MFI Pump - Open Heart Surgery - Page 2

pjh69911 · 01-10-2008, 06:24 PM

If its any help, I once measured the spring rates of the three sprigs on my 69E pump. The smallest(idle) spring was 45 lb/in, the intermediate was 184 lb/in, and the strongest spring was 900 lb/in. Initially, at low rpm, the only spring that is active is the smallest spring. At some above idle prm, the weights engage the idle and intermediate springs. At an even greater rpm, as the weights move further out, all three springs are engaged. I assumed that the sizable difference in spring rates results in moderate independence of their actions. For example, the idle spring is only 1/20th of the full load spring. The intermediate spring is about 20% of the full load. Hence at full load, the idle adjust does nothing, the internediate spring does little, but the full load spring influence is predominant. At intermediate speeds, the idle does little, the full load spring does nothing as it is not engaged yet by the counterweights, and the intermediate spring here is most important. I just wish I had a bench setup to better know the approximate rpm of the transitions. I have heard, that the idle spring only really works at idle. It also seemed that the spring rates corresponded to a squared function, as would be expected in the centrifugal force from a rotating counterweight.

Jeff Higgins · 01-10-2008, 08:07 PM

Well, I didn't get around to tearing into that pump tonight. It's my younger son's birthday, and as much as I tried to convince him that would be "quality time with dad" he wasn't buying it.

Anyway, starting to piece all of this together from the Mercedes pump diagram, the photos you guys have posted, and some of your observations.

I think the white screws affect preload on the heavier (900 inch pounds?? wow...) spring and the black screws affect preload on the intermediate spring. The center screw obviously adjusts preload on the small center spring. I'll have time to tear into that little piece that fits over the springs tomorrow night. I'm curious what lies inside that would make all of that work.

Henry Schmidt has mentioned several times that the space cam does nothing over 4,000 rpm. The clamshells are full deflection at that speed. That is engine rpm, so actual pump rpm is only 2,000. Bernie told me the same thing. That must, however, be plus or minus whatever range the adjustment of those white screws can vary it. Clearly, if my suspicions are correct and those screws affect preload on that heavy spring, that will affect the rate at which the clamshells can open and, ultimately, the rpm at which they open fully.

jcge · 02:00 PM

pjh69911 - I think you are correct - the springs in combination deliver a piecewise progression to the space cam with increasing engine speed (roughly proportional to the square of shaft speed).

Jeff - no disrespect to Henry - but I just can't get past the nonlinearity beyond 2000 pump rpm on the 911/72 chart on the previous post - what else could cause this.. (Note that I think 911/72 is not a street car motor = 2.8 RSR)

Jeff Higgins · 01-11-2008, 06:12 AM

John, I'm having trouble reading the numbers on the chart, or even what the different axis represent. Is that fuel volume at various throttle positions vs. rpm?

304065 · 08:09 AM

I sent Bosch an email asking for the document, they have replied with "working on it."

Edit: we already HAVE This document, it's the "MFI repair instructions" booklet available here. But we have a VERY poor scan, much worse than the good scans of Check Measure Adjust.

I'll keep you posted.

Jeff Higgins · 01-11-2008, 09:03 AM

Actually, John, the manual we have is the "WJP"; the missing one is the "WPP". On page 28 of the WJP, step 74, it has us assembling the governer springs and the spring retainer, the little cover with the five (two black, two white, and center) screws in it. Then it refers to the WPP for testing and adjustment. So, we still need that VDT-WPP 711/1 B manual.

I suspect we would also need the Bosch MFI pump dyno to start messing with the black and white screws. Being relatively sure I'll never stumble into one of those at a garage sale or anything, I'm going to get bold and head back to the chassis dyno and start turning screws.

The cover that holds the outer, spring loaded hex head that we depress to engage the center low speed screw has a seal inside of it. The seal appears to seal the the outer diameter of the spring retainer up against the governer case, rather than that little cover. In other words, I believe I can remove that small cover without all the oil inside draining out. That means it can be an on-the-dyno adjustment, albeit not quite as convenient as the other two screws.

Like I said, I'll dig into that retainer tonight and confirm or dispell my suspicions about what goes on inside of it. If I'm right, I can probably use those two white "high speed" screws (per the Mercedes diagram) to lean out my 3,000-4,000 rpm rich condition. Or, it may take an adjustment on the black "low speed" screws. Or a combination thereof. It will be awhile before I get back to the dyno to do this, but I will certainly report back when I do.

David E. Clark · 01-11-2008, 09:07 AM

Quote:

Originally Posted by john_cramer

... Edit: we already HAVE This document, it's the "MFI repair instructions" booklet available here. But we have a VERY poor scan, much worse than the good scans of Check Measure Adjust ...

John Cramer, I think the document John Jenkins is referring to is the one called "Bosch MFI Pump Test Instructions" listed as "(soon)" in John Hunt's Bosch Mechanical Fuel Injection page. I believe that the poorly scanned Bosch MFI repair instructions and the supplement that we all have from John Hunt's page are different documents. The "test Instructions" have been listed as coming soon for as long as I can remember!

I have a three page document dated 23 August 1972 entitled: "Bosch Injection Nozzle Testing Equipment and Quick-Clamping Device" which details the use of the Bosch Injection Nozzle Tester along with a short letter from Geoffrey F. Clarke, the Service Operations Manager for the Porsche/Audi dealership in Grand Rapids Michigan noting the price of the tool ($142.00) and recommending it's use. I'm also certain that this is not the document that John Jenkins is referring to. The letter I have is an original so I can make a good scan but the Nozzle Test Instructions are a very poor photo-copy. I will be happy to send this to anyone who would like to see it with the caveat that the scan of the photo-copy is bad!

I believe that the "Storage Instructions," also listed as "(soon)" are the ones available on pages 52-56 of CMA version 4532.21 called "Storage of Fuel Injection Pumps - Porsche 911 T, E and S". I have the original of this document if anyone needs a copy.

304065 · 01-11-2008, 11:45 AM

Quite right gentlemen, I misread the middle letter.

Will let you know what the Bosch folks say.

pjh69911 · 01-11-2008, 01:33 PM

Jeff,
You are correct in that you can access the white and black screws by removing the small access cover. The oil level, however, is a problem and quite a bit will come out. I tried, though not successful, to park my car on a very steep driveway to prevent this. You might just seal off the oil feed and keep the oil level lower, I think that could work. Good luck. paul.

304065 · 02:43 PM

I see the little plate you are referring to. I think you would lose the oil in the process. In any event, the oil feed is a pressure feed uphill through the banjo fitting at the oil pressure sender, so without the engine running, the pump is not going to refill. You could slip off the drain line to the breather cover and insert something to vacuum the oil out, then refill after the adjustment was done. If you were lucky, you would only use two rolls of brawny on the top of the engine. . . .

Sealing off the pressure feed permanently is not a good idea insofar as the pump will run out of oil. Gus told me that in order to seal the pressure feed and the return, and run the pump on its own independent oil supply, you have to have the red RSR-style breather installed (that's why all the race pumps have it). Otherwise all the oil will blow out through the seals. . . He also said you should change the pump oil every time you change the engine oil, which in my case is after every race. . . and you have to remove the pump to do that properly.

Having fought the pump for hours to get it back on, the next time it comes off my car is when the slide valve MOTEC goes on!

Jeff Higgins · 01-11-2008, 03:36 PM

Here is a photo of my breather setup, John. My pump is sealed off from engine oil and is full of Mobile 1, with the return capped by a VW drain plug and the feed capped by a Solex float bowl screw. The breather was made by drilling and tapping the little cover forward the warm up thermostat and screwing in the unused return fitting. On top is a pipe elbow, a lenght of hose, and a motorctcle crankcase breather. Works like a charm; clears the stock air cleaner, and the pump absolutley does not pressurize.

Jeff Higgins · 01-11-2008, 03:43 PM

I dug into the governor spring mechanism today. It looks like it must be identical in function to what the Mercedes diagram shows on the first page. The three springs are different weights and lengths, with the longest and lightest being the idle, and the shortes and heaviest the high speed. Here all all three in the housing:

Here is just the idle spring in the housing:

Mid range:

High speed:

Jeff Higgins · 04:19 PM

As you can see, the three springs sit on different parts of the housing. Those three parts move independantly. Here is just the housing, showing the three nesting surfaces of the respective springs:

Obviously, the center post is for the idle spring. The next ring out is for the high speed spring. That ring's position fore and aft is, in fact, affected by the white screws. The outer ring is the mid range spring seat, and it is fore and aft position is affected by the black screws. I didn't want to dig any deeper into this piece, as there is a snap ring that I simply cannot get off easily holding the works together. I can, however, pull it apart far enough to see inside. I also played with screwing the white and black screws in and out. Sure enough, those rings move in and out. Those screws actually "click" just like the idle screw, utilizing the same flat springs that bear up against their hex heads. Here is a photo:

So that heavy high speed spring has me a bit baffled. It appears as though it may be all of that 900 in/lb rate; it's like a valve spring. Pretty darn hard to compress even with a C clamp, where the mid range spring can be compressed a bit by hand. I'm wondering if the flywheights are even heavy enough, and if they spin fast enough (half engine speed) to even compress this spring. Maybe it's just a buffered stop, limiting travel of the space cam without it hitting a hard stop. Just guessing here; "thinking out loud"...

It is clear, though, that the further the flyweights are allowed to open, the leaner it will run. "Left is lean" on that externally adjustable center idle screw. Left backs it out, and allows the flyweights to open further and pull the space cam further back. It would stand to reason that holds true for the mid and high speed screws as well.

Anyway, it's getting clearer what I need to do (sort of). My guess is the black and white adjustment screws set the overal range of adjustment in which the pump can be adjusted with the "normal" adjustments. They move the center of that range up and down. How much? Hmmm... time to find out.

My pump runs rich in the 3,000-4,000 rpm range with the idle screw at full lean. Any leaner and its hex head is backed out so far it clears the flat springs and doesn't click anymore. I need to re-center its adjustment range. The part load screw on the rack is about 1/4 up the range from full lean to full rich right now. It has a lot of adjustment left towards full rich.

So, here is the plan. I'g going to leave the white high speed screws alone for now. I'll back out the black mid range screws some amount and leave everything else alone. When I dig into the running pump, Ill see how far out they are, and how much further out they can go. Then I'll make an educated wild-ass guess on how far I'm going to back them out.

Keeping in mind that (I hope) I'm just moving the range in which the other two adjustments work, hopefully I'll be able to use them on the dyno to to do the fine tuning. Stay tuned...

chan1408 · 01-11-2008, 06:24 PM

Hello Jeff and all others - great stuff here. I've learned a lot in the past few minutes.

I just left my two pfennig worth on the other MFI thread re thermostat - while any of you are so deep into disassembled pumps, can you verify the mechanical logic of the thermostat? My understanding is that it is supposed to compress the thermostat mix rod until the mix rod spring is at coil bind or to some other stop when the engine is full hot. Can anyone confirm? Sorry for the hijack, again, great info guys.

Thx

Chip

Jeff Higgins · 01-11-2008, 07:49 PM

Cold thermostat position:

Hot thermostat position:

Note in the "cold" photo, the plunger on the left has it's inner rod fully extended and visible. It is fully compressed in the "hot" photo. Also note the effect on the eccentric slot on the right, and the rod it controls. You can see how it affects the position of the compensating lever on the end of the control rack.

304065 · 01-12-2008, 05:52 AM

Here is some info on specs on the extension of the thermostat rod.

MFI thermostat data. . . and distaster!

pjh69911 · 01-12-2008, 07:27 AM

Jeff,
When you start adjusting, write everything down so you can always return to your previous settings. I'm not sure what range each spring covers, at 3-4k you may be into the heaviest spring. Remember, the centrifugal force rises as the square of the speed! You might make a moderate adjustment on the screws to see the effect, and then adjust as needed. Paul.

69911e · 01-13-2008, 04:44 AM

Quote:

Originally Posted by jcge

I've seen a few posts on this forum that report the space cam as having no influence on injection volume beyond 4000 engine rpm (tell me if I've misinterpreted), but I don't think this is correct. The following document Y 428 904 601 for 911/72 motor style shows non linearity well beyond 4000 engine rpm (maybe the reference is to 4000 pump rpm = 8000 engine rpm..)
John

It does look like it is pump RPM, ie 8000 engine RPM. It looks like a decrease in flow after 4000 RPM. I wonder if this is not due to a mechanism in the pump, but rather flow restrictions for the inlet or outlet to the pistons or possibly the springs for the piston (returning it to the cam).

I wonder if caution in the high speed spring adjustment is needed to stop the spacecam follower from going too far and going past the edge of the spacecam.

This one of the most interesting threads I had ever seen on this BBS. Strange how this adjustment wasn't presented in most documents. Please post any results.

Ed

pjh69911 · 01-13-2008, 05:42 AM

Jeff,
I found some information in a mercedes manual that references tuning of MFI pumps. It is for the 280 application, a 2.8L motor, max rpm 6500, and the pump internals look similiar, though I have not measured the spring sizes or stiffnesses. They suggest not to ever adjust the white screws, because at full load they direct you instead to change the rack screw, "Alteration is not authorized because the adjustment of the white screws would also result in an alteration of the full load range or in a displacement of the entire performance graph which could then no longer be kept under control". It also suggested, that the lower partial load(which I guess relates to the middle spring) is measured at 1600 rpm, and the upper parial load(heaviest spring) is measured at 3200 rpm. So if you are too rich in the 3-4k range, and wanted to follow their tuning algorithm, you would have to lean out the full load rack screw, and then richen up the black lower partial range screws, and the idle screw. Paul

jcge · 09:42 PM

Quote:

Originally Posted by Jeff Higgins

John, I'm having trouble reading the numbers on the chart, or even what the different axis represent. Is that fuel volume at various throttle positions vs. rpm?

Jeff- its not the best picture in the world - but the bottom (X) axis represents the pump speed in RPM - it extends to 4250 pump rpm == 8500 rpm at the crankshaft

The left (Y) axis represents fuel volume delivered (in cc's) for 1000 pump revolutions

The curves on the chart represent the different throttle poistions (expressed as both the throttle angle, and the corresponding control lever angle, hence the 2 angle.......CORRELATION)
WOT is the top most curve, down to the idle curve (0 degrees) at the bottom

(and peaks on the top WOT curve at about 67cc/1000 pump strokes at 3750 pump / 7500 rpm for this engine)