[MFI] One cylinder much leaner than the rest?

[422flat6]

My 2.4L MFI motor is running very rich. I'm working my way through Check Measure Adjust and I'm making good progress, but I'm not ready to adjust the mixture screws on the pump yet.

It was fouling plugs pretty rapidly. I found that the cap and rotor were badly burned, showing some pretty gnarly carbon tracks. I swapped those out with new ones and it is no longer fouling plugs. It's still too rich, but it's stopped fouling plugs and I'm not worried about getting stranded anymore. I've been watching the color of the plugs, and (unsurprisingly) they've all been very black up until I swapped the cap and rotor. Now with the new cap and rotor, plugs 1-5 are pretty evenly black with carbon but #6 is bright white! I've attached a pic of #6 after about 200 miles of combined city/freeway driving. Clearly this needs to be sorted out first.

I had assumed that all cylinders were running about the same degree of way too rich since all plugs were about the same color before, but perhaps the bad cap and rotor were just masking an issue with #6. I need some input on what to do next. I already tried swapping #5 and #6 wires thinking I might have damaged the wire with all the spark plug changing I've been doing, but it appears that isn't the case.

Here's what I've done so far, listed in CMA order:

• Air cleaner cartridge: clean

• Compression: 190-195 psi all cylinders (9.4:1 compression ratio)

• Plugs: NGK BP7ES gapped at 0.60mm with feeler gauge, tried another new one in #6 and it's just as white... that seems to rule out the slim chance of a defective plug

• Wires: appear to be new (Beru), swapping #5 and #6 has no effect on plug color, #6 electrical continuity checks out and resistance is 1440 ohms which is the same as #4

• Dwell: 39-40 degrees, meets spec of 38 +/- 3 degrees

• Timing: 5 ATDC with vac line on at idle, 6 BTDC with vac line off at idle, 30 BTDC at 4000 RPM. I think the distributor has been recurved to account for higher than stock compression ratio

• Fuel pressure and flow: not checked, need to buy the tools

• Injection nozzles: not checked, would need to be sent out

• Injection timing: spot on

• Correlation: I don't have protractors, but per the P250 procedure my pump rod is 114.5mm (yeah, out of spec by 0.3mm) and my left and right throttle links are within 3mm of each other. All rods installed at no preload

• Synchrometer readings: 25 kg/h on cylinders 1,2,3 and 6 at 3000 RPM. 18-20 kg/h on 4 and 5 at 3000 RPM. I plan to take the throttle bodies off and check and set blade position and stops using the vacuum cleaner flowbench method, then check individual link lengths. I doubt this is contributing to my spark plug issue.

• Emissions test: not done, need to buy the tools

The motor was rebuilt in December 2014, bottom end included. Supposedly the MFI pump was rebuilt but I have no proof of that. I'm running 91 octane California ethanol fuel. I've cleaned the thermostat and it has 25 pairs of expanding washers in it plus one thick and one thin shim. The thermostat is getting hot air from the heat exchanger, fully leaning the pump out at operating temperature and working well. The microswitch and speed switch are hooked up and functioning properly. The throttle linkage seems reasonably tight. The throttle bodies have been recently rebuilt and have new bushings. I did try pulling the #6 wire off with the engine idling and there was a small RPM drop, but unfortunately I didn't measure it to report exactly what it was.

The cold start enrichment solenoid is hooked up as per stock configuration. I intend to put it on a push button switch, but have not done so yet. The car won't start cold without it hooked up. I haven't tried to start it hot with the solenoid unhooked or just unhooked the solenoid after the engine is running. I might try this next to see if maybe it's continuously weeping fuel into the velocity stacks. I did test it with +12V and it makes a positive thunk-thunk. This seems like a long shot though.

I'm thinking I should do the fuel pressure and flow check and then get the injectors tested. Or maybe check the #6 fuel line between the pump and the injector for a clog? I put a bottle of Techron fuel injector cleaner in the tank yesterday.



I haven't started a proper introduction thread since I don't have many good photos of my car yet. I've been so busy learning and trying to get everything sorted. I'll try to get some soon and post another thread for an intro. Until then, here's a photo:

[lateapex911]

Great car and color. (I have a similar yellow version) And welcome!!
I'm not much help, I'm afraid. I USED to have MFI on my 2.4, but back in the 90s ended up with a 2.7 and Webers, (Now PMOs) Killing myself over that blunder.

Anyway...you have good methods. You mentioned disconnecting #6, and noting a slight drop in rpm, but not recording an amount. It's all tough to do alone. A quick check to see if it's running equally is to just reconnect and pop another one off, and compare.

It sounds like there are injector issues..I'd think that it wouldn't be a line blockage...since the injector is the bottleneck, so to speak. I'd consider a quick injector swap between 6 and another cylinder. If the problem remains the same, you know the injector isn't at fault.
Once you determine that, concentrate on finding out if indeed the cylinder is getting less fuel. Swapping or substituting the line would be next.

If thats all fine, then.....uh oh, it's the pump.

[MDamen]

Sounds like maybe the #6 injector might be clogged a bit and not delivering as much fuel as the other injectors. The plug in the picture looks brand new. Hard to tell that there is anything wrong. Generally it should have some tan deposits. Can you post a picture of the other plugs for comparison?

[BURN-BROS]

Good looking car.

With the engine idling, disconnect the aluminum connecting rods to the throttle arms.

Then close the air bypass screw at #1, check air flow. Then turn bypass screw back out before moving on the next cylinder.

What are your sync readings for each cylinder?

If the fuel lines were re-plated, then there can be debris and blockage in the line or now in the injector(s). Remove the injector and look inside for foreign material. The injector has an internal screen that prevents material from getting into the nozzle. Test the hard line for blockage.

[patz]

If you have the ability to measure injector flow, do so. Might be as simple as dumping in a bottle of Techron.

[David E. Clark]

I'd start by confirming that you don't have a bad plug - rare but it happens. Swap another plug into #6 and see if you get the same results. If the results are the same with the second plug, you might have a blocked or leaky injector. Remove the injector and give it a good ultrasonic cleaning or if an ultrasonic cleaner is not available, bath the injector in Berryman B-12 or similar for at least a couple of days.

You can also test the injector by attaching a Mitivac style vacuum pump to the nozzle end of the injector and draw a vacuum. It should hold a vacuum if the injector nozzle isn't "leaking". If the injector is leaking or clogged, a new injector should solve your problem. I see these for sale from time-to-time on the big auction site.

If neither of these solves your problem, you are probably looking at a fuel delivery issue which is much more complicated but still correctable. Let us know your results after swapping plugs and/or cleaning/testing the injector.

[reddogmotrsprts]

Good news: That plug looks perfect to me; a slightly brown/tan color with a fine chalky surface. If it was ashy white, that'd be a sign that it was too lean.

Bad news: Seems you have 5 cylinders running rich, not just 1 cylinder running lean.

One easy test would be to swap the injectors between #5 and #6 after putting a new plug in #5. Run the car and check both plugs.

[MDamen]

These injectors do not open until about 150 psi if I remember correctly. A mityvac will not generate much vacuum. A mityvac wouldd show you the injector that is running rich because it sprays at too low a pressure.
A clogged injector is usually just a partial clog with a bad spray pattern, not totally clogged.

[reddogmotrsprts]

Any progress?

Years ago, we made an injector tester out of PVC pipe with a couple valves, along with a glass jar with a nut mounted in the center of the lid. Screw the injector into the lid, fill the pipe with fuel, apply compressed air to whatever PSI you choose, open valve to test crack pressure and then see how much fuel is injected into the glass jar over a set period of time. The glass jar allows you to inspect the injector spray pattern as well. If you do this, it's wise to include a fuel filter in the loop.

[lateapex911]

Cool setup, but make sure the lid is substantial, 150PSI could blow things apart.

[Sway Bar]

MFI system and the injectors open at about 220-250 PSI. A perfect injector will chirp and atomize the fuel to a fine mist and should hold at least 190PSI without leaking. A partially clogged or dirty injector will spray wet fuel, drip etc. and won't hold much pressure.

Bench testing/cleaning will be the only way to see how 'good' your injectors really are and then you can go from there.

[reddogmotrsprts]

Originally Posted by lateapex911

Cool setup, but make sure the lid is substantial, 150PSI could blow things apart.

The reservoir doesn't need to hold pressure, so the main limitation will likely be the ability to create a pressure source above the limits of schedule 80.

Nowadays, I would probably try one of the cheap diesel injector testers from eBay. I just checked and they start at around $50. You get a product designed for much higher limits with much less work and the question simply becomes whether you can add pressure finely enough to perform a reproducible test.

I actually have a set of MFI injectors to test, so I guess I'll play guinea pig here.

[422flat6]

Hey, yes after a short delay of getting the right tools for the job I took the injectors out and sent them off to a specialist for testing. That should be done by next week. I'll update when I get the results.

I found this video about MFI injector testing to be educational https://www.youtube.com/watch?v=nmkiBz65Wjs

[422flat6]

I sent the six injectors out to Mark Jung for cleaning and testing. He said they're all working properly, holding pressure and the spray patterns look great at idle and higher RPM. While the injectors were out I checked the #6 hard fuel line for blockage and found none. I reinstalled the injectors and put new plugs in and went for a drive yesterday. It ran much smoother and I didn't hear/feel any lean popping at all so I was excited to pull the plugs this morning and see what they looked like.

No change. Bummer. #6 is still bright white and #5 is still black.

My plans going forward are:

• Check the wires in the dark for leakage/shorts

• Visually check spark at #6 (need to recruit a helper for this)

• Measure fuel pressure and flow

Other than those tests, is there anything else it could be besides the pump itself?

Here's a photo of all the plugs from before the injectors were cleaned. 1-6 left to right. I only checked 5 and 6 of the new plugs I installed after installing the clean injectors so I don't have a photo of those, but 6 is still white and 5 is still black. In fact, the new 5 is blacker than the 5 in this photo.

[lateapex911]

This is going to sound crazy (And might be!) but what about getting 6 calibrated glass cylinders, pulling the injectors, positioning them to get the flow of the injectors, then cranking the engine over (Be sure disconnect the central plug wire on the dist) for say 15 seconds and comparing amounts?

Also, I am no plug whisperer, but are we sure thats a rich fuel condition? Might it be oil? Does the motor smoke on start up or on the overrun?

[422flat6]

Time for an update on my progress. Since my last post I focused on correlation, thinking that maybe #6 was getting more air than the other cylinders. I pulled the throttle bodies off and cleaned them with B-12 Chem tool, including the idle air screw passages. They have not been rebuilt, but they are nice and tight and do not appear to need it.

Using my home vacuum cleaner, a synchrometer and a set of feeler gauges, I adjusted the throttle stops, idle air screws and linkage lengths to make each throttle flow equally. This worked really well. Each side is driven off the center throttle (#2 and #5), and the center throttle lever has a nice flat spot that rests on the throttle stop screw, perfect for using a feeler gauge to open the throttle by a known amount. I adjusted the links for identical flow numbers on all six throttles at idle stop, .1mm, .25mm, .5mm, 1mm and 2.8mm off idle stop (measured using feeler gauges on the center throttle). It took about 16 hours of work to get them all set, including my learning curve having never done this before. One tip if you're going to try this... set your idle at 950 RPM and know the flowrate needed for that before removing the throttles from the engine. Then measure the flowrate on the vacuum before changing anything and use that as your target idle flowrate. It will be different because the vacuum and the engine will pull different amounts of vacuum.

Back to my story. I reinstalled the throttles and started the car. Idle was right on the money thanks to pre-measuring the flowrate needed. I warmed up the car and then set it to about 2200 RPM with the hand throttle and measured flowrates through the plastic stacks. 4-5-6 were flowing more than 1-2-3, so I adjusted the right hand side link from the main cross bar to the #5 throttle. Now I had all six throttles flowing evenly on the car with the engine running (HOORAY). I double-checked my pump link was 114mm. I think this is about as good as you can do without a set of P228b and P228c protractors.

I went for another drive and the car ran a lot better with less sputtering and hesitation. I didn't really expect this, since it should still be way too rich. Today I pulled plugs 4, 5 and 6 to look at color. 4 and 5 are now evenly black with dry carbon, where previously 5 was leaner than 4. #6, however is STILL BRIGHT WHITE.

So the problem remains. At this point, could it be anything other than the MFI pump?? A test like Jake describes measuring the fuel output of each injector would definitely show if the pump is sending less fuel to #6, but it seems kinda dangerous squirting all that open fuel around while cranking the engine over. Thoughts?

[reddogmotrsprts]

Can we see some photos of the injectors now?

I stand by my original assessment: that photo of #6 shows ideal air:fuel. If you had bright white scale which flakes off, that would be a sign of a lean condition. Instead, you have a light brown/tan surface that is very fine, with no significant build-up.

Look at the back of a Hayes manual for a color pictorial. At least, I think I've seen them there...

[422flat6]

I didn't take any photos of the cleaned injectors. They came back with a clean bill of health. There was no debris on the internal screen and the tip was free of crust when I installed them.

I agree red dog. I don't mean to say that I think #6 is too lean, I'm just saying it's leaner than the rest. Maybe that's phrasing it backwards. The problem is that 1-5 are too rich and I can't lean out the mixture to fix that unless all the plugs are more or less the same. If I did that then #6 would go too lean. My thinking was that the only way to individually adjust mixture other than tear into the MFI pump is with air i.e. throttle correlation. Now that I've done that and the imbalance problem hasn't been solved can we conclude the pump needs to be rebuilt?

[Jonny042]

I agree that your next step is to have a look at the pump. You're pretty resourceful (to say the least!), you should have it out of the car in an couple of hours, and who knows, the problem may be immediately apparent once you take the side covers off and have a look at the plunger movement. There's a chance something has caused #6 to hang up, and the spring pressure isn't enough to make it follow the cam. It also could be the non-return valve at the top of the plunger, or simply that cylinder has gone out of adjustment.

If you haven't purchased Michael Burgess's book yet I'd highly recommend it. Even if you don't plan on building your own pump dyno and rebuilding your own pump, it's a wealth of knowledge. I'm going to assume you've found:

Ultimate MFI resources thread

[Peterfrans]

If injectors check out and correlation is spot on, then my guess would be that the pump element for the lean cylinder is out of adjustment.