Long term fuel trim out of spec bank 2

Well, subject says it. Short term fuel trims are all on-spec, long term on bank 1 (driver side) is correct, but long term on bank 2 is out of spec (+17 over, adding fuel to correct for lean condition)

I swapped the oxygen sensor on that side for a new one (acdelco) and it didn't change. I figured that was easy to do and the most likely thing to blame. It wasn't of course.

Tonight I did a fuel injector balance test- here are my results, pressure drop in PSI listed below

#1 25
#2 25
#3 26
#4 25
#5 27
#6 25
#7 27
#8 26

Average drop 25.75, GM spec says +/-1.5 is ok so 24.25 to 27.25. According to the test the injectors pass. I figured either the oxygen sensor or a fuel injector would be to blame, but again neither one tests bad.

What are some other things that may cause the PCM to think I need more fuel only one one side of the motor?

vacuum leak somewhere. exhaust leak upstream of the O2 sensor.

Joel - thanks. Exhaust leak would be something I didn't check for. I'll do that too.

exhaust leaks aren't always noticeable if they are small especially with headers, the headers let more sound into the engine bay and hide the leak sounds. many times you can feel the leak with your hands, obviously be careful around the hot headers. also you can use a piece of paper/tissue to see if it gets blown around by the exhaust.

I used the heater hose to the ear diagnostic methodology first, no leaks heard. Then the strip of paper method, nothing there either. So I did the right thing and went and got a smoke machine. Jammed the stopper into the tailpipes and filled the exhaust with smoke- no leaks found (except for the hole drilled in the rear mufflers, of course) also smoked the intake manifold- no leaks found there either. So, still looking as to what may be going on.

Seem to be having a heat-dependent poor idle issue also. May or may not be related.

hmm... I don't have any other ideas at the moment.

Nice job on the diagnostics so far.

There aren't a lot of things that are whole bank specific, AFAIK, but the fuel injector feed tube is a loop in most recirculating fuel systems like these. Perhaps there is a restriction of some sort in the cross tube between the banks or some other disruption in that area as that would make the banks see different fuel pressure.

That's possible. I do think I have another fuel rail I could swap and see.

I am still scratching my head on this one. The only things I can think of that would do this would be

1: an injector that's not dropping enough fuel
2: a vacuum leak specific to the right bank
3: an exhaust leak between the head and O2 sensor on the right bank

One thing I'm wondering, since I am still getting what feels like a random misfire when the motor goes into closed loop: if a cylinder doesn't fire and dumps unburned air/fuel into the exhaust stream, would the O2 read rich or lean? I'm thinking lean since the unburned mix would have a lot of oxygen relative to what it should be. So maybe this is some sort of ignition problem.

I did some testing just now, trying random things, and I have two interesting things that came out of it.

First - when I capped the vacuum fitting at the throttle body that goes to the purge solenoid, the engine ran much better.

Second - when I capped the vacuum fitting that goes to the fuel pressure regulator leaving the FPR with no vacuum, the bank 2 trims returned to normal, and bank 1 went rich.

I'm not sure if that sounds wrong, it makes sense to me since we want to drop less fuel at idle.

Ok, so latest update on this: I pulled the fuel rail, replaced the fuel pressure regulator and swapped the injectors from left/right- also replaced the o-rings while the injectors were out. There was no difference in the result, so it doesn't appear to be related to fuel supply.

One thing the right bank has that the left doesn't is the EGR tube- could be a leak there that's hard to detect so will have to figure that out.

Well it took far longer than it should to find a rather significant exhaust leak, and it is of course at the mating surface where the EGR tube and fitting on the header meet. Obviously it doesn't meet exactly right and that fitting normally doesn't have a gasket. But I think it might need one. If anyone has had a problem there and fixed it let me know. I can try making a gasket, maybe there's a RTV or o-ring that can take that sort of heat?

The manifold to exhaust and EGR to intake are metal reinforced graphite to they hold pressure and handle the high temp.

Something like this is probably very similar.

[You must be registered and logged in to see this link.]

If there is a leak there at the exhaust header, I don't see how that would affect only one bank though, as it is a single tube to the intake.  It would still need to have some kind of mismatch in the intake itself, I think.  If the header is giving strong negative pulses at certain rpm/load points it might be introducing air/oxygen into the exhaust and messing up the O2 reading, though, I suppose.

If it is the O2 reading getting messed up, that should show up if you look at and log the O2 voltages for both sides to compare them,

Booster I think that's the theory anyway. A leak there would result in a lean condition.

It's a problem, but I don't think it's "the" problem. I did notice just now that while it leaks pretty badly when the exhaust is cold, once the exhaust warms up it appears to stop. I do know that when driving sometimes when slowing to a stop you can smell exhaust being brought into the cabin (but not always) - so maybe I just found an intermittent leak? (ugh, great)

A vacuum leak specific to that bank still seems the most likely explanation. I did notice that the counts are much worse at idle than they are while cruising, which does correlate well with engine vacuum.

lamune wrote:

Booster I think that's the theory anyway. A leak there would result in a lean condition.

It's a problem, but I don't think it's "the" problem. I did notice just now that while it leaks pretty badly when the exhaust is cold, once the exhaust warms up it appears to stop. I do know that when driving sometimes when slowing to a stop you can smell exhaust being brought into the cabin (but not always) - so maybe I just found an intermittent leak? (ugh, great)

A vacuum leak specific to that bank still seems the most likely explanation. I did notice that the counts are much worse at idle than they are while cruising, which does correlate well with engine vacuum.

I had a block off plate kit on mine and the gasket they sent was junk. It failed in the first month and got very loud because it was thick and allowed whole cover to spin on the single bolt. I pulled the manifold and tapped the whole for a pipe plug so pretty foolproof and no good if you want to keep the EGR.

Yep, the leak will make the O2 indicate lean and it will richen the mixture to compensate so cylinders will run rich in reality when the O2 says all is good.

I did ask the guy who makes the headers about that and he suggests that some high temp RTV is the best way to seal up that joint.

That being said Booster I ordered a blockoff plate set. I want to see if blocking off EGR has any impact on the fuel trim numbers. I really wish it were easier to get in there though, that EGR pipe bolt is really hard to get to!

EGR tube blockoff plate arrived and installed. Re-ran the test and now the fuel trims are within normal range.

When I got the EGR tube out of the header the whole end of it (and the metal mounting plate) were covered in soot, so it was definitely leaky.

[You must be registered and logged in to see this link.]

Now you need to find out if it was the leak, the EGR valve, the controls or something with the manifold internals. Of course you could just use the EGR plates at both ends and be done with it.

Yes, not done yet though it appears (so far) to be limited to the EGR pipe at the header fitting.

I'm also wondering if my cam has enough duration that I can dispense with EGR entirely. My exhaust duration is 217 and the LT4 exhaust duration is 223, so it's pretty close.

Disabling EGR didn't make any perceptible difference. In fact I'd say it ran worse.