Need some input -> Problem with 3.2 transplant going lean and misfiring after two min

[Wayne 962]

I did this yesterday and took a video of it but the video wasn’t that good. I’ll recreate today. Bottom line, the mixture does get richer but the car does not run any better.

Wayne

[ischmitz]

Wayne - that’s a strange result. I’m trying to consolidate the statements that changing the mixture doesn’t appear to help and your other finding that can be paraphrased as “different VE of bank 1 and bank 2”

I don’t quite understand how this is possible unless there is something severely different in how much air, fuel, or spark goes to the left vs. the right side.

You mentioned 90 degree intake flow (next to impossible, since at idle there isn’t much air flow to begin with), partially plugged intake or exhaust (maybe but unless you’d have a cat on each side not very likely)

How about this: someone swapped out a broken cam with one with a different grind.

[Wayne 962]

Originally Posted by ischmitz

Wayne - that’s a strange result. I’m trying to consolidate the statements that changing the mixture doesn’t appear to help and your other finding that can be paraphrased as “different VE of bank 1 and bank 2”

I don’t quite understand how this is possible unless there is something severely different in how much air, fuel, or spark goes to the left vs. the right side.

You mentioned 90 degree intake flow (next to impossible, since at idle there isn’t much air flow to begin with), partially plugged intake or exhaust (maybe but unless you’d have a cat on each side not very likely)

How about this: someone swapped out a broken cam with one with a different grind.

Indeed, very true. This car has me starting to cast doubt upon the measuring tools that I'm using, and/or starting to think that there may be something unusually wrong with the engine. Either that, or I've completely overlooked something obvious. But in 11 pages on this thread, I think something obvious would have been pointed out to me by now. I mean, it is still possible that the wire harness is slightly messed up and not driving enough current to either the injectors or the ignition. My gut feeling though is that particular problem would get worse as the engine revs higher (it doesn't appear to get worse).

From Wikipedia:

Volumetric efficiency (VE) in internal combustion engine engineering is defined as the ratio of the mass density of the air-fuel mixture drawn into the cylinder at atmospheric pressure (during the intake stroke) to the mass density of the same volume of air in the intake manifold.

It's certainly hinting like there's something potentially off left-to-right. Today I'm going to take the muffler off (again), and replace the bolts with a "quick release" system (studs and c-clamps), and drive it off to a parking lot (away from my neighbors) to measure the mixture out of the left bank and the right bank. I'll also redo my video from yesterday where I was poking the AFM flap with a wooden stick.

The most likely thing I can think of would be some type of fuel pressure drop through the fuel rails caused by a bad pressure regulator or bad pressure damper. My colleague who works at SSF mentioned that he was working with someone who had a 1988 BMW 5-Series that had similar issues. They replaced nearly everything, and then when they replaced the fuel damper, all of the issues went away. Since I have Steve W.'s spare fuel rails, injectors, regulator and damper, the logical thing would be to start taking a closer look at that. Or, I can just wait until Monday when the replacement 964 air fuel meter comes and try that first.

I think I'll measure left / right fuel mixture using the LM-1 and see how that works first.

-Wayne

[jjeffries]

Wayne,

As my Dad would have said, "Keep your chin up, lad!"

John

[Steve W]

Originally Posted by Wayne 962

It's certainly hinting like there's something potentially off left-to-right. Today I'm going to take the muffler off (again), and replace the bolts with a "quick release" system (studs and c-clamps), and drive it off to a parking lot (away from my neighbors) to measure the mixture out of the left bank and the right bank. I'll also redo my video from yesterday where I was poking the AFM flap with a wooden stick.

-Wayne

I would be easier to just weld another bung on the other header, and you can always retest in the future much easier, or even run two LM-1s. I can loan you one of mine if you want. The thread is 18x1.5, same as a spark plug. That's what we've done in the past and doesn't take that long. Make sure it's after the collector so all three primaries are sampled.

[Wayne 962]

Originally Posted by Steve W

I would be easier to just weld another bung on the other header, and you can always retest in the future much easier, or even run two LM-1s. I can loan you one of mine if you want. The thread is 18x1.5, same as a spark plug. That's what we've done in the past and doesn't take that long. Make sure it's after the collector so all three primaries are sampled.

I agree that would be the best solution. But, I don't know how to weld (bought a brand new Miller unit like a decade ago, and "learn to weld" has been on my to-do list ever since), and these are stainless steel which I hear is much more difficult. In non-Covid times, I would have just driven the car to the local muffler shop and had them weld it on. But it's not that easy these days.

I found a good parking lot to take the car to. Hopefully I won't drop the hot muffler on any important body part.

-Wayne

[firstclassof86]

Wayne,

I had similar symptoms on my '86 Carrera, which got progressively worse. Until the engine would not turn over. Car was running very rich, especially the cylinders on the left side( drivers). I am still in the process of replacing the regulator and damper(waiting for my parts). Found fuel in the vacuum hose that connects to the damper. Maybe you have a small hole in the diaphragm (regulator/damper) and for some reason, the driver side cylinders are getting more fuel from the intake than the right side ones(fuel from vacuum lines into intake manifold).

I must have a larger hole in one of the diaphragms, since car was smoking and fouling plugs.

also, the intake manifold was cold to the touch(very cold< even condensation ).especially on the left side. most likely from the fuel leaking into the intake manifold.

[Wayne 962]

Originally Posted by firstclassof86

also, the intake manifold was cold to the touch(very cold< even condensation ).especially on the left side. most likely from the fuel leaking into the intake manifold.

Cool deal, it's good to hear that my theory is on the left/right side imbalance is not completely crazy. I had not thought about perhaps measuring the temperatures of the left and right manifolds. Although I'm skeptical, having more data is usually not a bad thing...

-Wayne

[Steve W]

Originally Posted by Wayne 962

I agree that would be the best solution. But, I don't know how to weld (bought a brand new Miller unit like a decade ago, and "learn to weld" has been on my to-do list ever since), and these are stainless steel which I hear is much more difficult. In non-Covid times, I would have just driven the car to the local muffler shop and had them weld it on. But it's not that easy these days.

-Wayne

Chip Foose once said “I can’t weld but I can grind!!”

[Wayne 962]

So I took the car to a local medical plaza where I could run it without headers and will only puss off people that aren’t my neighbors.

I’ll post the videos soon, but the results are what I expected 1-2-3 is running okay and 4-5-6 is running leaner. So there is that disconnect between left and right.

Now the trick is to figure out if that is caused by a mechanical issue or a fuel injection issue. The compression check came back fine and I ordered a new leak down tester to replace the one that broke - it should be here tomorrow. So I will check that - compression and leak down.

Just sitting here leaning against a concrete wall waiting for the car to cool down to bolt up the muffler.

Wayne

[ischmitz]

Cams?

[Wayne 962]

Videos showing the mismatches side-to-side. I think this test here confirms what Steve W. originally predicted / figured out about 5-6 pages ago, that the air-fuel mixture coming out of the muffler was reading lean while 1-2-3 was reading richer, leading to the thought that 4-5-6 was running leaner than 1-2-3.

Video #1 - Cylinders 4-5-6 (left side) - idle, running quite lean. The car goes super lean and then stumbles and then recorrects. Not sure how important this clue is.



Video #2 - Cylinders 1-2-3 (right side) - idle, running close to 14.7. This is what we've seen when the O2 sensor was plugged into the 1-2-3 side header previously. Since I tuned the airflow meter mixture to be 14.7 when the LM-1 wide-band O2 sensor was plugged into the 1-2-3 header, this makes sense.



Video #3 - Cylinders 1-2-3 (right side) - idle and about 1500-2000 rpm. The air-fuel mixture on this side is very good / close. You can hear the car sputter and miss though as this is revving.



Video #4 - Cylinders 4-5-6 (left side) - idle and about 1500-2000 rpm. Same thing as when at idle - runs much leaner than the 1-2-3 side.



-Wayne
-Wayne

[ischmitz]

So in summary left side sucks at idle (lean) and all else is well?

[Wayne 962]

Left side runs lean in 1500+ as well...

Although this looks bad, it's getting me closer to figuring out the cause:

- Fuel (blockage in the lines?)
- Air? (blockage in the runners?)
- Mechanical / timing / compression (I'm not sure how likely that is1)

I think ignition is pretty much excluded as a potential culprit, as even a missing / dead cylinder wouldn't necessarily cause the mixture to change - the O2 sensor doesn't detect unburned hydrocarbons (HC).

-Wayne

[ischmitz]

Originally Posted by Wayne 962

Left side runs lean in 1500+ as well...

-Wayne

Hm, not easy to see in the videos. Too me it looked like only one of the idle video showed 18+ while the rest was inline around 14.

- Have you had the intake off to check the gaskets?
- Any chance one of the injector o-rings got damaged during install?

[scarceller]

Your warm running issue is not likely the CHTs given your numbers above. Once fully warm if the CHT reads below 200ohms then cold start enrichment is no longer used. You said that after 10 minutes you read 155ohms, it's not the CHTs.

Unplug the O2 sensor, get engine fully warm, what's the AFR at? You need to know if you are lean or rich. A properly setup 3.2L should be at 13.8 to 14.4AFR fully warm with O2 disconnected. This engine does not like to run lean at idle, if it's lean it will result in lean mis-fires and drive the WBO2 nuts! You need to set base mixture at around 14.0AFR and leave the O2 sensor unplugged.

Other thing I noticed is the extra 90deg rubber elbow going to the AFM, this has me concerned. Can you remove that elbow and put the AFM onto only the first elbow, put it as if it were in a normal 911 configuration. Not sure why that extra elbow was added? I've helped on a few 914 conversions and none have that crazy intake setup.

RHS vs LHS mixtures not the same usually means bad injectors, swap injectors between RHS to LHS, does the issue move to the other bank? Just swap 1,2,3 with 4,5,6 and test bank mixtures again.

[Wayne 962]

Originally Posted by ischmitz

- Have you had the intake off to check the gaskets?
- Any chance one of the injector o-rings got damaged during install?

Those are good suggestions, so to double-check, I ran another smoke test this evening (4th one). I also pulled the injectors out of their mounting holes just to confirm with my eyes that smoke was in the chamber (it was). No leaks appeared there. I also presume the intake gaskets would be fine too if there was no smoke leakage there either...

-Wayne

[Wayne 962]

Originally Posted by scarceller

Other thing I noticed is the extra 90deg rubber elbow going to the AFM, this has me concerned. Can you remove that elbow and put the AFM onto only the first elbow, put it as if it were in a normal 911 configuration. Not sure why that extra elbow was added? I've helped on a few 914 conversions and none have that crazy intake setup.

Thanks, that's a good suggestion - I did remove it tonight. Injectors are all rebuilt and I've been swapping them around a bit. Steve W. lent me some used one that are old, but may still give good clues as to what is going on. I swapped out the fuel damper tonight with the one that Steve W. lent me - I'll have to test it in the morning to see if there is any difference. Changed the 2-year old fuel filter too while I was in there.

Static pressure with just the pump running (hotwired from the DME plug) is about 40 psi on my gauge, which may be high. I have a brand new regulator I may swap in since this seems slightly out of spec.

-Wayne

[Wayne 962]

Okay, I installed the pressure damper loaned to me by Steve W., and I removed the 2nd boot on the intake - no change. The car behaves exactly the same. I'm going to recap all that has been done so that perhaps it might be come obvious what the issue is. I also might get some ideas too.

Summary: The car (seems) to run well (slightly erratic idle) when started cold (I have numerous videos of it starting cold). Then after about two minutes is just starts to miss and flub. Revving the engine will work, but the car will miss when revved. When driving, power feels down, and when there's the miss, there's a matching dull thud clunk coming from the engine. My gut is that it is not a mechanical engine issue, otherwise it would run poorly on code start up - I cannot really think of a mechanical mechanism that would cause the engine to run fine for 2 min and then just take a dump like this is.

I'll spend some time on a recap post today, it will help clear my mind too.

-Wayne (seriously bummed)

[Steve W]

Originally Posted by Wayne 962

I think ignition is pretty much excluded as a potential culprit, as even a missing / dead cylinder wouldn't necessarily cause the mixture to change - the O2 sensor doesn't detect unburned hydrocarbons (HC).

-Wayne

Actually this is incorrect. A missing cylinder will cause that cylinder to act essentially as a straight air pump, pumping all that O2 that came in from the intake out to the exhaust. That extra oxygen is detected by the O2 sensor which causes it to sense that car is running lean. You can see missing on a wideband as spikes that go lean (> 18-19:1). Excessively rich mixtures (> 9:1) can also fool a wideband into displaying a very lean mixture because the excessively rich mixture fouls the spark plugs causing missing. Usually in such instances, that can be detected by excessive hydrocarbons that you can smell and if at a standstill, burns your eyes and nostrils.