MFI Delivery Rates

[jluetjen]

It actually sounds like the Kuglefischer MFI system that BMW used for their 1.5 liter F1 Turbo engine. That too was an "electronically controlled" MFI system.

[KobaltBlau]

This is a way cool idea, guys. Be sure to post here as you work on these projects.

Originally posted by michel richard
MFI becomes available for a wide variety of engines.

But for bigger engines you still have to get custom stacks ($$$)

[KobaltBlau]

Originally posted by jluetjen
It actually sounds like the Kuglefischer MFI system that BMW used for their 1.5 liter F1 Turbo engine. That too was an "electronically controlled" MFI system.

do you know more about this, John?

[michel richard]

One thing that would be useful to make progress would be an old MFI pump that I could take apart and use as a mule for things that will need to be fabricated.

If anyone knows of one for sale, please let me know. I'll pay a reasonnable amount for an old pump, does'nt need to be an actual working one.

I'll also post on the classifieds.

Michel Richard
michrich@videotron.ca

[Henry Schmidt]

Originally posted by KobaltBlau




But for bigger engines you still have to get custom stacks ($$$)

Available from Supertec Performnace from $1500.
41 mm butterly, stacks, linkage and custom fuel lines.

[NevenM]

Henry

What has been done to the pump? It looks seriously modded

Neven

[KobaltBlau]

Originally posted by Henry Schmidt
Available from Supertec Performnace from $1500.
41 mm butterly, stacks, linkage and custom fuel lines.

Very nice, Henry. Not that expensive at all, actually.

[power]

Have you seen this before?

http://www.pelicanparts.com/techarticles/911_MFI/Bosch%20_More_MFI_Repair_Instructions.pdf

[NevenM]

Yes but I'm not sure what 'adjustment distance' means and if it is the rack position it seems stange that it appears to not be directly related to delivery volume ie Under part load
400rpm, 4.4mm = 12.5-13.5 and
3000rpm, 3.8mm = 12.5-13.5

So it seems to be a measurement before the space cam, what I want is measurement after the cam, ie the rack movement

Neven

[thastings]

Henry,
-Does that include the MFI pump?
-Will the stock '72 air cleaner fit?
-Is this a good "street" setup?
- What other changes will this require to maintain the "stock" 911T driveabilityand performance?
-Stock fuel pump?
-Will the idle-speed adjustment, etc still be required?

Terry Hastings

[KobaltBlau]

Terry, maybe you know this but those stacks are for a bigger motor. I believe that the motor pictured is a 3.5 race motor. I also don't think that the MFI pump is included; this is just the induction. For 2.7 or less you can use OEM parts, depending on your RPM (high RPM motors need more air, of course).

[jluetjen]

do you know more about this (The BMW F1 Turbo injection system), John?

This is per "BMW - Bavaria's Driving Machines" by the Auto Editors of Consumer Guild and Jan P. Norbye. As far as "coffee table books" are concerned, it's really pretty good with excellent text and pictures. They also put out a similar book on Porsches which is just as good. If you ever run across them at a used book sale, they are definitely worth the money. Anyhow, this is from an interview that the writer had with Horst Rech, the BMW Motorsports engineer who supervised the development of the F1 engine.

The cylinder head from the F2 engine needed little modification for its F1 application. An aluminum casting, it has four valves per cylinder splayed at an ange of only 20 degrees from the cylinder axis and closed by duel-coil valve springs. Intake valve head diameter is 35.8 mm and the exhaust valve heads measure 30.2 mm. Valve timing has not been disclosed, but Rech admits there is considerable overlap. Because of this, a time, high pressure fuel injection system is used to prevent unburned mixture from being ejected into the exhaust system, thus protecting the turbine from overheating. The Kugelfischer pump operates at an injection pressure of more then 440 psi, and provides the maximum fuel quantity required throughout a 100 degree crankshaft angle range. The usual cone-cam arrangment for fuel metering in the normal Kugelfischer system is replaced here by a spiral that is moved by an electro-servo motor into the position calculated by the engine's electronic control unit. The spiral can run from one end of its travel to the other in only .04 second.

The spiral idea is pretty clever since it provides a more sensitive adjustment of the rack then a stepper motor may be able to accomodate.

Here's a picture that I found on the web, but unfortunately it seems that the pump was mounted under the intakes and so is almost never pictured.

[KobaltBlau]

wow John, that is very cool. Thank you for putting that post together. It seems that even modern EFI could not deliver the fuel in such a short time, perhaps only the new common rail diesel injection is in this ball-park for a fully electronic system (just a guess, but it runs at much higher pressures than gas EFI).

[michel richard]

I'm not sure I understand what is meant by a "spiral" in this case. I read that the servo motor acts on this "spiral" . . . is it really like the rack that is found in our Bosch pumps ?

It looks more and more like Neven's idea of a servo motor acting directly on the main rack is the way to go, as opposed to a motor (servo or stepper) acting on the little lever that the normal MFI thermostat acts on.

The more I think about it, the more Neven's idea sounds better. The force to move the rack is quite small, I believe: the repair book says to use a "biro" to move the rack, using only very little pressure to see if it's stuck. I suppose the force required could be greater when the engine is running, but there is no indication anywhere that that is so. Certainly, the force required at the gas pedal does not appear to be any different between a carbed car and an MFI car, and the force required to move the "fuel" lever on the pump when the engine is running is small.

Can't wait to lay my hands on a "play" pump to do some real figuring. None on ebay right now, though.

Michel Richard

[jluetjen]

I'm not sure I understand what is meant by a "spiral" in this case. I read that the servo motor acts on this "spiral" . . . is it really like the rack that is found in our Bosch pumps ?

Here's how I understand it. The pump itself including the rack is conceptually similar to the Bosche pump. BTW, some sources describe the BMW Turbo F1's pump as being Kugelfischer pump manufactored by Bosche. There may be some spin involved here since I generally see Bosche's name next to Porsche and Kugelfischer's next to BMW.

Anyhow, here's an explosion of the key MFI components:


Note how the back and forth movement of the rack (#9) rotates the plunger assembly (#7 via #8). The way that this rotational movement varies the fuel volume pumped by the plunger assembly is as follows:

Note that the fuel flows in from the left and out the top. Notice how the spiral groove in the plunger can vary the actual closing event based how the plunger is indexed. Examples #1 through #3 shows the plunger alignment for starting (full rich) when the groove never sees the port on the left as the plunger moves up and down. As a result a full stroke of fuel is injected out of the injector.

Example number #4 through #7 show the plunger alignment under normal operation. While the plunger stroke is the same, the groove on the side of the plunger can delay the actual closing event by upwards to half of a stroke, kind up like a long duration cam can reduce the CR of an engine. The result is less fuel is pumped out of the nozzle.

Now going back to the BMW example imagine that you pitch everything in the first picture before the control rack head (#2) and replace it with a stepper motor and computer. What would happen if the minimum adjustment step of the motor was too large compared to the required adjustment on the rack? You'd most likely find that will flutter back and forth between the two closest steps.

Instead, picture a threaded screw pushing against the control rack head and the motor turns the screw. This would allow virtually limitless adjustment of the rack based on the pitch of the screw and the rotational speed of the motor. I suspect that this is the type of arrangement which the BMW engineer was describing.

[NevenM]

John

Firstly they used a 'servo' and not a 'stepper', Servos are analoque, so wouldn't have a discrete increment. The spiral would be how they converted the rotational position of servo to the linear position of the rack, AFAIK Bosch own Kugelfischer.

The Futaba Servos have a response time of 0.2s for a 60 degree change, I think this will be responsive enough

Neven