MFI airflow question

[logician]

When I measure the airflow on my '73 911S (MFI) with a synchrometer, I find that the airflow at each of the left stacks is pretty much 75% if the flow at the right hand stacks. So for example, at 2700 RPM, all three right hand stacks read 16, while the left hand stacks read 12. At 730RPM, the left hand stacks read 3.5, while the right hand stacks read 5. The left hand mixture screws are all set at approximately 3/2 turns, while the right hand ones are set at approxmiately 1 turn. Another observation is that the left hand stack -- the one with the lower airflow -- seems to pop at idle and also at 2700RPM.

So my question is what might cause this difference between all of the left stacks vs. all of the right stacks? The throttle bodies are separate, each with separate stops, so you would think that any error would apply randomly to the cylinders, and not apply equally to all of the cylinders on one side. The throttles are all on their stops at idle, so I don't think it's that one side is getting more throttle than the other due to the linkage from the cross shaft.

The one thing that all of the cylinders on each side have in common is the cam. If the cam is timed wrong, could that cause a 25% difference in airflow? Any other ideas? Thanks in advance for any help.

-Juan

[avendlerdp]

Do a compression test and if you find all the cylinders on one side low then it could be a cam thing. Otherwise look for worn out linkage parts on your MFI stacks.

[Jim Richards]

Juan, it sounds like the links that connect each bank of stacks/throttle bodies to the cross bar are not set quite right. There's one link per side. I'm seeing this airflow discrepancy on my engine, too. I'm going to balance mine out this weekend and see how it turns out. Of course, I will measure and record the lengths of these links before making any changes in case I make things worse!

[jstobo]

You do not have mixture screws on the throttle bodies, those are air screws for each intake. They should be adjusted so that you get an idle of 950 RPM with even air flow through all stacks. The mixture is controlled from the MFI pump. Sometimes carbon builds up in the air passages and changing the air scews have no effect. If that happens, remove screw on offending cylinder(s) and spray Chemtool B-12 liberally into the passages. Be careful, this stuff eats paint. If you have set the idle with the screw on the microswitch you need to readjust per the manual. Download Check, Measure and Adjust from the tech articles site here at Pelican. It will tell you how to adjust the air flows and linkages if they need to be adjusted which I doubt unless someone was in there before you. Hope this heps.

Jim

[logician]

Hi Guys,

Thanks for the replies.

Originally posted by Jim Richards
Juan, it sounds like the links that connect each bank of stacks/throttle bodies to the cross bar are not set quite right. There's one link per side.

What I'm seeing has nothing to do with the linkage. The throttles are all on their individual stops at idle.

Question: I start out with an air screw set to 3/2 turns and see an air reading of 5 at idle. Then if I close down the screw completely, how much should the air flow be reduced? Does all the idle air come through the adjustment screw?

-Juan

[Peterfrans]

The airflow at idle is both through the air screw and through the gap at the throttle valve (and depending on condition of the throttle bodies also past the throttle shaft). If the throttles cannot be balanced any further with the air screws, you could attempt to adjust with the throttle stop screws (for instance all throttle stops screws on the right hand bank a 1/4 turn out). In this case you would also need to re-adjust the links. Be methodical and note down the changes that you made, to be able to back-track

You probably already went through the CMA and other relevant MFI stuff on the tech section, if not that is a good place to start.

[Grady Clay]

Juan,

In addition to a cranking compression test, I would measure the actual cam timing and check the valve adjust on both sides.

Are all six throttles sitting on their stops? Same when you connect the linkage?
Do they come off the stops at the same time?
Is there any radial play in the cross shaft? (Missing bearing?)
Are the idle air passages in the throttle bodies clean?
Does it appear as if the throttle bodies are different from each other?
Are there air leaks between the throttle body and head? Between throttle bodies and stacks?
Is the vacuum retard connected? Leaking?
Are the six cold start nozzles in place? Hose connected?

You can check to see which side is lean by popping a rod off one side and slightly opening that throttle with the center linkage. The side that is more lean will pop and slow, the richer side will speed up.

Any chance it could have different cams L-R?
Do all the heads have close to the same casting date?
Are the heat exchangers the same?
What kind of muffler?
What do the plugs look like L-R?

Best,
Grady

[logician]

Grady, Thanks for the reply. I'll check the things that you suggest.

Regarding lean condition, I notice that the left bank is popping (out the stacks) in its current state, even without opening the throttle on that side. The poping seems to come from cylinders 2 and 3, and happens all the way from idle through 2700RPM. The thing is, that's the side with the lower airflow (richer), 12 on the sychrometer, vs 16 on the right side at 2700RPM. Any ideas on what that could mean?

-Juan

[304065]

Well, the throttle plates are not REALLY "at their stops." Somebody has monkeyed with the ones on the right, because they are easily accessible, whereas, the throttle plate stops on the left (port side) bank are inaccessible without a special tool.

You might try removing the throttle bodies and making up a "flow bench" consisting of a shop vac with appropriate reducer and seal on one side, and the synchrometer on the other. Or you could send them off to be flowed by Matt Blast at Eurometrix, or Henry Schmidt. They will come back with the linkages adjusted and marked with witness marks (paint on the shafts) so they are all pulling equally, then the fine adjustment is set with the air bleed screw. (Which technically IS a "mixture screw" insofar as it controls the air portion of the air-fuel mixture, but I know what you mean).

I also like Grady's idea of checking the cam timing.

[logician]

I disconnected all the throttle linkages, and that doesn't change the airflow at all. So the problem has nothing to do with the linkage.

I tried closing down the air screws altogether to see how much air flows past the butterflies. At 700RPM, the left hand side allows 2.5 units of air, while the right hand side allows 5.0. So it would seem that all of the butterflies on the right side are open a lot vs. the left bank.

So I'm starting to think that the two banks are simply calibrated differently. One thing though, the throttle screws seem to be set the same between the two throttle bodies. I used a feeler gauge to measure the distance between the head of the screw to the locknut. So it doesn't look to me like someone went in there and messed with the screws. Maybe the difference in airflow is due to a difference in the construction of the levers.

Question: If I adjust the left hand screws to 4 turns I could get the same airflow as the right when the right is set to 1 turn. Is that a reasonable adjustment? Or would that be too much of a difference between banks, and might that cause other problems?

-Juan

[jstobo]

I don't think the amount of screw adjustment side to side is important. Throttle bodies will wear differently, carbon buildup may be different etc. The important thing is to have the air flow balanced. Appears that you can do that and keep idle where it belongs. So adjust away.

[304065]

Juan, if you disconnect the throttle linkages and the airflow doesn't change, that just means your throttle linkages were set correctly.

I still think the throttle stops are set differently on the right-hand stack. Try removing the stacks with the linkages removed, and use your feeler gauge between the edge of the throttle plate and the stack, at a consistent position (like the 6:00 position, closest to the rear of the car). If you measure a larger gap on the starboard throttle plates, that confirms that the plates have been set more open on that side.

Once the air bleed screws are closed, there's just no other explanation than that the plates are open a different amount, or the throttle bodies are more worn on one side than the other. Unless somebody stuck a smaller gasket between the throttle body and the cylinder head.

[logician]

John, I agree that the throttle plates are open a different amount between the sides. It just appears to me that the throttle stop screws are set very similarly between the two sides. I'm thinking maybe the hardware is slightly different on both sides, so the same screw setting achives different degrees of openning. I suppose another possibility is that one side is rebuilt and tight, while the other side is leaking. I know that the MFI pump has been rebuilt, so the system has been reworked at some point.

Why would the gasket thickness affect the throttle stops? That sounds like an interesting possibility too.

Actually, one last check that I'm going to do is measure the vacuum on both sides to make sure that isn't the cause.

One other question. It seems that the side that is poping is the richer side -- the left side with the less airflow. Does that make any kind of sense? I thought it was the lean side that is supposed to pop.

-Juan

[304065]

Juan,

What I meant about the "smaller" gasket is that someone could have fitted a gasket with a smaller diameter than the "s" throttle body opening, either between the stack and the throttle body, or between the throttle body and the head, that is protruding into the airstream.

Can you get the stacks to even by keeping the right side screws mostly closed and the left side mostly open? That would seem better than disturbing the throttle stops on the starboard stack.

In my own case, they are all over the place, and nobody ever told me NOT to move the throttle stops until I had already monkeyed with them. You COULD close the throttle stops all the way down on all the stacks and balance the right side by loosening the lock nut and closing the throttle plates all the way. Some counsel that this might cause the plates to stick in the bores when the engine heats up, but I've never experienced it.

At the end of the day, if you have a stable 900 rpm idle, no problem.

[logician]

So I measured the vacuum, and I find that at 2200RPM, the vacuum on the left side (lower airflow) is 8 inches of Hg, while the vacuum on the right side (higher airflow) is 12 inches. That reduced vacuum is consistant with the reduced airflow on the left side. What that tells me is that the difference in airflow is not a result of anything different in the throttle bodies, but rather due to something different in the engine that causes the left side to suck less than the right side. Strange.

I'm not inclined to think it's due to a blockage in the (stock) exhaust system because it would have to be a huge blockage to affect the flow at idle. The engine seems to run well under load.

I did check the compression a few years ago, and I while I didn't write down the results, I didn't note anything unusual at the time.

Any ideas? Is that difference (33%) in vacuum between the two sides unusual? I still note that the left side is poping during the tuneup procedure, even though that is the richer side.

-Juan

[avendlerdp]

Still think you should run a comp test. If you have a cam timing issue it will show up as well as any other problems. Start with the basics so you know what's up, then move to the next issue.

By the way, are the left and right T.B.'s interchangeable? If so you could swap them and see if the problem switches sides.

Alex

[davidppp]

Right..If Ive got this right..

One one side the flow is low, and this same side has a lower vacuum?

This almost certainly means the cam timing is different.

May have jumped, or just set wrong by the assembler..

It is quite quick to check with a clock guage..

Kind regards
David

[logician]

Originally posted by davidppp
Right..If Ive got this right..

One one side the flow is low, and this same side has a lower vacuum?

Yes, exactly. Which side would be suspect, the one with the lower flow, or the one with the higher flow?

So when you say check it with a clock gauge, you mean just time the cam -- if the lift isn't right, then you know something is fishy? Is there a way to measure the profile of the cam? Given the screwy problem, I'd like to also check that the right and left cams match.

-Juan

[davidppp]

Funny enough, I do not know which might be out!

At idle/slow speeds, a wrongly timed cam could likely pull more air..tho' I'd guess you have E cams, which should be getting efficient at 2200rpm..

Yes, Its certainly possible to check the cam timing and profiles in situ..but tedious..

Kind regards
David