A little (really!) MFI help, please...

[JohnKo]

Going through the "Check, Measure, Adjust" (CMA) routine, both CMA and the full manual refer to making "fine adjustments" to pump timing at the drive belt by loosening three allen-head screws. I've tried to find pictures, but I don't get it. If the crank is at the proper F-E position, and the belt is on and tight, how does loosening those screws free up anything to move and/or make an adjustment? Am I just having a brain fart? TIA
John

[Grady Clay]

John,

That is Porsche’s idea of ‘really fine tuning’. It makes relatively little difference in the overall scheme of things but worthwhile while you are doing everything.

There are not simple screw ‘holes’ in the pulley but circumferential ‘slots’ that allow the bolts to be loosened and the pulley rotated relative to the MFI pump drive hub … or is it the cam? I’m off in the wilds of central Missouri. I’ll correct this with images when I get back to Denver. The adjustment is only a few degrees but it is designed to adjust finer than one tooth on the belt.

One of the ‘arts’ of CMA is very through C&M, make everything work correctly and only then adjust to perfection – whatever that is. I tend to lean to ‘non-perfection’ where there are still some lean running issues but overall it works great. Of course MFI is really a ‘race’ system that somehow Porsche got to comply with emissions – for a while. MFI always shines at full throttle (well, should).

Best,
Grady

[JohnKo]

Thanks, Grady.
I'm to the point in CMA where it calls for checking pump timing, and I just wanted to understand the 'mechanism' before getting in there.
Safe travels,
John

[JohnKo]

Well, the plot sickens! CMA led to checking pump timing, resulting in the discovery that it was ~180 degrees out. I'm amazed the car ran at all. After having set it right, I got all excited by the having finally found the source of my problem. It runs better, but the big let-down was that my fundamental problem remains -- I get a lot of loud backfiring in the exhaust, on accel and decel, which suggests to me that it's running very rich. I've tried making a small mixture adjustment, but it doesn't feel like it's supposed to, and I worry I've damaged something.
Any ideas out there?

[JohnKo]

Originally Posted by JohnKo

Well, the plot sickens! CMA led to checking pump timing... it was ~180 degrees out.

Allow me to clarify that it was not ME who made that error -- PO or PO's mechanic.

[Fast Corners]

A few benefits from past experience with MFI.

You probably did adjust the pump correctly but I have known more than one mechanic to set up the MFI on the wrong cylinder by accident. You need to read the CMA manual carefully. On page 16 it says set the engine ti TDC of #1. The rotate the engine 360 degrees (Basically TDC on #4). Then turn the crankshaft to the FE mark. After this you adjust the pump to match the pump shaft timing mark with the pump housing mark. The engine will run 180 degrees out but not well.

Don't worry about damaging the pump you really can't. Now damaging the engine is a different matter. I doubt that you have. Make sure that the pump is timed correctly (FE mark past #4). If you are confident in all of the other adjustments here are my suggestions.

Check the function and the adjustment of the micro switch. The switches are pretty failproof but you never know. After you have done that, remove the silver Bosch fuel injection bok from the electrical panel on the left side of the engine compartment. Carefully and gently shake it next to your ear. If it rattles then some of the components have come loose from the circuit board. This box is basically a tachometer that lets the pump know to shut off the fuel delivery on overrun (decelerating conditions with the engine engaged to the wheels) conditions. If this is not functioning properly, you get raw fuel being dumped into the exhaust pipe...resulting in back firing (and in extream conditions a large flame comming out of the tail pipe...It is quite impressive actually).

Let us know what you find out. You will need patience and be very careful and sure about your measurements and adjutments. MFI is not Carbs. You shouldn't fiddle or tweek things to much. The book says it all...Check, Measure...Adjust. What it should say after those is Verify (A/F mixture) and then LEAVE IT ALONE.

:-)

BTW, what is your car, what engine, condition of throttle shafts pump etc.

[304065]

Just to be clear, the pump is set to 40 degrees after #1 overlap TDC. So you first set the engine to #1 TDC on the COMPRESSION stroke, then rotate 360 degrees, then 40 degrees to the FE mark.

The adjustment of the pump sprocket with the three hex-head screws is so that the line on the pump sprocket lines up with the line on the pump housing.

Doing that with a mirror and shop light while hanging over the engine is enough to make you want to remove the decklid (pull the pins so you don't have to remove the hinges from body or lid) or remove the pump entirely, set it and then reinstall the pump. Have the official factory curse word* handy.



*The official factory curse word is: Himmel-herrgott-sakrament-zefix halleluja-scheissklump-verreckts! (Not suitable for German-speaking minors)

[JohnKo]

Thanks to you both for some great tips. The cars is a '70 911E with <50k miles, so I trust that things are mechanically in pretty good shape, but it's not had a lot of recent use.
I appreciate your attempts to clarify the pump timing process, because I labored over it many times to make sure I understood it. The timing is now correct, and yes, working on it in the engine compartment is... uh... challenging!
In line with the CMA methodology, I have confirmed that the microswitch, fuel cutoff et al are working as they should.
I'm actually getting as much or more exhaust backfire at higher rpm as on decel, and even some at idle. To me that suggests over-rich conditions. I'm thinking now about disassembling and cleaning the MFI thermostat components, the logic being that perhaps it's not fully allowing the system to 'revert' to 'normal' settings when everything is warm.
Any and all comments are appreciated.
Thanks, John

[Fast Corners]

Sorry I missed the backfiring on Accel in your original post. I would first confirm the rich condition by pulling the plugs. If they are black and sooty, it's running rich. The tail pipe is also a fair indication of this as well. It is important to differentiate between a rich idle and a rich high speed condition. The MFI pump has two mixture adjustments one for idle and one for under load. This as well as the adjustment process is well covered in CMA.

After reading through this thread again I wonder if the push rods are all adjusted properly. if the pump is sending fuel based upon throttle position and the correlation of the other rods is incorrect, the engine may be starving for air thus creating a rich condition. The protractors are rare so checking/adjusting the correlation is tricky without them. Start by verifying the center to center distance on the pump reglator to the cross shaft rod. This rod must be 114mm +/- 0.2 mm. I used to do this by spinning both ball ends to the same side of the rod and measure from the base of one ball socket to the top of the one on the other end and take the measurement. You need to adjust it to the correct length keeping in mind that when you spin one end to 180 degrees the the other side you have to add or subtract (which ever way you turn it to get it there) 1/2 of the thread pitch.

Now I have simplified this a bit. I start as above by measuring from the top of one socket to the bottom of the other one (when they are on the same side). I then measure the overall length of the rod from the tops of the ball socket housings. The subtract the difference. this gives you the number to subtract to get the 114 mm measurement center to center when the sockets are on opposite sides of the rod.

The thermostat only richens the mixture until the engine reaches 130 degrees IIRC. At that point the disks have expanded to the point where the actator rod has pegged the lever in the pump to the stop. I would save the disassembly of the thermostat until other items have been ruled out as being the problem. If your heet exchangers and the hot air feeder hose are in good shape your thermostat is probably working fine. Usually a visual inspection through the hot air inlet and exit ports will let you know if the disks are in decent shape or not.

[JohnKo]

Thanks, again.
It's finally starting to register. I'd been thinking rich mixture because of backfiring in the exhaust, but the plugs are decent and I never see any black smoke. The problem remains even after the engine is up to temp. (when the thermostat should be all finished). Leads me to other things, I suppose. I'll look at correlation. Also wondering if the PO screwed up the valve timing somewhere along the way (although compression is very even). The plot sickens! Any other thoughts on the cause(s) of random exhaust-side backfiring at idle, at part-load, and at higher rpm?

[mhackney]

JohnKo,

I had a similar issue with my 70S - backfiring after warmup, decel, etc. Started through the CMA and got to the thermostat check. I pulled and cleaned and tested my thermostat and it was working but the problem remained. Then I looked in the heater tube that warms the thermostat and this is what I found:



Yes, that is a dehydrated mouse. It was keeping the thermostat from warming up properly and leaning the mixture as it should at operating temps. Removed the mouse, and no more backfires.

Cheers,
Michael

[JohnKo]

Thanks! I know that tube is clear, but every little bit helps.
Regards,
John

[JohnKo]

UPDATE: I cleaned the thermostat and managed to remove some gunk and free up its movement; it seems to have improved the situation, but not entirely. Anyway, the warmer it gets, the better it gets. Feels good.

[304065]

Any competent Porsche driver tracks faster if brakes come smoothly.

Air filter, compression loss, plugs, dwell, timing, fuel pressure, injectors, belt, correlation, smog.

NEVER DEIVIATE FROM THIS SEQUENCE

Cleaning the thermostat is a good idea, do you have 25 pairs of discs? How many spacers? Remember WHEN IT GETS HOT IT GETS LONGER-- it's possible to fool the MFI pump into thinking it's already heated up by the introduction of additional spacers for testing purposes.

[Grady Clay]

Backfiring on trailing throttle can also be from a pin-hole exhaust leak. This can be at the muffler gaskets or head-to-heat exchanger gaskets or from the muffler or header pipes. The leak can be so slight that it is hardly detectable under acceleration.

A test method that sometimes works is to run the engine above idle with the hand throttle and restrict the tail pipe. Investigate with a stethoscope and open hose. Not infrequently it is necessary to remove the muffler and heat exchangers to summarily replace the gaskets with new and dress the flange surfaces.

While the exhaust is off, you can pressurize the parts and look for leaks. A common place for a muffler leak is at the lip that contacts the two ‘tangs’ of the engine mount muffler support. The common places at the heat exchangers are at the 3-into-1 junction (particularly the underside) and at the muffler flange.

One more piece of trivia for CMA2.

Best,
Grady

[JohnKo]

Thanks, guys. I appreciate the help.