Another MFI correlation attempt. . . LONG

[304065]

Well, it's a long, painful story, but I had finally determined that my MFI throttle-bodies needed to be rebuilt, due to my inabilty to synchronize them and reduce the idle below 3000 rpm.

So the process began. Removing the aircleaner and stacks was simple, just undo the M10 nuts, dont drop the washers down inside the engine or bind the cross-shaft when removing.

Slightly more difficult is removal of the throttle bodies. The 12mm nuts are the high-temperature lock type, like one would use on the exhaust system, and while not torqued very high, one needs to use an obstruction wrench or swiveling socket to get them undone. Having no such tools, I had to use an open-ended spanner, laid down atop the green shroud, would make a small movement, flip the wrench over and make another, and so on. Took about one hour to remove the throttle bodies.

Notably, to remove the outboard nuts for cylinders 2 and 3, the injectors had to be removed. To remove the injectors, and the throttle bodies themselves, the injection lines had to be removed, which is fairly simple, 17mm.

Now, here's the fun part. I held the throttle bodies up to a bright light, and on the starboard set, with the throttle stop screw dialed all the way out, I could clearly see light coming around the edges of the throttle plate-- looked like a solar eclipse. This was most pronounced on #4.

In the middle of all this I lost my storage space and had to move the cars. But how to get the car on the trailer without a working engine? Everything had to go back on.

So I thought I would give it one more shot to see if I could make them work without a rebuild.

Anyway, that's a long-winded way of building up to the fact that during the reassembly I tried the shop-vac flow bench trick. Unfortunately, I didn't have time to build a proper setup with an adjustable flow and manometer-- so I just ended up wrapping a couple lengths of duct tape around the shop vac nozzle, which then fit perfectly into the base of the stack. My synchrometer fit nicely in the other side.

With the vacuum on, all the throttles read approximately 3 kg/hr.* When one was too high, I made sure the air bypass screw was seated, then noted the readings again. Closer this time. Finally, I set each one to 3 kg/hr exactly, using the throttle stop. Interestingly, you want each plate to be slightly open, so setting to 3kg/hr was the lowest setting I could repeatably measure while ensuring that the throttle plate wasn't completely closed.

1 CFM = 2.081 kg/h for air. So 3 kg/hr is about 1.4 cfm.

With individual plates set, it was time for the linkages. I remembered Grady's admonition to clean the ball cups of any old grease and only use a light oil for the setup phase, so I spent a few minutes cleaning the cups. I was able to get all the rods set to the correct length this way, with a little play in the cups. Unfortunately, I didn't have the time to mic the cups or the balls, so I don't know if they are worn beyond limits, but next time I shall.

When you set the rods perfectly you'll know. They snap into place with little resistance-- you have to take care that you don't inadvertently make the side of a cup contact the ball when you tighten the locknuts.

Anyway, using the improvised flow bench, I was able to measure the throttles again with the linkages connected. This time, the flow for each was about 3.5 kg/hr for all of them, some slightly higher or lower (the resolution of the synchrometer is actually pretty fine) which I figured was good enough to start with.

So it was time to put them back on the motor and try it. Unfortunately, I had a bonehead moment, I forgot the new gaskets from head to throttle and throttle to stack. But I had no time, I figured the seal would be pretty tight, so (horror) I put them back on without gaskets. I have to take them off again anyway (foreshadowing). . . .

It was easier to get the throttle bodies back on by reversing the process. Does ANYBODY have a convenient method or tool for getting the inboard nut for the #1 throttle body back on? I fought with it for way too long. Also, tightening torques aren't that high, but has anyone figured out a socket/wrench combination that works well? A 12mm on a 1/4" drive ratchet was the only thing I could get to work, maybe I need a wobble extension.

Anyway, after putting everything back and double-checking the high-pressure injection lines, I fired the engine.

Result. . . .3000 rpm idle!

Now, having the idle up so high when the pump was at the stop tells me that I still am pulling too much air. So I tried the synchrometer on the now-running engine, and there was a great deal of variability in the individual stacks. Unfortunately, #4 would not close down far enough, even when I pressed on the linkage to be sure it was all the way closed, to deliver a synchrometer reading of less than 7 kg/hr. So I tried synching everything to 7, which just made the problem worse.

So it's off they come for another adjustment to see if I can get it right again. This time I'll use the proper gaskets, too, to prevent any EXTRA air from getting between the head and the throttle body. Also, some kind of a manometer between the vac and the throttle body with a valve for adjustment would be helpful to make sure I have the same flow at the beginning of each stack.

I don't have a set of protractors. I bet we could have some sets made by the guys who have CNC machines.

[svandamme]

John , when you work on the correlation , did you have your throttle cable connected?

i noticed that it's a royal pita , to do anything on the TB's, with that one attached, the hand throttle and pedal linkage on mine always affected the control rack , even if only a little bit, it's there..

mine wouldn't idle properly at all, so i needed something to keep the rack slightly open, so i used the shutoff screw to control this while testing

for idle adjust, you could actually remove all linkage completly
and fully work with the throttle stops and bypass screws to set correlation. they are guaranteed to be in the correct position for idle this way

once you get correlation , you can simply attach the individual linkages , while keeping an eye on the correlation , if it changes , the little rod is not correct in lenght.. keep adding , till you got them all hooked up

then the 2 links to the control rack, again , everything should be at the stops, and the correlation should not change..

it's slow , but it does work well

[304065]

Stijn, excellent points.

Yes, I had fiddled with the throttle cable to make absolutely sure it wasn't too short and pulling on the cross-shaft, which would open the throttle plates.

I remember from the Lee Rice/ esses article, one can set correlation with all the linkage rods removed to have a pure throttle stop/air bypass idle.

I'll try that trick when I R&R everything again.

[Jandrews]

This is an interesting thread...please keep us posted, John.

One question...what happens to the idle if you close down all the air adjustment screws tight?


JA

[svandamme]

Originally posted by Jandrews

One question...what happens to the idle if you close down all the air adjustment screws tight?


JA

technically , those screws are only there for the tiniest of adjustments

they shouldn't affect idle to much really , they are mostly there for the perfect correlation, you can't correct a 30% difference in correlation with them , if it's off that much , you have to readjust the throttle stops first... the screws are for that last little bit of difference...

can't control the idle speed with those either, at least not effectively

- set the throttle stops with the correction screws almost fully closed (fully closed, then 2.5 turns out again)
- when things are very close to right, finish off with the correction screws

[Grady Clay]

John,

I’ll add one final thing to Stijn’s procedure; everything needs to
come off the stops at EXACTLY the same time, six butterflies
and the pump.

Now you are ready to check correlation. Correlation has a great
deal to do with how each of the lever arms goes across center
(where the arm is perpendicular to the rod.) That is what this
wonderful chart is all about.
IMAGE: MFICorrelation01.jpg “
"
© Dr. Ing. h.c. F. Porsche A.G.
CMA p. 18

There are basically nine positions that the throttle bodies must
track the pump by this chart. The closer the better. It is critical
you always push on the ball-socket of the link going to the
accelerator pedal so you don’t move anything in an unusual
direction.

I sympathize with the effort you went through. There is real
advantage to being able to do this start to finish, uninterrupted.
You still have to take things apart multiple times too often.
Remember; don’t count on the throttle body going back on in
exactly the same place each time, they don’t have locating
dowel pins.

Best,
Grady