Finally getting around to CMA on my MFI

[304065]

Declan, that's a pretty good cleaning job, bravo!

Now, take a pipe cleaner and run it down in the air corrector screw holes to be sure they are clear. Look at the screws-- if some DPO has buggered the tips they won't give you a good seal. They should be straight before being reinstalled.

For my own TBs I made a home-made flow bench with the shop vac I have for cleaning the interior of the car. It has about a 1.25" hose which I wrapped with duct tape so it would seal inside the throttle body on the underside. The STE synchrometer fits in the high side. Start up the vacuum and hold it tight so everything seals, you can see how many Kg/hr of air are passing the plate. Start with the air corrector screws all the way closed. Once you set the plate stops so they are all flowing LESS than 3Kg/hr, you can then balance them out with the air corrector screws.

The reason I say LESS than 3Kg/hr is because that reading is JUST off the closed position-- you don't want the plates resting against the magnesium because wear will be accelerated and they could stick in the bores-- but you don't want them too far open as you will pass too much air and have too high an idle. If you have too high an idle (with the distributor advance dialed back to TDC at idle so advance doesn't push it up) at least the good news is that with your '72- style throttle stops it's fairly easy to loosen the lock nuts and reset the plates to the more-closed position (but you lose the luxury of doing this off the car, you have to do it with the engine running.)

So, less than 3Kg/hr with your home made flow bench should do it, remember you can always adjust with the air correctors. NOW. . . set the rods so they are EXACTLY on the balls with no grease on the cups. Check with flow bench again, nothing should have changed. If any cylinder shows too much air, you had that rod set wrong, start over. Now grease the ball cups. Check again. Still all the same? Great, now put paint on the throttle rods so they don't move and you KNOW they are set. Then put the TBs back on the car with NEW gaskets top and bottom. Use M8 nuts that are 12mm ATF (across the flats) these are easier to get on and off with a swivel socket, and stainless wave washers. Do NOT drop any hardware down the intakes, have a magnet handy for when you do! Good luck!

p.s the TBS in the picture were just professionally rebuilt and set with a real flow bench. They will not be adjusted when on the car- only the air corrector screws will be used for fine-tuning, maybe only half a turn here or there. The procedure I describe above did work for setting the stops, however.

[declang]

Originally Posted by john_cramer

Use M8 nuts that are 12mm ATF (across the flats) these are easier to get on and off with a swivel socket, and stainless wave washers.

Aha. I thought the 13mm nuts were a pain to get off. I had to just loosen them about half a turn with the socket and do the rest by hand.

I had to google "wave washers". Never heard that term before.

Is there any particular type I should get ?
And, (to save me measuring wrongly) what thread size are the nuts ?

Last question for now, any tips for cleaning the stacks, or just get the
toothbrush out again ?

Thanks,
Declan

[304065]

Declan, every M8 nut in the car is an M8x1.25. In non-engineer speak (I'm not one either) that's an 8mm metric thread with a 1.25mm pitch. The special 12ATF nuts are commonly used in the exhaust system (where they are an all-copper lock type, a BMW part number) or can be found from specialty suppliers. I THINK the 12mm atf is a Japanese spec? Like you might find on a motorbike. Our host can certainly get these for you. Ditto the wave washers, standard M8 size, these conform to DIN 137B (a German spec), any Wurth dealer should have them. For the pundits who say, "use aluminum" I say "you better enjoy the taste of chewing gum, because you will be chewing a lot to put on the end of a ratchet extension to retreive those non-magnetic aluminum washers!" (Stainless should be SLIGHTLY magnetic, enough for you to fish it out of the (hopefully valve closed) intake port WHEN (not if, we all do it!) one drops down inside!

Stacks are easy, get a bottle brush. If you use water, be sure to dry them thoroughly, as magnesium corrodes easily enough, in the presence of water looks even worse.

[declang]

Thanks for the info John.

I'm missing one of the screws for the plate stops !!
I can't be 100% sure if it was there or not when I took the TB off the car. I made apoint of checking the 4,5,6 but am not 100% sure if I checked the 1,2,3 TB or not. I guess these are M4 ?
I still might find it in the drain-pan I was using for the clean-up operation if it came out during cleaning

I've just ordered some bits from our host - gaskets, distributor cap, etc, and found a supplier for the stainless wave washers over here, so by the time I'm finished cleaning & adjusting ( & being divorced for using the vacuum cleaner on "that old car" ) , I should be ready to assemble again.

I ordered a dial gauge & z-block as well, so should be able to get an idea of what cams are in there.

Is there a recommended plug I should be using, for my temperature zone ? I've read a lot of the threads on here about plugs, but still unsure what would be best for my climate here in Ireland

Thanks again to all for the help so far. I'm enjoying all this learning

[Jeff Higgins]

The NGK BP7ES has worked well for me in both my 2.4 T and my 3.0, both running MFI. Our Puget Sound climate is much like yours; kind of cool and drizzly much of the year. An average summer might see a few days running up into the 90's; and average winter might see low 20's to teens for a week or two.

[declang]

I had a go this evening at John's flow bench system, using the vacuum cleaner.

The bypass screws were in pretty good condition, just a little carbon build-up but I had them steeping in brake cleaner overnight, so it scrubbed off easily enough.

I could get all but one flowing less than 3kg/hr. Cant remember if it was #1 or #3
Best I could get was 3.5, so I set them all to 3.5kg/hr.

Hopefully I'll get some more time tomorrow evening to put the linkages back on & ensure they dont change the airflow.

Today I managed to order a decent timing light, and am on the hunt for am LM-1 next, since our host stopped selling them

[304065]

Declan, OK, but prepare yourself to remove them and reset to lower than 3.5. The concern is that you will have too much air passing the plates at idle even with the air correctors dialed all the way closed, resulting in the classic "2000 RPM IDLE" so common among MFI enthusiasts.

I would see if you can't figure out why that hole is pulling more air than the rest. Carbon buildup blocking the plate open? Air leak out the throttle shaft hole (likely!) What is the flow with the throttle stop screw all the way out of the hole and the spring holding the plate shut?

Wayne stopped selling the LM1? Titan offers them for a special price of $299. http://www.titanmotorsports.com/lmdiairame.html

[declang]

Yeah, as I typed the last post, I was thinking I should investigate further why one was higher !

I could get it to 3 or slightly below, but then it seemed to bind/catch when moving it off the stop, which I didn't like, so I found that at 3.5 it was much smoother .

I thought I had them pretty clean, but mabye there is some carbon I missed in close to the butterfly shaft. I'll try & investigate further tomorrow evening.


I've found a few sources for the LM-1. Should get one ordered tomorrow I think.

I guess our host is preparing for the introduction of the LM-2 to replace the LM-1

Cheers.
Declan

[Grady Clay]

Declan,

This is a situation where you need to carefully investigate ALL the engine details. Pulling more air (fuel?) from a particular cylinder can have many causes – some not very desirable.

This is where multiple tests for cranking compression, cylinder leakage, valve timing & clearance adjustment, vacuum leaks, imperfect injection nozzles, fuel pressure, MFI pumps, an ignition issue and much more can lead you down the path to confused diagnosis. This is the major reason everyone familiar with MFI is so determined (anal?) following CMA and the geriatric extensions. Without these (and more) tests and extensions, it is VERY easy to miss something simple and be lead down the path to complicated (expensive) routes that don’t solve the problem.

I probably have done and supervised way over thousands of Porsche MFI engines. I think I still own around 20 (resting quietly). I can’t tell you how many times I (and my employees) were ‘fooled’ by skipping some simple test because we KNEW it was OK Lessons learned.

There is a great deal archived on Pelican (and a few others). A search on CMA and CMA2 will yield some important links. It is critical to have both the CMA booklet (4532.20 & revisions) and the Factory Workshop Manual pages (Vol I, SF 1-45).

While MFI was relatively complicated when new, today we have the issues of ‘geriatric MFI’ and the engines with similar conditions. None of this makes it easier to diagnose and repair these systems. These were (and are) ‘state-of-the-art’ race systems of the day. They deserve to be maintained in that status.

Best,
Grady

[declang]

Quick Update.
I had another good look and I did miss some carbon in #3

I now have 2.5kg/hr on #1,#2 and #3.

While I was investigating I noitced that by opening the bypass screw it goes up to amost 7kg/hr then drops off to around 5 with the screw toally removed (sucking air in the actual screw hole).


I still have not found the missing stop screw, and still am not sure there was one in it at all !
(I keep thinking, surely I would have noticed something so obvious, but mabye not. I never had reason to adjust them on the car, and it's #3, so hardest to see when on the engine.)
I made up another screw to suit.

Declan

[Grady Clay]

Declan,

This tells me you need to remove the throttle bodies off the engine and see what you have. First I’ll encourage some ‘sporting’ driving and repeat the cylinder leak tests.

Best,
Grady

[declang]

Grady,
The figures I'm talking about are with the TB's off the car, using John's vacuum cleaner method.

Simply vacuum cleaner in bottom of TB and synchrometer in top

I was trying to achieve John's figure of sub 3kg/hr

Does it make any more sense now ?

[304065]

Declan, that sounds good to me. As a final step, verify that at 2,5Kg/hr the throttle plate IS being held open by the stop screw. The way you would verify that would be to slightly back off the screw and watch the reading fall. If 2,5 is the absolute bottom you can see, then you might turn the stop screw a tiny, tiny bit to ensure that the plates aren't all the way closed.

Then verify settings with the rods ungreased, then greased, then seal with paint, then install. The time spent getting the readings equal will translate to easier correlation and better power balance when the engine is running.

Declan, one more thing. . . Grady is the one who gave me the idea for the "home-made flow bench. . ." he built one years ago using a vacuum pump and some manometers. . .

[declang]

John,
2.5 is absolute bottom. If I back off the screw, i can push the arm further than the spring will take it and get down to under 2, but it feels like I'm forcing it beyond what I should. Not that I'm applying much pressure, just more than the spring can provide.

Would I be better to up them all a little, to just a shade below 3kg/hr ?

I still have to set the other TB set, its still on 3.5.
Continued thanks

And, Grady, Sorry for "over-explaining" the "home-made flow bench" bit.
Fantastic Idea, and so simple.

Cheers,
Declan

[304065]

Declan,

Unfortunately it's not possible for me to say "2,89 Kg/hr is the perfect flow reading for a 900 rpm idle" because I haven't yet flowed my own rebuilt (above) throttle bodies on the bench or on the car. So what it translates to in terms of an idle reading, I'm not sure. What I am sure of is that if you get them reading the absolute minimum you can with the plates BARELY open, you can always adjust the idle higher with the air corrector screws if it needs it.

3Kg/hr times six is 18Kg/hr, the fuel consumption spec is around 225 g/HPh or 225 grams per horsepower/hour, so 18,000 /14.7 (stochiometric) = 1,224 / 225 = 5HP. So the engine puts out around 5 hp at idle. Actually, the way you would really do the measurement is figure out the horsepower at 900 rpm, fuel consumption and air/fuel ratio and divide by six. But this kind of stuff only happens in the lab. . .

[declang]

John,

OK, I thought maybe the 3kg/hr figure was lower than the required airflow for a good 900rpm idle, meaning the additional air needed would be supplied by the bypass screws, and the "number of turns out from closed" bit from CMA.

Any ideas what the airflow per stack would be at idle ?
I'm expecting it'll be between 3 & 7, since I got 7kg/hr with the bypass screw fully open. (Hope my logic is right !)

Or, is this where the "geriatric MFI" bit comes into play ?

Slainte,
Declan

[declang]

In my continuing "tool buying" mode, I've bought a set of circlip pliers, another set of combination spanners, some torx bits, a decent timing light with advance control, an innovate LM1 with the RPM kit, a leakdown tester (still unsure about this*), a dial gauge & z-block, etc etc and now I realised I haven't got a compression gauge/tester.

Is there any particular type/make of compression tester I should get, or are they all pretty much the same ?



* Leakdown tester : I have a Gunson 77005 Cylinder Leakage Tester,


http://www.gunson.co.uk/item.aspx?item=3441&cat=666 but I'm unsure if I'm measuring correctly. With #1 at TDC compression, gauge set to zero leakage, plug in the hose from ppark plug hole & get about 5% or 6% leakage.

Where I'm a bit lost is that I cannot see any way to zero the leakage gauage at anything higher than 1 bar.
I simply have the air supply from compressor directly onto the leakage tester. Mabye I should have a valve of some sort to stop the air supply & seal the gauage ?

If I leave it set at the 1 bar, and reading about 5% leakage, then turn the engine by hand, I can open a valve, hear the air escaping, and see a big value of % leakage.
That makes sense.

I just cant help thinking I should be able to do this at a higher pressure, so must either I'm doing something wrong, or my tester is faulty ?

Any advice ?

Thanks,
Declan

[304065]

Declan,

I have the same gauge. I, like you, originally thought that I would use 100 PSI (makes the math easier) or 80 PSI (the old aircraft standard) but in practice I wasn't able to get more than about 35 psi. If you turn the pressure relief valve higher, then you can't zero the leakdown gauge. This is the identical "problem" you are experiencing.

In truth, it's not really a problem at all, the gauge gives very accurate leakdown numbers using the lower pressure. Some (Grady) advocate using very low pressures anyway, because 1) less tendency for the engine to rotate off TDC, opening a valve and throwing the reading off, and 2) the low pressure will develop leaks that would be sealed by higher pressure. Weighing against that is the old rule of physics that says that the airflow through an orifice is directly proportional to the size of the orifice and the pressure behind it. Maybe throw a square root in there somewher. Anyway, the bigger the hole the more the flow, but given the internal violence of an internal combustion engine, there's rarely such a thing as a small hole. e.g. leakage around the rings from normal wear shows up as 3-5%, BROKEN rings show up as 45%. See, for example, this thread: Leakdown Results 1966 911 Normal

Anyway, nothing is wrong, proceed as normal. . .and post your leakdown figures!

You may want to consider using a piece of tubing, plastic or metal, around the 14mm hose that threads into the spark plug hole-- this makes it easier to wind in there. Also, try spinning the hose with one hand and guiding it in with the other-- you'll get the hang of it eventually. Finally, I use a "whistler" compression tester that's been cut off and the hose slips over the small quick-disconnect fitting that goes into the tester-- makes it much easier to identify TDC on the compression stroke. Then again, if you see 100% leakdown, rotate the engine away from TDC OVERLAP. . . and hold the Guinness till the job is done!

Good luck!

[declang]

OK, Thats what I wanted to hear
Thanks John.

Will post the figures when I get all 6 tested.

[declang]

And today I.....












did nothing !

I ordered a compression tester, so I suppose that counts as something constructive !



Declan