MFI Pump - Open Heart Surgery

[luckydynes]

Originally Posted by 911st

I love this picture.

For fun lets think of this as a fueling map.

The left side axis from top to bottom is RPM.

The bottom axis from left to right is TP (Throttle Position).

At idle the cell in the lower left cell determines the fuel. If it takes 10hp to maintain idle that is about 70#TQ.

Going to WOT the active cells skip quickly accrost the bottom of the chart to the 100% throtle section. The cells at the far right starting from the bottom and progressing to the top as RPM advances. Thus, if you have a representive or calculated WOT TQ curve you can map this section pretty easyly from 1000rpm to red line.

Steady state cruse would mostly be the cells progressing on the diagonal from the lower left hand idle rpm cell to the top right hand cell which would be the cars top speed.

Just off the cruse line will be acceleration cells. If the TP is advanced a small amount out of the steady state line, fuel might be increased say 10% to take us from the 14's to the low 13's to promote acceleration.

Get the WOT and the steady state sections right and the other stuff might fall in to line.

Yes that is very cool ... and at 2 distinct rpm points the engine likes to go a bit richer (the low spots). An engine wants more fuel for the air it's breathing as it reaches peak efficiency from cams/intake and then drops off as I understand it ... second bump means a second "rich" point for some reason ... could be something to do with cam/intake combo causing a couple of required "fatter" spots.

Be nice to see how well a pump tracks a/f ratio just during a full throttle pull ... I think I saw some graphs on here.

[911st]

Corrected.

Sorry for the quality but here is your basic RPM by Throttle Position map.

[luckydynes]

Originally Posted by 911st

It might simplify things to start by thinking of the space cam as three different space cams. One for low rpm (where TQ is increasing) one for mid range RPM, (where TQ is flat and or peaking) and one for high rpm (where TQ is falling).

We can then use the three stage spring sets to determine when each section of the space cam will be used and at what rpm.

Again, we do not care about HP, just TQ. TQ is the weak line on the following chart.

It this case the low rpm 100% throttle angle might be set for something like 125# TQ, the TQ peak section for 175 lbs, and the high rpm section for 150#.

There will be transitions between each slice of the space cam to sooth the transitions between them.

The space cam values or stylist heights will basically mirror the motors VE table at all RPM and throttle angle operating points.

This is all just a thought.

Looks like a couple of distinct richer bumps on this dyno graph on this post ... but you mention springs effecting the way the stylus moves up the cam ... there is a lot going on there if changing the springs is a known way of tuning. I'm still learning about all this.

edit: I got it backwards ... the "bumps" I'm referring to are in the direction of leaner.

cheers

[356RS]

Originally Posted by 911st

Sorry for the quality but here is your basic RPM by Throttle Position map.

Keith, You have it backwards. Idle is at the lower right corner and WOT would be at the top left.

[911st]

Mark,

You are right. It has been some 8 to 10 years sense I played with one of these.

Anyway the thought is still valid. Cruse is a diagonal, throttle lift is to one side of cruse, acceleration is to the other, and WOT is top to bottom on one end.

Get the WOT and cruse right and the rest should not be tough.

Here it is with a rough guess as to TQ at some of the points for a motor making 200 peak FT lbs TQ.

50# at 1000rpm is 10hp, 70# at 3000 is 40hp, 180# at 7000rpm is 241hp.

Corrected.

[911st]

Please be gentle, this is just to show that we do not have to be spot on at 120 data points or programing cells so to speak, on the space cam.

Get the WOT and cruse right and the rest should fall into place without to much issue is my thought.

A lot of cells like on throttle lift are kind of easy.

[911st]

Two 'maps' above corrected. Thx.

[jeffc280sl]

Be nice to see how well a pump tracks a/f ratio just during a full throttle pull ... I think I saw some graphs on here.

How well a pump tracks afr during wot is determined the positioning the stylus on the far left side of the space cam. AFR at wot (full load) is set by adjusting the rack screw. CCW is richer and CW is leaner.

Under partial loads the front to back stylus location is determined by the black and white screws. The upper rpm range is adjusted by the white screws which set tension on the heaviest spring inside the flyweights. CW adjustments (more spring preload) will create a richer condition and ccw a leaner while under partial load. The lower rpm partial load is set by the black screws.

I posted a wot graph on my car about 15 posts back.

[911st]

The spring tension screws are not really for adding fuel. They are for getting the stylist to the right place on the space cam.

In general, running a lower spring tension will let you progress to a richer part of the space cam earlier but up to a point.

However, before the high RPM end of the WOT part of the space cam and after the point where TQ peaks for said space cam, the AFR will actually get leaner.

For example, say if one is using a 2.4 T cam that makes peak TQ at 4000rpm, after that point the space cam would start to get leaner. If one has a build and cam that makes peak TQ at 5500rpm, one would 'probably' be well served to tighten the spring tension or cut down the weights to try to get the low part of the space cam to not come in until closer to 5500rpm than 4000rpm.

So in some cases, tightening will get you fuel, and in some cases loosening will get you fuel.

You would think a 2.4S and 2.7RS that share the same cams and basic injection would have space cams that look close to the same. I was surprised when I took a 'S' space cam out to put an RS space cam in a pump. The 2.7 had deeper cuts and the low section or valley covered significantly more of the space cam.

For example, if the 2.4S has to support say 50 to 158 ft lbs of TQ (a range of 108 from idle to peak TQ) and the 2.7RS has to support 50 to 188 lbs of TQ (a range of 138) the cuts or valleys in the RS are going to be about 22% deeper. Thus, the stylus needs to progress 22% deeper on a 2.7RS motor to support the range of operation between idle and the peak TQ point.

Yes you could use a 2.4S space cam in a 2.7RS build and position the rack to get the extra fuel need for proper AFR at TQ peak but then your idle fuel would go from something like 14.5/1 at idle AFR to about 11.5 AFR i this example (fuel for 88 v 50# of TQ).

Sorry if I get carried away, I love this stuff.

I am not an expert, just what I have learned to believe.

(I am guessing at the idle TQ numbers so my delta could be off but the basic conclusion would be the same.)

[Flieger]

The peaks corresponding to drops in VE are likely to do with the exhaust pulse resonance/scavenging. At certain valley points, the scavenging and intake "supercharging" work to increase VE to the point that other spots become lean peaks. This is why MFI is more work to tune for exhaust system changes.

The peaks and valleys are also narrowed because of the path of the stylus. One must take the gradient of the two-variable function (throttle angle and speed height) in cylindrical coordinates to find the instantaneous rate of change/ curve for the fuel metering height.

How much of a difference is there between the pointy stylus cams and the later roller cams, other than length?

[911st]

Agreed!

There is a part in the rpm range of a 911 motor where the exhaust seems to back up and then become extra efficient that creates a bump in the VE table. This might not be as big as we think if we stay with a stock or factory sport muffler.

Aftermarket mufflers can seem to exaggerated this issue.

As MFI is a pure 'reference" injection system with no air flow metering or allowance for changes in air flow expectation, any changes in air flow can through AFR's off quickly.

Thus, the same reason we might want to go to a new exhaust is the same thing that can through things out of whack.

I to would like to know more about the pre 2.4 pumps. I understand most pump builders prefer them for highly modified motors like working with an RSR cam or such. They seem to have a simpler space cam and I think the stylist is roomered to be more stout and less pron to issue.

[jeffc280sl]

911st

We agree that the spring tension screws are for getting the stylus to the right place on the space cam. What the stylus reads determins afr by adding or reducing the fuel supply via the rack. If the stylus is reading the deep part of a groove as rich you could go potentially go leaner by turning the white or black screws in either direction up to a point. Don't forget there is some overlap of the white and black screw adjustments. What I mean is that changing the white screws will effect the lower partial load afrs to some extent.

[Flieger]

Originally Posted by jeffc280sl

The MFI is not efficient that's for sure. With a hand held O2 meter you can watch afr bounce all over the place while driving. One thing I know Bosch was concerned with is afr while coasting. When the throttle is on its stop and the engine is still producing rpms during a downhill coast you can watch the afr meter go lean. Vacuum timing adjustments are also crude by todays systems.

I forgot to add my thoughts on this.

When the engine is above 1500rpm and the throttle is on the stop, there is no fuel going to the cylinders because the shut-off solenoid/microswitch/RPM transducer should be functioning. That would be the reason for the lean AFR- an intentional measure to save fuel and prevent pooling fuel on the butterflies causing a surge/backfire/poor running when the throttle is opened again.

[jeffc280sl]

I have removed the fuel shut off solenoid from my pump and can report the space cam by itself does lean the mixture when coasting. I also think when activated the solenoid pushes the rack to the idle position versus shutting off the fuel supply. Minor point

[356RS]

Originally Posted by jeffc280sl

I also think when activated the solenoid pushes the rack to the idle position versus shutting off the fuel supply. Minor point

Actually, the solenoid ends up lifting the stylus off the space cam and shuts the fuel off.
As your RPM drops towards idle, the solenoid is then deactivated and fuel flows again.

[911st]

I love the way MFI crackles and pops on throttle lift so if I build another MFI motor I would take that solenoid off to.

I removed my Thermostat to and in its place put a thumb screw that I could use to quickly adjust the base AFR quickly.

I did add a button on the dash to enable the factory primers and never missed the thermostat. However, I do live in California. It seemed my best start up was not to use the primers and to just push the throttle to the floor and energize the starter. Then the second it started to kick, I lifted my foot and set my had throttle.

One of my thoughts it to build a small port SC motor opened up to 35-36mm intakes with just exhaust and MFI keeping the stock cams. It would also be striped of all the pump add ones except maybe the altitude pot. That might go to if I keep my external adjuster and run a full time wide band. Programmable MSD would be a nice addition to this. With the rich/lean conditions that even a well developed MFI can run, MSD and or twin plugs are a good addition.

[911st]

Mark,

Again, cool info.

Is there an easy way to easily change the leverage ration between the stylist and rack?

This might allow us to use a stock space cam to work over a wider TQ range on a built motor.

Larger plungers could do the same thing. 25% larger plungers would take us from a 2.4 to a 3.0 motor and the stock space cams might be more adaptable.

Also, have you done any tests to see what the max delivery of a street MFI pump might be? That would really be interesting to know. I think the Turbo MFI pumps had larger plungers to support 500+ HP but am not sure.

[kenikh]

Originally Posted by 911st

Larger plungers could do the same thing. 25% larger plungers would take us from a 2.4 to a 3.0 motor and the stock space cams might be more adaptable.

Intriguing - how much meat is in the plunger bores? This would be a "simple" bore and ream operation for the larger plungers. The harder part is making the plungers, which would have to be machined, then precision ground, which is very specialized equipment; not something the home machinist has around.

Mark, being in the PNW and relatively close to Boeing, there are numerous shops up here that do precision contract work and could certainly do a run of plungers, if one were so inclined.

Intriguing indeed, although I will play devil's advocate and suggest that we are really moving into the academic versus the practical. Jeff Higgins car runs basically perfect AFRs now that he has dialed in the tuning. Not bad for a high revving, extremely torquey, twin plug 3 liter with hot cams on a 2.4T pump, no? I am not saying it is PERFECT, but certainly has the potential to be if he swapped for an 'S' pump.

[356RS]

[QUOTE=911st;5267681]

Is there an easy way to easily change the leverage ration between the stylist and rack?

Keith, Maybe you can. Below is the Stylus to Rack lever system. The lever in the center could be modified in length to change the ratio. Also the lever's pivot point is an off-center adjustment just for that, although the ratio change is very small.

[356RS]

Originally Posted by kenikh

Mark, being in the PNW and relatively close to Boeing, there are numerous shops up here that do precision contract work and could certainly do a run of plungers, if one were so inclined.

I have found several manufactures in China that already make the plungers and cylinders that are very close to the PED 6 KL specifications, but they only make the 1-2-3 side plungers, for all the in-line pumps. The 4-5-6 side plunger faces 45 degrees off.

I have installed 2.4 plungers/cylinders in an early 2 liter pump and can get close to 70ml per 1000 strokes at WOT @ 3500 pump RPM.