CMA and all kinds of other useful MFI stuff can be found here

Pelicanparts, where else?

you can download a nice clean, high resolution copy of Check, Measure, Adjust from the link in my signature area... good luck and please post your results for us MFI types.

d.

Shuie,
John Hunt’s compilation of technical articles and advice is great, study it carefully. http://www.pelicanparts.com/techarticles/911_MFI/TipMFI.htm

Does anyone have other good sources?

Best,
Grady
gradyclay@hotmail.com

Shui,

Lot's of good advice from everyone. Through all this keep in mind that the MFI system has a reputation for being difficult only because people don't understand it. Basically, it is no more difficult, amd mo less reliable than any other fuel delivery system.

After being sure the lengths of all the rods are correct, especially the one from the pump to the accelerator linkage, and that the idle ports are clear, and the screws all out a few turns, and that the butterflies are resting on their stops and closed, and that both hoses from the heat exchanger to the MFI pump (there is no third hose to the airbox) are clear, here's what I'd do. (Except for the clear hoses, my guess is that all the other adjustments are substantially correct or it wouldn't have run well before.)

1) Check the cold start solenoid as someone suggested. You can either look and see if fuel is squirting into the stacks when cranking a cold engine, or you can remove the wire from the solenoid and apply 12 volts. You should hear a click. If you hear the click with 12v, but don't hear it or see fuel squirting when cranking then probably the hot timer thing (can't remember what it's called) isn't working and has failed in the off position, thereby rendering your cold start solenoid ineffective. Just bypass it, i.e., jumper it, thereby ensuring that the cold start valve will always work, whether the enigine is hot or cold. Those hot timer things are notorious for failing, and lots of guys (like me) keep it permenently bypassed. Under these conditions you need to keep the pedal fully depressed when starting a hot engine.

2) Once you're sure the cold start valve is doing its job, I'r remove the warm-up solenoid. Don't just disconnect it. Remove it. Check it out as someone has suggested by seeing if the little protruding rod protrudes even farther when hot. The range of movement from cold to hot is small. I don't recall exactly what it is, but it's on the order of 6 or 7 mm, maybe 8 or 9, but not more.

3) If it's not working, clean it as others have suggested, and check again.

4) Now, with all of the above accomplished, it's time to adjust the pump. My technique (I've rebuilt 4 911's with MFI systems, and the rebuilds included the pump, to give you a basis analysing my advice. Not a big number, but not zero either.) is to lean out the part load adjustment a few clicks at a time (keeping track so you can always return to where you were) until it begins to idle poorly, and then go rich a few clicks from there until the idle is good. This will result in a nearly perfect adjustment. If you want to, you can then fool with the idle adjustment on the pump, but it doesn't do a whole lot to put it mildly.

3) You will than want to refine the idle speed and balance the stacks with the adjusting screws on the stacks. Balancing is easy too, but that's another lesson.

One last comment. People seem to find the Gunderson CO tester useful. I never have. It's too finicky. You can get one rwading one minute, and another the next, and they can be very different. One thing I've done, which is interesting, and also useful, is to temporarily install an O-2 sensor in the tailpipe to monitor fuel/air ratio. It's kind of neat to watch how it changes under various driving conditions, and it's possible to use it to adjust the mixture to the precise stochiometric ratio, but even that isn't necessary. Do the click to lean and stubling idle, then a few clicks rich from there, and you should be just fine.

Have fun.

MAn I would look at those points FIRST. Being that it was running fine. It just sounds like the points closed up on the way home. I know I said that before , but sometimes it is a simpler fix that you may think.

Actually as soon as you can..get rid of the points. Put in an optical system like the Crane unit. Set it and forget it.

Shuie,
targa911S is right. Silple is most likely. Also optical/magnetic, no distributor is better but not for this discussion. We want to get you running correctly.
Give us some more diagnosis/symptoms, Shuie.
Everyone benefits from diagnosing the problems.

The Factory publication “Mechanical Fuel Injection – check, measure, adjust” is available at:

http://www.scatliff.mb.ca/pelican/MFI_Check_Measure_Adjust.pdf

Thanks Derek.

Does anyone else have other good sources?

Best,
Grady
gradyclay@hotmail.com

Guys, as a way of introduction for those of you new to the Porsche world, Grady Clay knows his stuff. His shop was THE place to take a Porsche in Denver back in the 70's...Welcome to this board Grady, your advice is more than welcome! :D

MFI is not really that difficult as long as you understand it. Having teethed on it way back when (was at dealership then), I have worked on a few. Porsche published "The Ten Demandments" for MFI back in the 70's. We still go by it today, even though we usually can eliminate some of them if we know the car.

Here is what I would do:

Compresion and leak test. Check condition of plugs, wires, cap and rotor. If ok then move on.

Dwell set to 37 degrees, timing set,

Check pump timing. Correct if necessary.

Check length of pump rod. Correct if nessary.

Start car and check co - set to around 4% at idle. Note if the oil is fuel diluted, you will get a false co reading. Also if the stacks are allowd to heat up you will get as false co reading.

Drive car and check for light throttle response and if surges. If so the part load may need to be richened. Be careful as you can go so rich on part load that it will override the idle circuit and you will fight eachother. I usually lean out the part load untill I have slight hesitation on acceleration and then richen a couple (2) clicks.

Hope this helps.

WOW! Thanks a lot everyone! Ive been able to find some good posts on troubleshooting MFI through the search, but this is one of the most informative single topics that Ive seen.


When I picked the car up, it was almost completely out of fuel. I didnt run out of fuel before I filled it the first time, but it was really low. Since it is a low comp engine, I didnt think of using anything but the regular 87 octane gas. The station attendant knew the car and told me after I filled it that the previous owner had always used the best gas they sold, not the cheaper stuff I was using. I didnt think anything of it and hopped back on the road to start a 750 mile drive home.

When I pulled off the interstate at the end of that tank and came to idle, the car died. I was able to restart it and pull it into the gas station on the exit ramp. I filled up and, no start. I got a pushstart and hopped back on the interstate.

At the end of that tank, I pulled the handthrottle on the exit ramp and filled up with the car running, adding a quart of oil. Like a complete dumbass, I stalled the car with a huge backfire when I got back in it and couldnt get it started again until it cooled down.

Next gasstop same handthrottle routine, but I managed not to stall it. I made it home on that tank. I downloaded CMA the next morning. I checked the things I was comforatble with, but when it came time to measure and adjust it went to my wrench. The car is still in his shop as of today.

Any chance I threw everything off from changing to a lower grade fuel? When I brought my wrench the car and gave him the rundown, his first suspicion was bad fuel but not necessarily lowgrade fuel.

Thanks again everyone! I really appreciate the help.

Shuie,
I agree it's not low grade fuel. Driving on the highway at normal speeds won't stress the octane rating on your '72T. Perhaps on the track, but not on your driving home. Not to worry. the issue with octane is detonation. In your drive home, I can't believe that is an issue.

If the mixture was off, I understad why it died at idle. If the mixture was off (rich or lean) that could cause the big backfire.

I don't understand why it wouldn't start on the starter motor yet would push start. Did the temperature change?

No, there is no chance that "low-grade" fuel caused a problem.

Best,
Grady

It's sounding more like points with every post.

targa911S, What do you mean by points?

Grady, the starter would turn over fine, car just wouldnt fire.

The more I read, the more I think Im going to find a ignition/timing problem. How do I 'check the points'?

Hi Grady, I mean ignition points. It just sounds to me like the rubbing block on the points themselves wore down from the combo of no use for a while (lack of lube) and the long sustained rpm of a 700+ mile trip. As it was running OK when he left on his journey then slowly decayed to backfiring. Especially that has me thinking the point gap closed up. Plus it won't start and run other that at higher rpm. You have real good credentials and I am just an old time diy air cooler but it sure seems like that could be the problem to me. Sure a lot simpler fix than dinking with the MFI when it could be avoided. In my experience MFI seldom goes out of whack all by itself. Just my .02

Two years ago a I purchased a 911T with MFI from a volkswagon mechanic for a song. So I , by necessity had to learn a thing or two. Seems to me if your engine cuts out at warmer temperatures, your thermostat is working quite well.What it does is , as the engine temperature increases the thermostat will lean the mixture out . That is the discs expand pushing out the rod which pushes a lever that leans out the rack. I actually purchased an extra pump thinking that I would have to replace the one in the car. So I did some observations and experiments. One thing I learned , as far as the order of the thermostat washers , it 's not order but orientation. They should go like this - ()()()()(). Try disconnecting the thermostat hose from the thermostat. That way when the engine warms up the mix will not lean and your engine should still be running. If thats the case then the rack needs to be adjusted to enrich the mix, as per check measure adjust. Also as far as the hard cold start goes, check your plugs for fouling. MFI runs rich so plugs tend to foul. I just had this problem and a new set of NGKS did the trick. The specified Bosche are too cold. Also go though the whole ignition, plug wires , rotor , cap , coil , CD box etc. Oh but check the simple things first as mandated by the ten demandments.

Two years ago a I purchased a 911T with MFI from a volkswagon mechanic for a song. So I , by necessity had to learn a thing or two. Seems to me if your engine cuts out at warmer temperatures, your thermostat is working quite well.What it does is , as the engine temperature increases the thermostat will lean the mixture out . That is the discs expand pushing out the rod which pushes a lever that leans out the rack. I actually purchased an extra pump thinking that I would have to replace the one in the car. So I did some observations and experiments. One thing I learned , as far as the order of the thermostat washers , it 's not order but orientation. They should go like this - ()()()()(). Try disconnecting the thermostat hose from the thermostat. That way when the engine warms up the mix will not lean and your engine should still be running. If thats the case then the rack needs to be adjusted to enrich the mix, as per check measure adjust. Also as far as the hard cold start goes, check your plugs for fouling. MFI runs rich so plugs tend to foul. I just had this problem and a new set of NGKS did the trick. The specified Bosche are too cold. Also go though the whole ignition, plug wires , rotor , cap , coil , CD box etc. Oh but check the simple things first as mandated by the ten demandments.

What a great post; all the gurus of mfi chiping in!

Sounds to me like the car isn't getting fuel.....but the CMA should be adhered to at all cost.

Not much to add here, especially as I haven't played with my MFI for AGES since getting it 'right', but I have to say this board is where it's at when all the paid mechanics fail you!
Better to learn how to do it yourself.

Bob Spindel! I have some 4+ year old printouts of your advice (which I followed and won!), so it's great to hear from you again. fwiw, I tune my car the same way re. mixture.

I have to say that not many people feel comfortable doing it that way and would rather use the Gunsons or similar CO tester.
Ear and nose is better :) but I appreciated that's a skill that takes time to learn.
Warren- excellent!

Good luck

Ed- spot on.

I think it's a common thing that bonds the MFI guys. Nobody initially knows much about the system. But of course in reality it's just about learning it. Anyone with a mechanical bent can understand it, so well done on your efforts.

The T.Stat washer orientation is correct. You can check if it works properly by heating it externally. I think I ended up using a butane torch carefully, although obviously watch out for situations that place your eyebrows at risk..

If all else fails console yourself that even a completely disfunctional MFI system will allways look better than CIS fungus atop your 911 longblock.

So, what's the consensus on plug choice?

Tom