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

 Paul,

 I'm just starting construction of my fuel tanks for a Zenith CH-640.
 I'm opting for a four wing tank system. I'll likely be using a 16b
 conversion,
 fuel injection based on your designs. With the discussion on-going about
 in tank fuel pumps and fuel return lines, I am asking if I need to make
 accomodation for anything special. Presently, my tanks have no provision
 for
 a fuel return or in tank pump. Would it be necessary to design in any of
 this
 stuff?

 Thanks, C. Smith

 Yes you might think of adding a large removable panel in the root
 ribs for internal pumps. The 640 has a nice thick wing for internal
 pumps. Looks like the tanks are in the leading edge. If not
 practical use two lines from each tank. One for feed according
 to the plans and one for return fuel.

 Paul Lamar

 Correction, I meant 13b engine.
 C. Smith--


 So the EFI system with a fuel return line would preclude the use of a
 fuel flow meter to  keep track of fuel used and burn, unless someone
 knows of a meter that subtracts the return flow?
 so with a 4 tank system, thats 8 fuel lines, and a tank selector valve
 ganged 4x4? 4 separate fuel pumps, and control wiring?
 No other way to do this? I'm bummin' out.

 C Smith

 Unfortunately that is one of the drawbacks of a fuel return system.

 With the 555 system all injector pulses rates and widths are the
 same so one can build a simple flow meter that consists of diode
 and large capacitor that will sum the pulse width and rate from
 one injector. That sum can be read by a digital voltmeter to give
 you a flow rate. It is calibrated with a pot so it can read
 out GPH to the nearest .01 gallon. Very simple. Computers systems
 do it the same way except they do it in software.
 I might build the diode, capacitor and pot into one of the PC
 boards. You have your choice of read outs. One can probably test
 the scheme on a car EFI system.

 Paul Lamar

 Just a quick reminder (in case anyone forgot) that the rising-rate
 vacuum-referenced fuel pressure regulators are intended to be 1:1 and
 increase fuel pressure 1psi for every 1psi manifold pressure increase
 to keep a *constant differential pressure*.

 By design the differential pressure across the injector should be
 constant at all times, irrespective of manifold vacuum or boost.

 Ratios other than 1:1 exist, but those are generally for radical
 supplemental fuel pressure regulators used when applying forced
 induction without any other means of fuel enrichment. With rates like
 7:1, these are a fairly bad idea.


 In other words, there is no necessary modification to a 3D fuel table
 to take this action into account. The injectors can be assumed to
 operate linearly (neglecting opening/closing time effects). Even if
 the differential pressure was changing, there is no need to have a
 separate fuel table and pressure sensor, as this information will be
 incorporated into the 3D map when the engine is tuned.



 Paul: Regarding flow measurement from injector pulsewidth, remember to
 consider injector dead time (opening/closing) which becomes very
 significant at low duty cycle or high RPM. Different injectors have
 different properties in this department, which vary with applied
 voltage.


 -Chris

 No computer. No 3D maps. The load on an aircraft engine never varies
 with fixed pitch prop load. It is a much simpler system than a car
 engine.

 If we don't use a rising rate regulator the engine might be too rich
 on the bottom end. I am just using it to supplement the linear mass air
 flow sensor voltage controlled TLC 555. Just another way of tailoring
 the curve. It may not be necessary.

 I have a scope hooked up to monitor the dead times.

 Paul Lamar


 Paul,
 the reason we use manifold pressure to tweek the fuel pressure even on
 production cars is to keep an even pressure differential across the
 injector. on boosted applications, this is even more important.
 The simple cure to all these controls is to pwm the pump. It takes 2
 sensors, one on the fuel line, and one to monitor manifold pressure.
 The one
 on the manifold pressure can be quite accurate even with all the pulsing.
 all that is required is a long length of 1/8" od thick wall plastic
 tubing.
 the friction of moving the air through the long length of tiny tubing
 provides the damper to smooth out the signal.

 We have a pwm module we already use for controlling dual electric fans
 based
 on engine temperature. it may be simple for us to modify it to control
 fuel
 pumps instead.

 If you are interested, I will find out what the production sensors
 ford uses
 on returnless systems and work on adapting them. We can drive any external
 pumps with this as well as any of the submerged pumps.
 Aeromotive already sells a pwm drive module, I just don't like the price.
 Check out the pump the jegs is selling. I like the fact that you can
 use an
 fittings.
 http://www.jegs.com/p/JEGS/1167179/10002/-1
 Larry

 Not interested right at the moment as I want to keep this system as simple
 and cheap as possible with a lot of history. Also it is a risk as it is
 designed for cars and the PWM pump system may not be up to aircraft high
 duty cycles. Down the road may be OK.

 BTW I have upgraded the write up on the web site so check it out.
 http://www.rotaryeng.net/simple-cheap-555.html

 Paul Lamar



Just caught up on this thread & noticed the comment on:
"So the EFI system with a fuel return line would preclude the use of a
fuel flow meter to  keep track of fuel used and burn, unless someone
knows of a meter that subtracts the return flow?"

This product does exactly that:
http://www.matronics.com/fuelchec/RFC/index.htm

Dynon might have one too, I recall something in their installation guide
about this topic.

--Kenny A.

I appreciate the information everyone has contributed. The Matronics device
does do the math, but it appears it does nothing else but produce a pulse
train that must be processed by another device, why is it called a controller
when it controls nothing, it is a flow totalizer. The price is kinda high too,
about 1/3 the price of a used 13b.
What I'm looking for would provide the data in a format acceptable to an EFIS
like a Garmin G-600. I'm wading through the data on the Dynon site.
Thanks for the information.

I am not trying to be critical of the system Paul is engineering, he
has specific design goals. I was just hoping for another option. The
Fuel injected Cessnas I've had, didn't use a fuel return and simple fuel
flow sensors were easily installed and interfaced with the GPS.

In the end, I could probably
 engineer a device to do what I need.

I think I'll have to hold off on the tanks until Paul has finalized
the system, and I have a better understanding of the fuel system
requirements for the rotary engine. I can easily add an extra fitting
to the tank, but in tank pumps has got me worried. Does the pump fit
in the top, bottom or side? the skins are .025 (already purchased the
material) I don't know if that is suitable to support the pump.
Rough tank dimension is 32.6"x18.5"x9". Changing the material for the
side bulkheads is doable.
Sorry to bore the group with my questions, but I'm not an engine
expert and I want to get it right.

C. Smith

In my 555 system there is no need to have the pumps in the tank
if you have pumps that have fittings on the intake side (MSD
and others).
Here are a few plumbing suggestions for return systems.

Paul Lamar