- 0.5%. FP variation is somewhat inevitable, at least
now we can measure and correct for it.  (2) A large pulsation damper
gives me the feeling of a safety net wrt possibly large, very fast
FP changes caused by individual injectors opening and closing.
(Which may not be easily calculated out due to high BW.)

Now the next step is putting a fuel temp sensor just before the
rail and compensating for fuel temperature too...


Hell, I'm sold. Even if I can't get one of those specific Moto units,
I don't see why I can't get *something* to do the job, and a PWM
controller for the fuel pump is easy. A module from
www.powertrends.com is probably more than good enough. I'm just
glad this thread came up in time for me to use the ideas on my
current project engine.


   Chris C.