Alternative actuation of MFI fuel rod

[Vintagemotion]

Does anyone have experience with the below or will be able to help with input on 4.2, 4.3

1. Project:
Now that my MFI system is running great I like to conduct an experiment with a spare MFI fuel distributor that I bought.
The idea is to take advantage of the inherent MFI benefits (high pressure injection close to the inlet valve) and combine this with modern technology for:
a)close loop control
b)easy and better adaptability to different engine modification
c)better cold start performance (easier on the engine)
d)easy experimentation with fuel curves (maybe even as input to create new or optimized space cams)
e)better fuel economy (argument for the wife to justify the expense )
f) interest in a winter tinker project

2. Approach:
Substitute mechanical controls (centrifugal, barometric and temp) with a bolt on modern industrial actuator that is connected to the fuel rack. Based on the architecture of the MFI distributor it should be possible to maintain the originality so that it can be converter back.
The actuator would be controlled via aftermarket ECU like Megasquirt etc. with input from broad band O2, TPS, MAP. Hardware first:

4. Requirements:
These are the inputs for the layout of the actuator:
1.Determine complete stroke of the fuel rack from shutoff to max open = 15mm
2.Determine or estimate actuation force
3.Determine and exceed actuation speed (fuel rack movement) of existing system.
4.possibility to actuate position with ECU rather than frequency
5.operation of actuator on < 12 volt
6.resistance to 100 Celsius sustained ambient temperature
7.200mm max length (unless linkage repositions the actuator away form the back of pump)
8. vibration durability

Here is an example for an actuator:
Product Catalogs - Intelligent Actuators

Does anyone have experience with the above or will be able to help with input on 4.2, 4.3, 4.4

[jpnovak]

Cool project. I will be watching.

I am sorry I do not have any of the information you are requesting.

[Jeff Higgins]

Yes, very cool project. Like Jamie, I don't have that information, but it should be easy enough to get all of it simply by playing with your spare pump.

It seems entirely possible to remove the entire mechanical governor mechanism and replace it with a "smart" system running of various feedback inputs.

Bosch literature specifies fuel delivery rates for throttle position and rpm for the various Porsche engines running MFI. With access to that information and a pump dyno, you could easily set a baseline "map" . Unfortunately, neither the information nor the dynos are easy to come by. Our own Pelican "356RS" has built his own dyno, and has fuel delivery specs. He would be a good guy to get ahold of and start a conversation with.

[356RS]

This is great. I'm in. I can get the answer on 4.1 today and work on the rest this week. I'm working on a few customers pumps right now but would be glad to help gathering the information you need in a few days.

[philippeF]

I'm online too ....

Mark> bad springs for me

[Flieger]

Would this still count as Mechanical Fuel Injection?

[356RS]

Electric rack with mechanical pump?

[Vintagemotion]

Tonight I measured the stroke of the rack and the answer for 4.1 is it is 15mm (from complete open to shut off)

[jbell959]

I looked into doing something similar about a year ago. I was fitting a linear actuator in place of the thermostat but I haven't had time lately. There is a store not far from my house that sells a bunch of different actuators.

http://www.anaheimautomation.com/products/stepper/stepper-linear-actuator-list.php?cID=52

[HawgRyder]

You have to understand that the movement is not linear.
Because of the cam...the valve sizes...the port velocities...and probably several other things, the movement of the rack (enrichment) will be non-linear.
The controls necessary for this will be quite intricate.
But all said....the best of luck with it....and please keep a meter on the exhaust to check every step of the way.
Bob

[Vintagemotion]

Valid concern Bob if an industrial controller would be used. The plan is to use something like Megasquirt. This system in not linear like any fuel injection system. The Megasquirt guys have some fuel curves that can be adapted and manipulated with a PDA while engine is running. That is the beauty of using a programmable controller that is already setup for this.

Unless I’m missing something but the main challenges are:
1.Convert the variable frequency output of a Megasquirt ( or such system) to a variable voltage output, see diagram.
2.Deal with the inherent RPM variable output of the mechanical fuel delivery.

[356RS]

Rolf, Your answer for 4.1 "total rack movement = 15mm" is wrong. Using a 2.4 S (013) pump I recorded the following: Starting with the main rack pushed all the way back, we'll call this 0 mm. At 400 rpm & throttle closed (normal idle) the rack moves forward 2.25mm. At about 2.0mm the fuel starts flowing. Then at 3500 rpm and full throttle the rack moves forward to 7.25mm. On a 2.7RS pump I get 8.25mm with 3500 rpm and full throttle. So the normal rack movement is anywhere from 7 to 8+ mm.

[tctnd]

It makes perfect sense to use all the normal EFI inputs and process the signals to drive an actuator that controls fuel output from the MFI pump (I think that's what you are talking about here). It should be more efficient than the mechanical controls could ever hope to be and programming for a specific engine would be much easier. Presumably you would do this to retain the appearance and coolness of the MFI but gain the advantages of EFI. The same idea would work with a CIS system just by removing the air sensor plate and using a servo/actuator to control the arm.
regards,
Phil

[kenikh]

I looked into this, too, using an electrical servo tied to the thermostat plunger, utilizing an ECU to set fuel trim from Wideband O2 and TPS.

The 935s used a vacuum driven actuator to drive an enrichment circuit the same way, so it is a proven means to get there. You would still have to tune the mechanicals to baseline, so it souldn't be fully electronic, but would be simpler to implement with 9/10 of the capability of a fully electronic system.

[Vintagemotion]

Originally Posted by 356RS

Rolf, Your answer for 4.1 "total rack movement = 15mm" is wrong. Using a 2.4 S (013) pump I recorded the following: Starting with the main rack pushed all the way back, we'll call this 0 mm. At 400 rpm & throttle closed (normal idle) the rack moves forward 2.25mm. At about 2.0mm the fuel starts flowing. Then at 3500 rpm and full throttle the rack moves forward to 7.25mm. On a 2.7RS pump I get 8.25mm with 3500 rpm and full throttle. So the normal rack movement is anywhere from 7 to 8+ mm.

Mark,

Thanks for the correction. I took the measurements on the extra MFI pump I have which is not an S pump (019 ??). With the covers off I found the position in which the clamp screw in parallel to the rack and measured 7.5mm movement in one direction. Assuming for a worst case that mechanically the other direction would allow 7.5 as well. I felt it was conservative to use 15mm as a max for the actuator layout. Your dynamic measurements on the bench tester are much more accurate of course.

Rolf


This is my extra pump

[jbell959]

Rolf.. the pump in the first picture was re-stamped 019 and also stamped with a 2.7. It looks like that pump was modified to 2.7RS specs at some point.

You other pump is a euro 2.4S (021) pump. That should have the same specs as the (013) pump that Mark is getting his measurements from.

[Vintagemotion]

Originally Posted by kenikh

I looked into this, too, using an electrical servo tied to the thermostat plunger, utilizing an ECU to set fuel trim from Wideband O2 and TPS.

The 935s used a vacuum driven actuator to drive an enrichment circuit the same way, so it is a proven means to get there. You would still have to tune the mechanicals to baseline, so it souldn't be fully electronic, but would be simpler to implement with 9/10 of the capability of a fully electronic system.

Best case I was hoping to only use the rack and pump elements without the second half of the of pump all together (mechanical controls). Instead a very small actuator would be mounted.

[Vintagemotion]

Originally Posted by jbell959

Rolf.. the pump in the first picture was re-stamped 019 and also stamped with a 2.7. It looks like that pump was modified to 2.7RS specs at some point.

You other pump is a euro 2.4S (021) pump. That should have the same specs as the (013) pump that Mark is getting his measurements from.

Thanks for clarification. Not to get side tracked form the “Mechatronic” project…. But now I"m curious since my S is suppose to be US spec?? In regards to the extra pump: If the 019 designation is an RS spec what would it have been before? The 2.7 stamping seems original. Here are the other plates if that would give a clue.

[jbell959]

Hi Rolf... Your re-stamped 019 pump was a 2.4S 013 pump based on the other plates. None of the pumps (that I'm aware of) were stamped with a 2.7. That was done by the person that modified it as an RS pump

[Flieger]

What kind of sensors are you going to use for your computer inputs? The computer will need to know the air pressure, temperature, engine speed, etc. in addition to throttle input. Are you going to use a pitot tube or a static pressure port as used on aircraft? You could have the first Porsche since the PFM Mooney M20L to have an altimeter.