MFI thermostat data. . . and distaster!

[304065]

Well, in the interest of expanding the knowledge base here I decided to take some measurements when I was putting my MFI thermostat back together after determining that the rubber air hose was severed.

So the first question was, what is the length of the spacers at low temperature? Sorry fellas, I didn't have a k-type themocouple or even my cheapo digital thermometer handy, so we'll have to take a guess at the temp, but I'm guessing around 40.

Note that there are 25 pairs of discs and the thick and thin spacers are on the FORWARD end of the rod. They are supposed to be on the AFT end but I moved them just before the photos to see if it made a difference in operating length at temperature. It didn't.





So installed length is about 75mm.

It does NOT make a difference with the spacers on either end. At that length, the stack is SHORTER than the length between the bushing end of the thermostat housing and the ledge that the spring retainer rests on. Therefore, when you install the discs and the two spacers, there's significant free play or "rattle" in the discs. Also, with the discs cold, there's significant variation in the rod length -- it sticks out 15 mm if you push it all the way up to the spring retainer, or about 18mm if you pull it all the way to the bushing end. When it gets hot, and I blew 600F air through the housing in the same manner it gets 250F air from the paper hose, there is no rattle whatsoever, and it protrudes a full 19mm from the housing.

In installed operation, what actually happens is the spring inside the pump is strong enough to push the rod aft, so your range of governing is really 15-19mm protruding from the thermostat housing. All these measurements were made between the end of the rod and the part that holds the bushing, not the mounting face. My spacer measured 3mm and the phenolic gasket-type spacer measured .3mm. Interestingly, it is not possible to install the thermostat housing without either of those present.

Next, I installed the thermostat on the pump and measured the difference in the spring pivot height inside the pump. It was about 5mm of total travel, which roughly parallels the change in thermostat operating rod length.





So you can see how the pump rod length influences things inside the pump.

I then buttoned everything up and got ready to start the engine. I heated the thermostat up with the heat gun and verified the rod was extended, then fitted the cover plate so oil wouldn't splash everywhere.

I started the engine and got, instead of idle mixture, the usualy 10.3:1 rich running. Hmm. Something wrong here. When blipping the throttle to keep my new set of BP7ES plugs clear as long as possible, I felt the throttle stick. Uh oh. Removing the pump linkage revealed that it wasn't the throttle bodies at all, but rather, THE LEVER ON THE MFI PUMP WAS STICKING IN THE FULL THROTTLE POSITION. It would not allow the throttle bodies to return to idle.

After looking inside the pump (visually only) to see if anything was wrong, I concluded that the arm that operates the pin that rides on the space cam was freezing at some point inside the pump where I couldn't see, below where it entered the oil bath. GAME OVER.

Tomorrow I shall make phone calls to the usual suspects but I am not optimistic that I can get the pump rebuilt before Summit Point.

Has this happened to anyone else? I am aware of at least one other case where the MFI pump stuck at WOT causing an off-track excursion and significant body damage. I don't think it's possible to fix this without disassembling the pump, which I'm going to do once I send the spare pump off for a rebuild this week.

I had moved the idle mixture adjuster through its range of clicks, I know there is a failure mode where one of the adjusters will come loose inside and kill the pump, but I can't imgaine this is what occurred. This was a progressive deterioration, same problem manifest itself at Lime Rock, and the corner workers said I was blowing black smoke, and the data log confirmed very rich operation. I thought the thermostat hose would be the culprit but evidently fate has intervened. . .

I'm keeping the MFI faith but it ain't easy. . . *

* too bad Harry Pellow never wrote a book on MFI, it would be a great read!

[lateapex911]

Interesting...my story is similar, but different, I think. On a trip to California, my car suddenly started running very badlyin rush hour traffic outside Chicago. I barely got it off the road, but at big throttle opnings, it would run...barely. It was towed to Midwest Eurosports, and the tech there spent the better part of 3 hours the next morning with his head jammed in the engine compartment, and on the phone with the guru in Rockford Il. This was in '93. In the end, he said that the linkage had jemmed full rich. Amazingly, he was able to repair it, and all was returned to normal.

But it did serve as a reminder that, at the time (pre net) MFI expertise was rather hard to come by. I was lucky.

[304065]

Jake, I'm going to give Gus a call at Pacific Fuel Injection, maybe this is a common problem?

The pump lever DOES move, but it encounters significant resistance. There is no metal grinding or gnashing and no chips that I can see, just friction. I shot the pump arm inside and outside with PB blaster to no effect. The return spring for the pump arm is intact with tension.

I don't think the arm is binding, rather, I think that the space cam is unable to slide freely fore and aft. I intend to find out, even if it means sending the parts individually wrapped to Gus after the spare pump is rebuilt.

[lateapex911]

Well, if anyone in the country would know, it's Gus. There used to be a guy here on the East coast...(LI?) who was well regarded, and had the right equipment, but I really haven't heard his name come up in MFI converstions in years.

Of course, the problem with Gus is the backlog. Running MFI really does mean needing spares and the right gear to tune. You're almost there!!

[Zeke]

Wish I could help. I have been inside these enough to understand some basics. Why do they call it a "space cam?' Anyway, it seems as though the main rack is always free to move with no binding. There are forces exerted against various levers that change the mixture, I guess we all know or assume that. There isn't as much direct linkage as one would think. In order to troubleshoot, all the covers and the thermostat have to be off, IMO. Something has to become apparent at that point. I'm wondering if one of the set screws on the six, geared swivel gizmos loosened and it's binding there. Just one of a few possibilities that you would see right away.

[304065]

I just got off the phone with Gus, he says that binding of the throttle lever is an unusual problem.

Milt, I appreciate your weighing in here-- I checked the rack itself, which of course meshes with the adjusters on the pistons, and it's not binding, it moves freely back and forth. The governor levers all move freely. The only thing that binds up is the throttle lever.

It's rebuild time.

[304065]

Another view of the sticking part. Somehow this must transmit the throttle position to the cam roller (shown at bottom) which then transmits the correct rack position back to the governor.

Milt, have you ever had this apart to this level? I would be really interested in knowing how these pieces fit together.

[tadd]

John:
I've got three pumps in pieces (so far), well two and one with all the covers off. The one not fully disassembled is the super nice 2.2S (009) that is like pristine. It has been my reference for the crap 2.2S and 2.2L(?) - read 010 pump.

So observations... that joint is kind of unsupported for what it seems to do so if that pin elongates or wears then the motion between the two arms binds rather than moving smoothly (note: the 'C-' pump has less play than the 'F' pump). I say this as the two disassembled 'crap' pumps (stuck plungers) have play there (like sicssors that need the bolt tightened so the blades widen instead of cutting) but the nice pump doesn't have any play. It just rotates (swings?). I haven't really looked at them to figure out if it is the pin or the hole that has worn. My guess would be the hole since it is only a mild steel stamping.

Does yours have play (like doing a turkey wishbone?) or am I a victim of too few data points?

The wear also scales with price! 750 = 0, 400 = some, 50 = holy!

tadd

[MURF]

hans at h&r fuel injection is still around and on long island his # is 6315891600....in bohemia...he has all the test machinery to set up or repair what you have

[jm993rs]

Thanks for this explain thread !!!