MFI Thermostat Hose
 

So I'm getting the car ready for Summit Point and I can't figure out why the mixture is around 10:1.

So I do a quick CMA-- any competent porsche driver tracks faster if brakes come smoothly-- air-filter compression-loss plugs dwell timing fuel-pressure injectors belt correlation smog. Ok, so I did everything but the injectors but they were Ok last time I checked.

But the thermostat won't turn off? Look what I found!

http://forums.pelicanparts.com/uploa...1158449132.jpg

The old hose was hard as a rock and shattered, probably due to motion of the shock (it's a 72- large diameter shock) against the heat exchanger.

A quick trip to the FLAPS resulted in a 1.25" x 18" radiator hose until the "steam hose" can arrive from our host here.

Anyway, I put it back together and while I was in there I cleaned the thermostat. I found that the big spacer was on the forward end of the rod, and the little spacer on the aft end, after the 25 pairs of discs. So I put the big and little spacers on the aft where they belong according to the diagram.

Needless to say every mosquito within a 2 mile radius is dead and I may have to invest in six more NGK's tomorrow, the current ones are so fouled.

I think what happened was twofold: moving the big spacer to the aft end resulted in more of the rod sticking out of the housing, richening the mixture dramatically. Even though that's the factory orientation, it's wrong for my setup. So it's a simple fix to revert back.

While I was in there I decided to snap a photo of the pump lever for Milt's benefit and for future use in thermostat diagnosis.

http://forums.pelicanparts.com/uploa...1158449581.jpg

Grady, in another MFI Thermostat Thread , posted an excellent diagram of how the thermostat operates on this linkage which is reproduced below.

http://forums.pelicanparts.com/uploa...1158449621.jpg

So tomorrow it's:

1) Pull thermostat and move spacer;
2) Clean plugs, or install new plugs from stash;
3) Set part-load and idle mixture after screwing them up trying to fix mixture problem caused by broken hose.

The road is not always clear but all MFI problems can be solved if you KEEP THE LONGHOOD FAITH! *

*with credit to the Late Harry Pellow

Thanks for posting this. So according to this picture "A" does not belong in "B".

http://forums.pelicanparts.com/uploa...1158461361.jpg

John,

That old hose looks like it would make a good substitute for a baseball bat as a defensive weapon!

I replaced my stiff hose with a length of Gates Green Stripe 1" HD heater hose ... worked fine, even though it didn't have the same number of plies! That OEM hose seems to have 8 - 10 plies!!!

Warren, you are right, that old hose had a wall thickness of about 3/8", and multiple cord plies! Thank you for the suggestion about the Gates heater hose.

Unfortunately, the pump lever is now binding, so I will have plenty of time to wait for an OEM replacement!

John: All of the pumps I have seen did have A engaged in B. Your configuration would cause a lean mixture requiring a compensation in the thermostat.
ED

Interesting, eD, except there is absolutely no evidence of this pump ever running lean like that. On the contrary, the last two years of operation, including data logs, show that it runs between 10.1 and 12.8 to 1.

Why do you say lean? I am not disagreeing but I am curious as to how the linkage works.

John: The effective mixture is a combination of thermostat, barometric sensor, and rack position (plus other things). A change in one can be compensated by a change in another. My statement of lean was assuming an otherwise correctly adjusted mixture.

The barometric sensor and the thermostat work together to determine the "gain" of spacecam follower to rack movement. The effective gain of the settings you have can be seen at the curved slot at the right side of your picture (directly below A&B). The pin within the slot relative to the slot endpoints can be measured to quantify a relative gain change as adjustments are made.
Neither one of these should be move much from the factory setting unless the engine configuration is not stock. As far as I can tell the pin in the slot should never touch the slot's endpoints when the engine is warm otherwise the barometric sensor is effectively disabled.

ED

John,

With the thermostat hose broken,you would have done well to get 5 mpg at best.
A similar disconnection happened to my Carrera motor just before a long holiday trip.I wondered why I was getting poor mileage!

Cheers,
Dennis.:)

Realizing that this thread is taking a fork in the road, I will submit a lengthy discussion on the rod and the arm:

http://forums.pelicanparts.com/showthread.php?s=&threadid=292246&highlight=mfi+pu mp

bump

John, what did you find?

Best,
Grady

Grady,

Current status is here:

http://forums.pelicanparts.com/showthread.php?s=&threadid=304926&highlight=distas ter

However, I am going to send the backup pump to Gus and then crack open the one on the car. . . just to see what happened, and then send THAT pump to Gus. Will keep you in the loop for sure.

One thing that you will find as it relates to the thermostat is that is will function properly when hot.
The hot air from the heat riser is designed to heat the thermostat at a rate quicker than the engine. As the engine heats up, the oil that is pumped into the pump will heat the entire pump to engine temperature and in during so will heat the thermostat to hot running position.
This is designed as a way for the pump to know that you are trying to hard start the engine.
After about 15 minutes of running the thermostat is as hot as it needs to be with or without the heat riser.

If you doubt what I say, disconnect the heat riser, run the engine for 15 minutes and them lay your hand on the thermostat. Careful ! it's hot.

Wanted to send the MFI to Koller (lean rich problems I couldn't fix).

Luckily I discovered yesterday that in my MFI the pin is also not in the hole.

My nabour checked the MFI....same problem.

I fixed the problem, car is now running very smoothly and strong!!!

Found this thread afterwards.

Really don't undertsand why this happens apparently so often.

Any idea?

Roland

I have no idea, either, but I'm convinced that the pin should be in the arm. It's not all that easy to put it in there, so I doubt if it just "slips" out. Perhaps the altitude compensator is off and mechanics disconnet it. I don't know. I will continue to connect them until I know otherwise.

Wasn't easy to fix it but now the whole compensation unit is working properly.
It took me 1 year to find the problem!!
Very frustrating process.
I have no explanation for whatever reason one should remove the pin.
If you want to get rid of the altitude compensator it is easier just to replace it by a screw (without negativ effect on the whole sysstem.

Roland