Next question: Fuel pumps....

Lawrence E. Piekarski c1ilep at kocrsv01.delcoelect.com
Thu Aug 31 16:14:53 GMT 1995


>I need a small, low-current-draw, high pressure, low (relatively) volume fuel
>pump/system.
>
>Engine is a 600cc single cylinder motorcycle engine with 8000 max rpm.  Fuel
>demands work out to about 33 lbs/hour.  Current plan is to use 2 19lb injectors
>from a Mustang.
>
>Most (all?) automotive systems circulate fuel to the regulator and back to
>the tank.  This idea seems great for keeping the fuel cool and allowing
>quick response to larger fuel demands (less inertia of fuel in the line to
>overcome).  My problem is that I don't have a lot of current available for
>the injection system and I'm trying to save a few mA here and there (engine
>is on a Honda XR600R dirt bike).  I don't want to put too big of a pump
>in the system and waste current pumping a lot of fuel through the system.
>Are there any guidelines to follow for sizing fuel systems? 

Well, I have two thoughts.

1) Any any system, there is a spectrum of users, you have you 600cc dirt
bike and I have a 527 c.i. V-8 tractor-pull motor. Still, I just went through
what you are asking about. I have twin 900 cfm carbs mounted high atop a 
tunnel ram with a Mallory fuel pump mounted below the frame rails right next
to the fuel tank. The fuel pump had a built in regulator, which I had 
adjusted to about 8 psi. Problem: I idled out on the track and tightened the
chain, dropped her in gear and dumped the throttle. Since there was about
3 feet of fuel line between the pump and the carb, here is what happened.
When I dumped the throttle, the fuel began pouring into the engine (and 
I do mean pouring). The fuel pressure at the carb dropped to near zero. 
It took on the order of 250ms for the zero pulse to make it back to the
fuel pump regulator, at which time the fuel bowls were running dry since there
was no fuel line pressure to fill them. At this point the regulator notices
the zero psi fuel pulse and drives fuel up the line virtually unregulated.
Now, there is so much fuel pressure that the float holds back what they can,
but fuel starts running out the top. 250ms the regulator notices the excess
pressure and stops pumping fuel, and they cycle starts again. What I noticed
is that when I dumped the throttle, the fuel pressure would drop to 0, spike to
15+ psi and so on in a damped sine wave of about 2Hz. After 5-10 seconds, the
regulator finally "got a clue" and we were running. Naturally, the engine 
didn't run good with the alternate leaning/flooding going on.

Solution: I placed the twin regulators right at the carb. This eliminated the
250ms time constants. While you won't use near the fuel I do, you also don't
have a fuel bowl reserve, so constant pressure is imperative. The reason
for the regulator at the fuel rail is not to cool the fuel (though it is a 
nice side effect) but to ensure that on idle -> WOT transitions, you don't
have a drop in pressure which would cause leaning. You could probably 
compensate for this in software, but it seems needlessly complicated.

2) I think that is would be easier (i.e. less current) to pump moving
fuel back to a tank than it would be to apply pressure to stationary
fuel. Electric motors generate back EMF, they draw the most current at
a locked rotor condition and the least at wide-open-no-load. I would think
your current draw would be less actually pumping fluid (higher rotor speed)
and applying pressure to a non-moving or slow-moving fluid (less rotor speed).

In my thinking, it is easier to push a car 125 feet down the road and then
push it back 25 feet, than to push it 100 feet with the brakes partially
applied.
-- 
(* Larry Piekarski, Lead Software Engineer *)
(*      Luxury Car HVAC Display Heads      *)
(*      c1ilep at kocrsv01.delcoelect.com     *)
(*      Delco Electronics, Kokomo, IN      *)



More information about the Diy_efi mailing list