<snip
 I can see why Ford went to PWM pumps. I can also see why
 carbs are still popular in certain forms of car racing :)
 Paul Lamar

 Didn't someone say that the RX-8 was a return-less EFI system? If so,
 is the RX-8 system PWM or just internally regulated at the in-tank EFI
 pump?
 -Vince Orton

 Here is the RX8 fuel system schematic. Does not look like
 PWM to me. Lots of relays in there.

 Looks like it is done mechanically inside the pump
 at 56.6 psi.

 Until you brought it up and I looked it up I did not
 realize that perhaps the RX8 system is the way to go.
 No fuel pressure regulator. That is good and bad.
 Saves the cost of a regulator but adds the high cost
 of an RX8 pump and the pump must be in the fuel tank.

 There is a high and low speed switch as well.
 That complicates control. The computer controls
 this according to the demand for fuel. At low power
 levels the pump is slowed increasing the life of the pump.

 Perhaps this system should be investigated before
 building a PWM system.

 I think I will continue to pursue the intake manifold
 pressure controlled regulator route as it is simpler
 to deal with, has a long history and it has no requirements
 for mounting the RX8 fuel pump in the tank. Any pump
 can be used. This has the advantage of automatically widening
 the dynamic range of the injectors. No staging of injectors
 is required like is done in the RX8 fuel system.

 Perhaps some day an after market pump will be developed
 with an internal pressure regulator that does not require
 a two speed switch and is durable enough to run flat
 out forever. Aircraft engines don't have a low power cruise
 mode that is less than 50 HP like a car.

 Bottom line: I don't think the RX8 pump is up to it :)
 Paul Lamar

 --------------------------------------------------------

 Paul -

 My thoughts on your comments above:

 You're probably right on the RX-8 system using relays (not PWM). The
 diagram on page 1-14-3 indicates a multi-speed control for the EFI
 pump (probably two-speed). Not a bad solution for a simple,
 internally regulated pump. I can't find where you got the 56.6 PSI
 figure from on the manual pages you attached. Not important other
 than we have a target pressure for RX-8 injectors.

 Assuming that internal-tank PWM EFI pumps are a modern automotive
 standard, I prefer to use them. Many, many parts available, and well
 proven. (Millions of cars, right?)

 The RX-8 EFI pump seems COMPLICATED - jet pumps, sub-tanks, fuel
 reserve caps, etc. Things we'll never use in aircraft. I'd rather
 NOT base an aviation design on the RX-8 EFI pump. Let's stick with a
 PWM-compatible in-tank EFI pump available anywhere. 'Nuff said.

 Varying fuel pressure to increase injector dynamic range is a good
 idea, but only within limits. Remember that there are manufacturing
 variations between injectors that result in very minor flow
 differences at high pressure. However, reducing the fuel pressure
 significantly will result in reduced fuel atomization and magnified
 variations in fuel flow between different injectors. Again: a good
 idea within limits.

 Regarding the last paragraph: There is no reason that an EFI system
 that uses a MAP/RPM lookup table to determine injector pulse widths
 cannot add another table to determine fuel rail pressure as well.
 (The PWM-controlled pressure will slave off of that lookup value.)
 Viola! Increased dynamic range.

 Question: Are you saying that off-the-shelf (OTS) simple 12V/"run flat
 out" internally regulated EFI pumps don't exist? I assumed they did.
 (Oops) I am also assuming that electronic fuel pressure sensors exist
 in everyday average cars. Are these valid assumptions?

 I think we all agree that the pressure regulator (mechanical OR
 electronic PWM-based) optimally will regulate AT THE FUEL RAIL to auto-
 compensate for head pressure changes during climb, descent, and
 straight-n-level flight. If regulated in-tank at the EFI pump, the
 pressure will vary slightly at the rail, probably insignificant.

 *****

 My offer still stands. If someone (Lynn? Tracy?) has some RELIABLE
 MASS MARKET parts to recommend for:
 - PWM in-tank pump with integral check valve (CRITICAL: recommend!)
 - electronic fuel pressure sensor (CRITICAL: recommend!)
 - PWM controller (recommend, but I can design also)
 Then I will build a rig for the design and testing of a standard "open
 source" aviation PWM fuel pump system with redundancy built in.

 I'd prefer part numbers from Ford, GM, Mopar, Toyota, Mazda, etc., and
 if possible, make/models I can scavenge the parts from. I'd prefer
 OEM to catalog (Summit, JC Whitney, etc.)

 I am NOT in the automotive systems info flow, so recommendations from
 the list will REALLY help jump start me (hint, hint). Otherwise it's
 Googleland and hoping I find good parts sooner than later.

 By the way Paul, I'm at Chino airport, not Bracket.

 -Vince Orton


 "Are you saying that off-the-shelf (OTS) simple 12V/"run flat
 out" internally regulated EFI pumps don't exist? I assumed they did.
 (Oops) I am also assuming that electronic fuel pressure sensors exist
 in everyday average cars. Are these valid assumptions?"

 I don't know about that. Could be (other than the RX8) I have not
 heard about them. In any event they have not be tested in a
 continuous high power AC application to my knowledge and
 consequently the life is suspect. I know the Mooney elect. boost pump
 needs turning off after takeoff as it can wear out quickly. I had
 one fail on me and the carbon vane particles jammed the mechanical
 pump causing the engine to quit. Fortunately I was close to an airport
 and I had to dead stick it in.

 I don't see a fuel pressure sensor in the RX8 schematic.

 There is a description paragraph which was not worth scanning
 that specified the 56.6 psi setting.

 The only time we need lower fuel pressure is when the engine is idling.
 During normal operation it is probably not needed as I will
 set the pressure on the high end to be sure it is rich enough
 for full power on p-port engine. If four yellow RX8 injectors don't
 put out enough fuel on the top end I will go to four larger injectors.
 In that case it might be too rich on the low end
 idle. As a rough first measurement the fuel pressure drops 10 psi
 when the manifold pressure is 15 inches of Hg absolute. That is a crude
 measurement and I will get a better reading when I get the vacuum
 system sealed. Atomization is of little consequence when an AC eng.
 is idling. More important in a car.

 It is sometimes difficult to mount fuel pumps in AC wings.
 Most people don't do it. They mount the pumps on the firewall.

 Paul Lamar

Can you tell me why you would want to use a manifold pressure sensing
fuel reg. on a n/a aircraft engine?

Skip Gorman

Simple. It expands the dynamic range of the fuel injectors.

Paul Lamar