MFI Pump - Open Heart Surgery

[KevinG]

From John Cramer - "you can get WOT AFR graphs against the RPM and log the whole thing to see the differnce"

John - Can you do this with just the basic LM-1, or do you need additional accessories?

Thanks,

Kevin

[Peterfrans]

You will also need the LMA-2 (RPM Converter.) This cable contains a circuit to convert a tach signal into an analog RPM signal that can be logged by the LM-1. This cable also has 4 additional terminal screws to tap into and log other 0-5 V analog sensors from your vehicle.

[304065]

The best thing to do is use the AuxBox, it includes an accelerometer. Being able to log not only AFR but the rate of acceleration of the vehicle under load is about as good as it gets.

[jcge]

Just to clarify - If you're going to adjust the black (or white) screws, make sure that both screws are adjusted equally, otherwise the capsule/spring mechanism may bind, preventing (or even halting, or worse) a smooth progression of the space cam with rpm.

If you move +2 clicks with the first black screw, you must also move +2 clicks with the second black screw.

The same applies to the white screws - they must be adjusted as a pair.

Caution is required....I'm not sure of the limits of how many CCW clicks they can turn before the unit dissassembles....(but that would be bad....very, very bad in a running pump)

John

[David E. Clark]

Originally Posted by jcge

... Caution is required....I'm not sure of the limits of how many CCW clicks they can turn before the unit dissassembles....(but that would be bad....very, very bad in a running pump)
John

There is some help in the "ADJUSTING THE FUEL FLOW AT PART THROTTLE" section of the Glenn text posted above:

CAUTION: The adjusting screw should not be turned over 2 notches from its standard position.

These screws must be turned in pairs, white only or black only, but each adjustment affects the other and this is why these screws should be adjusted only with accurate equipment.

[Flieger]

Sorry to hijack the discussion, but I read in a parallel MFI thread that poping into the intake/airbox is indicative of a lean mixture. Backfiring through the exhaust was said to be a sign of running rich. I experience both at the moment. I noticed only the backfiring through exhaust when I ran the stock plenum. Now I have some homemade intake screens and I hear the lean ticking noise.

I do not know if the ticking is from the change in intake tract or if it was there before, just muffled by the plenum. The warm-up thermostat has been replaced with a screw adjuster and I recently turned it in 4 turns(4x360 degrees).

Am I running too lean? I still get backfires but there is less carbon on my Burzel after fast runs, so I thought I had improved the mixture.

Thanks

Here is a photo of the engine:

[jcge]

David - 2 turns limit is relative to the factory set starting point of the screws - if they have been previously adjusted (perhaps by a previous owner, mechanic, tuner etc...) you may get more than you bargained for and be outside the +/-2 turns without knowing it - just a thought.

And particulalry in Jeff's case, where he's already using a pump on a 3.0L motor...

Flieger - turning the thermostat mod screw in (CW) would definitely lean the mixture - but I can't offer you any other suggestions as to the backfiring / popping...AF measurement (ala LM-1 etc) might help steer you in the right direction. Suggest you consult the person who modified the pump.

[Flieger]

I lubed my throttle linkage with WD-40 and examined the microswitch as well. It appeared that the throttle lever would not always have enough force to make the switch click unless the throttle was released rather abruptly. I think this could be causing some of the popping/backfiring.

How much difference does the off-throttle fuel cutoff microswitch make to popping/backfiring? Is there a specific adjustment or test procedure?

Thanks.

[Jeff Higgins]

Originally Posted by jcge

Just to clarify - If you're going to adjust the black (or white) screws, make sure that both screws are adjusted equally, otherwise the capsule/spring mechanism may bind, preventing (or even halting, or worse) a smooth progression of the space cam with rpm.

If you move +2 clicks with the first black screw, you must also move +2 clicks with the second black screw.

The same applies to the white screws - they must be adjusted as a pair.

Caution is required....I'm not sure of the limits of how many CCW clicks they can turn before the unit dissassembles....(but that would be bad....very, very bad in a running pump)

John

All three pumps I've played with have the two black and two white screws screwed very close to all the way in. They are only three turns from seated on the average. They can screw out nine turns before the screw heads are flush with the aft edges of the flat springs that provide the "detent" action. At 12 turns they clear the springs and no longer "click" on them. At 15 turns from seated they fall out. Even if they did, they will be captured under that little cover and cannot go anywhere. They would probably scratch that up, and ruin the end of the capsule they screw into, but there is no way for them to fall into the governor assembly.

I think a safe limit would be flush with the springs. That allows nine turns of adjustment, or 54 clicks. That is a significant range, and may be enough to very radically alter the pump's application. It sure seems to be...

So, I have to admit I've gotten impatient waiting for my LM-1. So I got brave and hit the road with my tools and a note pad. The plan was to turn the black screws only, assuming they were the "mid range" as called out in the manual one of you posted. I started out driving it at "nominal", which is the configuration it was in on the dyno. I recorded my impressions of how it behaved at various loads and rpm's.

Then I started turning screws. I can have the side cover off, the oil sucked out, the adjustment cover off, adjustment made, oil back in and buttoned up in less than five minutes on the side of the road or in my garage. So, I backed the black screws out 5 clicks and took note of the changes. I eventually wound up with them 15 clicks from "nominal". The car ran much, much better up to 2500-3000 rpm, with none of the over rich blubbering it was suffering before. It was still blubbering between 2500/3000-ish to 4000/4500-ish, and 20 clicks made no further improvements, so I returned to 15 and moved on to the white screws.

I followed the same approach with the white. Back out five clicks; drive and jot down impressions. I wound up with the white screws 10 clicks out; I'm going to stop there for now, until the LM-1 arrives. This adjustment very significanlty cleaned up the range up to 4500, to where it now just barely, almost imperceptably, has a bit of that rich blubber left. Just at part throttle "cruise", but never at full throttle. I'm sure it's still on the rich side, and I don't want to go any further without test equipment, but I can tell you it is a whole different car.

Studying my A/F chart from the dyno, it appears the mix was headed towards lean as it was approaching 6000 rpm. In light of this, I backed the main rack out two clicks in an effort to be on the "safe" side. Next weekend I'll have the LM-1 in hand, and I'll be able to further refine the mix. I've gained some measure of confidence in which range each screw or set of screws affects the mix. It will become much clearer with the LM-1.

The car runs clean, and has that very sharp MFI throttle response that we all know and love. It would be "drivable" to me just as is back in the old days, (when we couldn't easily get real numbers on this stuff), even knowing it's still a little fat and I'm leaving a bit of power on the table (and gasoline on the road behind me...).

I think I'm far enough along to conclude that yes, Virginia, a 2.4 "T" pump can run a hot rod 3.0 liter and run it well. The LM-1 will help me make those last little incremental changes. Or maybe it will just verify that it will never be "optimal" everywhere, at least by today's standards. But I will get a wee bit closer with it and, like I said, it is eminantly drivable right now. And my God does this thing haul some ass...

[ischmitz]

wow, interesting read. I like it when someone tackles the "impossible" and de-mystifies it. subscribed.

[Vern Lyle]

Impromptu diagnostics

['71 911E Coupe]

Magnificent thread! I've read the whole thread and have learned so much!

I can't wait to see what your final results are after you get the LM-1, Jeff.

I'm sure your results will help many of us with MFI problem (including myself), and what I am getting from this, is that adjusting these other screws make a wild of a difference and get these cars running correctly.
These manuals that we all have (CMA, etc.) only explain the two adjustments, but these other ones are very important too, correct?

Ya think us guys with MFI problems may have to look into adjusting these additional screws too??????

Thanks!!! And keep up the good work!

[jcge]

Jeff - how's progress with your MFI part load tuning via dyno and/or LM-1 ?

Regards,

John

[Jeff Higgins]

Hello everyone, I'm back. Sorry about the delay; my LM1 took a bit longer to arrive than I thought, and when it did come, I was away on an overseas assignment for a few weeks. Anyway, I've had some time to play with it, so here is what I found.

First off, I'll speak in generalities as much as specific numbers. I don't want to lead anyone astray by emphasizing numbers specific to my oddball application. Suffice to say if you want to do this on your own car, you need to get an LM1 and have at it. My specific numbers would probably be useless to you, so just remember that.

My car started out excessively rich in the mid range, and progressing towards too lean on the top end (under full throttle). With the motor now fully broken in, the LM1 initially confirmed this. It was running at 10:1 or richer between 3,000 and 4,000 rpm, and progressed up to 17:1 at between 5,000 and 6,000 rpm. Pretty much just like on the dyno a couple of months ago. So, the first thing to do was to fatten up the top end into a safe range.

This was done with the usual main rack adjustment, without touching the single idle or the two black and two white screws on the end of the governor assembly. The center idle screw was already as lean as it would go anyway. Starting with the main rack as it was when the pump was rebuilt (tuned for a 2.4 "T" motor), it wound up taking 25 clicks to richen it up to where it ran at 13:1 at 7,000 rpm. That makes sense; it's now on a 3.0 liter with two points more compression and over .100" more lift on the cams; obviously a much thirstier motor. It now stayed between 11.5:1 and 13:1 from 5,000 to 7,000. The low end and mid range, of course, got even richer. All the way to 9:1. It was obviously time to go after the black and white pairs of screws behind the governor cover.

I wound up with both pairs of screws backed out as far as I dare. Like the idle screw, they are now flush with the top of the flat springs that provide the detent action. At this limit of adjustment, then, I was able to affect the low and mid range at full throttle, but to a lesser degree than I had hoped at part throttle. It appears as though these screws work at about a 3:1 dissadvantage to the main rack screw. In other words, it takes three clicks lean on them to offset one click rich on the main rack, just to keep their respective adjustment ranges where they were before the main rack adjustment. The white ones do in fact affect the upper mid range, say to 4,000-4,500 rpm, while the black ones affect up to 2,500-3,000 or so. There is a lot of overlap, so it's a bit hard to say for sure. I can say, however, that I think I've confirmed that nothing on the governor does anything at much over 4,000 rpm (engine speed).

Full throttle now stays between about 11:1 and 13:1 across the rev range from 2,000 to 7,000 rpm. I could live with that if it were a race car or track only DE toy. Part throttle cruise and idle, however, are both too rich. Part throttle now runs below 10:1 up to 3,000 rpm, but the slightest load raises it very quickly to 11:1. Idle is about 9.5:1, with no adjustment left. I have to say this relationship between part throttle and full throttle has me a bit confused; I would have assumed part throttle would be leaner, rather than richer than full throttle at a given rpm. Quite the opposite is true.

I'm starting to form some conclusions from all of this. It is apparent that the pairs of governor screws work in a far narrower range of adjustment than the main rack. The space cam has to determine the center of that range while the governor screws can only fine tune from there. They do not appear to be suited to drastically altering the application of the pump.

My thoughts are that just about any displacement or configuration of flat six uses about the same fuel at part throttle cruise, within a narrow range of variation. The bigger, more powerful, or higher reving it is, the greater its peak demand at full throttle/high rpm, with this end of the range showing several times the variation as part throttle cruise. It then follows that the greater the peak demand, the more "aggressive" the space cam must be to reduce fuel delivery at part throttle. So a "T" motor, with the least peak power and fuel demand, would require the least affect from the space cam.

In summary, I have had no trouble adjusting the "T" pump to deliver plenty of fuel for full throttle operation. There is simply not enough adjustment in the governor screws to cut that delivery enough for acceptable part throttle and idle. I do think, however, that any space cam more aggressive than a "T" might just get that range re-centered enough to take up the difference with the governor screws. I'm pretty darn close as it is; like I said, if it were a race car, I would be done.

[Drago]

Originally Posted by Jeff Higgins

...like I said, if it were a race car, I would be done.

Good stuff Jeff, but you forgot to mention the power increase you achieved by tuning with the LM-1. I know it's not been on a dyno yet but what does the seat of the pants tell you?

I can't wait for another ride.

[Jim727]

Jeff, this is great stuff. I was wondering what happened here.

You made a comment that "I can have the side cover off, the oil sucked out, the adjustment cover off, adjustment made, oil back in and buttoned up in less than five minutes on the side of the road or in my garage", which raises a question for a guy new to MFI. My understanding is that the MFI pump is fed oil by the engine oiling system. This makes it sound like there is another encapsulated oil not fed by the engine oil system - true or not? If so, what type of oil is it, what are you doing to drain/refill? It is one of those things that sounds like it is either trivial or a minefield.

My understanding (remembering that I'm an MFI neophyte) is that the space cam controls the timing of the injection pulse, but the flyweight system controls the quantity in lower RPM ranges. If that's correct (is it?), it would seem that Gus could change out your governor weights to get you into an adjustable range.

Terrific stuff - please keep the news flowing.
Jim

[Jeff Higgins]

Jim, you are right that stock MFI pumps run on the engine oil supply. A common modification is to isolate it from that recirculating supply. This is done by removing the feed and return lines, then providing a pressure vent. The holes for the lines are plugged with a Solex float bowl screw on the feed side, and a VW drain plug on the return. The engine breather cover is then replaced with a standard carburetted or CIS type cover (the one without the return line) and the oil pressure sender base (where the feed originates) is likewise replaced with an non-MFI unit. The pump is filled with Mobil 1 until it starts to overflow from the return line hole (just like topping off a transmission). It takes about half a quart.

I use a gear oil pump to suck the oil out of the MFI pump so I don't have to remove the pump from the car. Like this:



You will note that my pump is missing the warm up and decell solenoids as well. There is also no thermal/time switch on the breather cover; I wired the bypass solenoid on the filter console to a pushbutton on my dash for priming.

Here is the goofy little screwdriver I use on the governor cover to access the two pairs of screws:



Here is the governor cover swung out of the way to access the screws:

[Jeff Higgins]

Oops - I forgot the other part of your question, Jim. There are two distinct and seperate sections to an MFI pump; the pump and the governor sections. The forward "pump" section contains the plungers and a cam (that looks like a normal cam) to drive them up and down. This section is identical on every pump; so is the cam it contains. It is this cam that controls timing of the injection pulse. That pulse is delivered at the same time regardless of the pump's application; a "T" pump shoots fuel at the same time as an "RS" pump. They can both shoot the same maximum volume as well. Again, there is no difference in this section of an MFI pump between pump applications and designations.

It is the governor assembly that bolts to the back of the pump section that varies between pump application. Inside of that governor assembly, everything is once again identical for the different applications, except for the space cam. The flyweights and screws are the same in all of them. Space "cam" is actually a bit of a misnomer; it does not spin as one would expect a cam to spin. It looks nothing like a cam, either. More like a small metal hockey puck. It rotates about its axis for about 90 degrees or so when pulled up and down by an arm connected to the throttle lever. It moves fore and aft on its axis in response to the opening of the clamshell weights. It does not affect timing in the least; just volume.

[Jim727]

Jeff, thanks for the additional info. Didn't know that the flywheel weights were constant across all applications - it seemed that they would be the simplest way to control volume against rpm, but now it makes sense: make everything possible a constant to leave only one variable - the space cam. Not a simple variable, but....

Going to change the Mobil 1 out for Brad Penn?? (Couldn't resist)

My "goofy" ratcheting screwdriver went walkabout. Still not amused.

Superb stuff. Thanks again.
Jim

[304065]

I have to wonder whether the slope of the fuel curve is affected by removal of the thermostat. Have you considered installing a thermostat replacement cover with a pin that extends the same distance as the pin in the original thermostat does when hot? That acts on the governonr linkage and may take some of the richness out at both ends.

I realize best power is 12.8 but could you shoot for closer to stochiometric at 7,000 rpm (14.7) to bring the idle end up? Is that 7000 rpm under full throttle and full load (I assume it is, although to really get it loaded you would need to be at the maximum vehicle velocity in fifth, not easy to do)