Honda Odyssey Forum banner

Honda Odyssey Pilot Accord P0171 P0174, Lean Conditions & High Fuel Trims

29K views 38 replies 6 participants last post by  John Clark 
#1 · (Edited)
So, I have been working on this problem for a while. My 08 Touring has been running lean for the last few months and not getting very good gas mileage either. It hasn't set any codes...I only put the codes in the title for search purposes but the PCM has been adding fuel to correct for what it sees as a lean condition.

An oil change or two ago I connected my scan tool as I routinely do at oil changes. This particular time I noticed that my fuel trims were running around +10% on both short term and long term. If you add them together that's a total fuel trim of +20%. That means the PCM is adding 20% more fuel to compensate for a lean condition. When running the RPM's up to 2K or 3K the fuel trims stayed about the same which I figured ruled out a vacuum leak, however, I searched for one anyway but did not find anything. Scan tool indicated the EGR was commanded closed and indicating closed so I didn't suspect an EGR problem.

Since both banks were about the same I figured possibly a mass air flow sensor. I removed it and cleaned it to no avail. With nearly 170K on the clock I started thinking O2 sensors. That's where things get a little tricky. On the 05 and up Honda used air fuel ratio (wideband) sensors rather than standard O2 sensors. These can be much trickier to diagnose because they don't cycle rich-lean-rich-lean like normal sensors. So, with the mileage I went ahead and replaced both upstream A/F ratio sensors. That seemed to help and the fuel trims came down to under 5% on both short and long term. However, after a few hundred miles I checked again and my fuel trims were back up to 10-20% total between ST and LT again. WTH?

After another month or two of checking scan data and further research I figured it had to be either MAF or downstream O2's. I was able to drive the sensors rich and lean when desired and they appeared to work correctly. However, I decided to change the downstream O2's anyway. These are standard narrow band O2's and, in most cases, are usually just for checking catalyst efficiency and comparing to the front O2 sensor. However, it turns out that on A/F ratio equipped Hondas the PCM does use the downstream sensors for fuel trim adjustment. What I didn't know was to what extent. I mean, if Honda puts in the fancy wideband, much more accurate, A/F ratio sensors on the front how much control could the rears have?

So, I went ahead and changed the rear O2's last week. After some monitoring I saw the fuel trims come down a bit to around 10-13% total, so there was some improvement. A bit of driving around town and they were down to around 8-10% total. With a good repair I'm used to seeing an opposite short term correction that will equal the long term correction. For example, if the long terms are at +10% and you repair the problem, you'll see the short term tick down to -10% which will eventually push the long term fuel trims back to zero. If you simply add the short and long terms and they equal zero then you usually have a good repair. I wasn't seeing that. At that point, I figured I must have a MAF problem so I went ahead and ordered a new MAF.

The new MAF showed up today. However, before just throwing it in I figured I'd check the fuel trims so I could compare them to the post MAF installation fuel trims. To my surprise, my long term fuel trims were already at 0 and the short terms were around -1 to +2, nearly perfect, or even a little rich of normal with the PCM subtracting a little fuel. I can only surmise that it takes a bit longer than normal for the computer to adjust the fuel trims after a component change like that. I didn't reset codes (since it didn't have any codes) or disconnect the battery to reset all that...I just watched to see if the PCM made any corrections to confirm the fix. Apparently, I didn't watch long enough because the PCM finally brought things back to normal.

The reason for this post is that NONE of this is in the Honda FSM and there is VERY little about Honda lean codes on the Internet. If you get a lean or rich code (or even if you just have a lean/rich condition but no code like I did) all Honda says to do is check fuel pressure (not easy to do,) check the MAF sensor g/s at 2500 RPM, inspect valve clearances and then replace injectors if none of that fixes it. It says that if the LT fuel trim stays within -22 - +25 then there is no problem. The lean/rich codes set at -22 or +25% on LT fuel trim. The problem is that if the O2's are old and not working efficiently then you can have +22% fuel trim, crappy gas mileage, no CEL and it could eventually take out your catalytic converter. I often wonder if this is why there are so many surprise P0420/P0430 codes on Hondas with no prior warning.

So, my take on this is that if you have high mileage on your J35 engine it's probably a good idea to replace ALL FOUR oxygen sensors as a maintenance item. If you have a scan tool you can keep an eye on them like I did. If not, then I'd recommend just changing them by no later than 150K. I bought my OEM NTK sensors on RockAuto and found some service life info from NTK there:

https://www.rockauto.com/en/moreinfo.php?pk=2258828&cc=1441873&jsn=365

Essentially, it says for modern 4 or 5 wire sensors the service life is about 100K miles. So, if you're over 100K miles and scan data is showing much over 10% total fuel trim adjustment, I'd consider replacing ALL FOUR sensors. Don't just replace the two upstreams. The downstreams have a lot to do with controlling fuel.

Hope this helps someone...

John

Another document from NGK on sensor life expectancy:

https://www.ngkplugpro.ca/content/contentfiles/pdf/NTKO2-0308-1LifeExpectancyofOxygenSensors.pdf
 
See less See more
#3 ·
John, my 2010 EX-L, with only 44K miles, pulled lean codes a few months ago. Being that I'm currently crippled, I just kind of cleaned the two sensors on the intake system, and I didn't replace any O2 sensors.

Other than the long and short term fuel trim correcting for the lean condition, the car ran fine with normal gas mileage.

Then all of a sudden the engine light went out, and the fuel trim was normal. It hasn't come back, but........

Maybe if others have had the same mystery disease, we can get a better feeling for what is actually going on.
 
#4 ·
I don't mean to say that O2 sensors aren't the only cause of lean codes. At 44K miles I'd be suspecting something else unless just one failed. You can't really clean O2 sensors anyway. The part that senses the O2 is underneath the metal probe portion with the holes in it.
 
#6 ·
I don't believe in "magic." The one variable in the sudden change from lean to normal is filling it with gas. My wife did fill the tank at a QuickTrip in between when I installed the rear O2's and when I went to check before replacing the MAF. Did I just happen to check right after the gas stations changed to summer blend? I don't know. All I know are the facts above. NTK does recommend a service life of around 100K. I wouldn't say that at 100K you should immediately replace them but by 170K I just don't feel too bad putting in new O2's and a MAF, considering they are the heart of fuel control on any modern vehicle.
 
#7 ·
I think I'm running into something similar on my '09 EX-L. Short term fuel trims are around 12-14% at idle (after warming up), long term are around 8%. But what got me looking at it is the occational hesitation while accelerating. I'll try to clean the MAF later today and see if anything changes.

We've had the van for a year and a half and I have no idea how old the O2 sensors are, so maybe it's time I replace them anyway. Although that's a big chunk of change for all 4. Better than waiting for them to go back completely I guess.
 
#11 ·
So, this likely rules out a vacuum leak somewhere. It could be an evap leak, too. If the charcoal canister is not collecting any vapors, due to a large leak, then whenever the purge valve is commanded open the PCM will expect vapors to be pulled in and burning. If there are no vapors it will have to add fuel and the fuel trims will reflect that. If you don't get any evap codes or gas cap messages then that's probably ruled out.

How many miles on the van? Has it had a valve adjustment?
 
#10 ·
I replaced all four sensors on an 09 Pilot with 170k on the clock a couple of weeks ago. The owner reported to me the other day that his mileage has jumped up to around 17.5 mpg from around 14.5 mpg so he's happy with having a new set of O2's on his Pilot, too. He had some other problems that were causing a lean code and he bought the new sensors before I had a chance to look over everything else. I found cracks in the intake tube which, after replacement, helped some but was still a bit lean. We went ahead and replaced his O2's but I didn't see much change initially. It was only after a week or two that he reported the mileage improvement...which indicates if I were to connect my scan tool I'd probably see fuel trims at or near zero.
 
#12 ·
It has 260,000km. I have no maintenance records from before we bought it, but I doubt it ever had a valve adjustment. I have thought it does sound kind of "clicky", but I don't know what bad sounds like.

The "commanded evap purge" was sitting around 29%. I'm not sure what other values to look at for that. There's also "Evap system vapor press", which was around 32, whatever that means :)

I recorded my trip to the store to buy the MAF cleaner. I tried to record the trip back, but the program kept crashing on me. Then, of course, I got that lag pulling out of the parking lot. I hit the gas and barely anything happened. I let off, then pressed the gas down more slowly and it went as expected.

I did clean the MAF sensor and tried it again. No change really. Fuel trim sits around 16% at idle. When I accelerate to 3000rpm, it changes momentarily, but then settles back to roughly the same.

Long term is still 7.0% and 7.8%.

Both banks are reporting roughly the same across the board (O2 sensors, fuel trim).
 
#13 ·
So, that's around 161K miles. It probably should have a valve adjustment if it's never had that. Lean codes can be caused by valves out of adjustment. Valve adjustment is recommended at 105K with the timing belt and plugs but going to 140-150K is not a big deal. If the O2's have never been replaced then I'd probably replace all of them if you're planning on keeping the vehicle. In the US you can get all four from RockAuto for around $250.00US. Not sure if that's the best source in Canada or not. You can also buy them directly from NTK's website. They're not difficult to replace...getting the wire harness clips released is the hardest part.

Adjusting the valves is not for the faint of heart. The front bank and the rear intakes aren't too bad. The rear exhaust valves are a real pain to adjust.
 
#14 ·
Yeah, I've read up on doing a valve adjustment. I wouldn't normally mind tackling it, except I don't really have the time right now. We have 4 kids and are moving 4 hours away at the end of July. I'll ask our mechanic how much he'd charge.

When we got the van, I did the timing belt, spark plugs, and all the fluids. I'll probably bite the bullet and do all the O2 sensors now. RockAuto's shipping cost to Canada is usually pretty decent, but I'll shop around.

I realize the hesitation and the fuel trim could be two unrelated issues. For the hesitation, do you think the EGR valve could be at fault?
 
#15 ·
Before replacing my O2's I noticed just a very slight hesitation on acceleration. It was almost barely noticeable but I noticed it. It was completely gone after replacing the O2's. No guarantee you'll have the same results but with the mileage you have I don't think replacing all four O2's is wasting money, even if it doesn't ultimately solve the problem. They probably need it and you've now ruled that out. If you were at 100K or less I'd be a bit more hesitant to throw all new sensors in there. Also, it's a good idea to do a PCM reset and an idle relearn after replacing the sensors. This forces the PCM to recalibrate for the new sensors. It takes a higher end scan tool to do that, though. If you don't have a capable scan tool and have to take it in for the reset and relearn then I suggest cleaning the throttle body first if that hasn't been done. An idle check is recommended at 160K miles which basically consists of checking the throttle plate angle. If the throttle body hasn't ever been cleaned the angle will be above spec and the throttle body will need to be cleaned. It's best to just clean it and do the reset and relearn.

The EGR is a possibility, too. I haven't had to do any diagnostic on my EGR, though. If your scan tool supports EGR data pids then you can compare the commanded state to the actual state and see if it's opening and closing. If it was stuck open you'd probably have a rough idle. I have read, however, that "mysterious" issues have been repaired by replacing the EGR. At this point that's just throwing a part at it, though.
 
#17 ·
Front and rear refers to front and rear bank of cylinders. Upstream and downstream refers to before the converter or after the converter.

That sounds right. Double-check but on an 09 EX-L it looks to me like you need:

2 - Upstream Front/Rear: NTK 24346
1 - Downstream Front: NTK 24437
1 - Downstream Rear: NTK 24435
 
#19 ·
So, funny story. I went to clean the throttle body today. This morning while I was contemplating it, I thought about the air filter. I know I had bought one when we first bought the van, but I remembered not using it since the filter that was in there looked almost new. But then I didn't remember every coming back to check it. Sure enough, when I went out of the garage, I had a brand new air filter sitting there that I had bought a year and a half ago and never used.

I decided to clean the throttle body anyway, so I did that (it definitely was a bit dirty) and put everything back together with the new air filter.

After a quick trip out and back I plugged in my laptop again. And of course, all the fuel trim numbers are right down around 0 now.

The O2 sensors are already shipped, so I guess I'll still use them. Won't hurt. Who knows if the air filter/throttle body was the cause of all my troubles.
 
#20 ·
So, I probably should have verified to begin with but were your original fuel trim numbers you posted about positive or negative? Your post just said the percentage so I assumed positive fuel trim numbers, indicating a lean condition. However, a dirty air filter will usually pull fuel trims into the negative range. The engine is getting less air so the O2 sensors will sense a rich exhaust and reduce the amount of fuel, which would show in negative fuel trim numbers.

I don't think the dirty throttle body would have any effect on fuel trims, though it might contribute to a bit of a hesitation from idle. I'm surprised that after cleaning it you don't have any high idle issues. After cleaning mine my idle was too high and only a scan tool would fix it.
 
#21 ·
The numbers were positive. You understood correctly.

I was thinking about that too - seems odd that the ratio was what it was if a dirty filter was the issue.

I did have to reset the check engine light. While I had the air intake off, I turned the ingition on (not the engine) and put a brick on the gas pedal just to open up the throttle to clean it. After I put everything back together and took it for a drive, I noticed the check engine light on. I figured it just took offense to not having those couple of sensors connected when I had turned it on. So when I got back, I cleared those codes, and that's when I checked the fuel trims again.

So I don't know if clearing the codes had anything to do with it.
 
#22 ·
Yes, if you clear the codes you also clear the long and short term fuel trims and they reset to zero. They will work their way back up if the problem is still there.
 
#24 ·
Yeah, 900 is too high. You need a bidirectional scan tool to do the PCM reset and idle relearn. If you have a VCMuzzler you'll want to remove it before doing the reset and relearn as during the idle relearn process the ECT1 temp needs to reach 194F, which may not happen with the Muzzler installed. Then once you've done that you can reinstall your Muzzler. This is where a variable resistor type Muzzler comes in handy...dial it down to zero, do the relearn, and then dial it back up to the proper setting again.
 
#25 ·
You sure you need a bidirectional scanner?

I'm just looking at the manual now. It says this about the PCM reset (page 11-4):
Sean Tool Clear Command
If you are using a generic scan tool to clear commands,
be aware that there is only one setting for clearing the
PCM, and it clears all commands at the same time (CKP
pattern learn, idle learn, readiness codes, freeze data,
on-board snapshot, and DTCs). After you clear all
commands, you then need to do these procedures, in
this order: PCM Idle learn procedure (see page 11-410);
CKP pattern learn procedure;
Test-drive to set readiness codes to complete (see page
11- 100).
What I understand from that is that clearing the codes means I just reset the PCM.

Then the idle relearn instructions on page 11-410 don't seem to require a scanner at all either:
Procedure
1. Make sure all electrical items (A/C, audio, rear
window defogger, lights, etc.) are off.
2. Reset the PCM with the HDS.
3. Turn the ignition switch to OSM (II), and wait
2 seconds.
4. Start the engine. Hold the engine speed at
3,000 rpm without load (in P or N) until the radiator
fan comes on, or until the engine coolant
temperature reaches 134 °F (90 °C).
5. Let the engine idle for about 5 minutes w ith the
throttle fully closed.
NOTE: If the radiator fan comes on, do not include
its running t ime in the 5 minutes.
But good point on the muzzler. I wouldn't have thought of that.
 
#26 ·
100% sure. I've been this road before. When they mention "generic scan tool" they're referring to something that's not the Honda Diagnostic System (HDS.) Resetting the PCM is much different than clearing the codes with an OBDII scanner.

A bidirectional scan tool used to reset the PCM will reset everything and you have to do the CKP relearn (doesn't require a scan tool) and the idle relearn (does require a bidirectional scan tool.)
 
#27 ·
Now I know you have a ton more experience here than I do, so forgive me :) But I understand that when the manual says "generic scan tool", is means a non-HDS (and non-bidirectional) scan tool. But it also seems to suggest that those generic scanners can't pick and choose what it clears - it can only clear everything. So if the DTC's get cleared, everything, including the PCM and idle learn, is cleared.

Are you saying otherwise based on experience? Or some other documentation?

I still haven't pulled the trigger on that Foxwell NT510 that you suggested. But I think I will after we get moved.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top