Hi all!

I tried to post this earlier, but I think it got lost in the ether. New to the forum, and I brought a fun puzzle for all you Honda experts! I've done lots of google searching and haven't come up with anything that's quite the same conditions as this problem, and I've been beating my head against this car for the last several evenings, so I'm open to ANY ideas at this point. Here we go!

2002 Accord automatic, 3.0L v6 VTEC

Problem:
Somewhat rough idle, when trying to rev and hold the engine to about 2.5k RPM in park or neutral, it revs up for a second and then comes back down and almost stalls. Under very light through moderate acceleration from 1k to about 2.5k rpm, it will do the same type of thing, producing no power and dropping rpms until it almost dies, followed by a spike back to the normal power setting and another fall. That produces very heavy and randomized bucking. If I drive it like I stole it for just a few seconds (not WOT, but accelerating quickly), it performs fine, and then for the rest of the drive even driving gently it performs fine! Idle smooths out, no bucking under any conditions. But if I shut it off for a couple minutes and restart, it's back to the bucking and rough idle until I drive it hard for a few seconds.

It had a CEL and TCS light, so I read the codes down at the parts store and got:
P0401 EGR flow insufficient
P0131 O2 bank1 sensor 1 low volt
P0132 O2 bank1 sensor 1 high volt

Pulled the intake, and sure enough the EGR port was clogged solidly shut. Cleared it out, cleaned the throttle body while I was in there, reset the computer and let it do an idle learn. No change in the symptoms, but it cleared up all three codes, so I'm chalking it up to unrelated issues.

Other (possibly related?) problems:
- I suspect the alternator/voltage regulator is going bad. Headlights and dome light flicker at idle, and the door ajar, battery, and brake light will all flicker on and off occasionally while driving. Checked the running voltage and it is 14.6V, but I'm thinking there is a ripple that my multimeter isn't picking up.
- Oil leak at the VTEC oil pressure switch. I tightened the switch about 1/16 turn and it appears to be sealed now.

What I've tried:
-TPS sensor replaced and adjusted. (no change) Then tested the old one and it was okay anyway.
-Coolant level checked okay, changed coolant anyway and let the car warm up with the radiator cap off to purge air. (no change)
-Oil level okay, changed oil since it was time anyway. (no change)
-EGR port in intake cleaned, throttle body cleaned, EGR valve cleaned, IAC valve cleaned. (no change)
-IAC valve resistance tested, pin 1-2 51k ohms pin 1-3 and 2-3 10M ohms. (not sure if this is normal?)
-IAC valve disconnected while idling, idle speed changes but doesn't smooth out.
-EGR valve disconnected while idling. (no change)
-Unplugged coils one by one, no bad coil found.
-Voltage regulator connector at alternator unplugged while idling and driving, headlights don't flicker, but no change in drivability.
-Several computer resets and idle retrains throughout the process. (no change)
-Cussed at it, reasoned with it, threatened it, offered it chocolate. (no change)

I'm stumped because of how repeatable the problem is, and yet it clears up as soon as I drive it hard, and stays cleared up for the rest of the drive. Could it be something to do with the VTEC system, or oil leaking from it and fouling an O2 sensor that isn't throwing a code? Thinking it can't be fuel quality or pressure since that shouldn't clear up and stay cleared after accelerating hard... right? Also thinking it may be an open loop vs closed loop type of issue since it is only at the start of a drive until I wind it up... but really, throw whatever ideas you've got at me!

Thanks for reading this novel of a post, apologies for any ignorance, and THANK YOU in advance for any input.
-Thomas

 

It's a 3.0, it likes beer, not chocolate.
What about the MAP sensor?
This would have a huge impact on how the engine operates. Although TPS is important Hondas can run pretty well even with a shot TPS, MAP will cause havoc as it needs the MAP to know what the engine load is. Dat Speed Density Lyfe. MAP and IAT sensor are used by the PCM to calculate what the air mass entering the engine, as well as calculating load using the TPS.
I'm guessing the diaphragm/crystal is failing and when you mash on it the MAP gets 'fixed' until the engine is turned off, then diaphragm rests into a position that doesn't allow it to move smoothly.
If you can get a vacuum pump, not the swedish kind, and test your map while monitoring voltage change anybnon linear change of voltage vs inHg may indicate a fault in the MAP.

 

Good thinking, thanks for the quick response Mike! I'll dig through my parts/tools bin and see if I can fabricobble a vacuum test apparatus. And I'll make sure it isn't the Swedish type FOR SURE. That might get some weird looks from the neighbors...

I'll try that before pouring beer into the intake. But I may have to pour some into my intake to really get into the fabricobbling mindset.

-Thomas

 

Have you tried live data on your scanner? Make sure ect is right, and the engine goes into closed loop ok? Also are the upstream o2 sensors equal and responsive?

 

Quick update, MAP sensor bench test with a 5V supply and a vacuum pump checks out. Here's the data.




So unfortunately it looks like the MAP is okay.

I don't have a scanner, just borrowed one in the auto parts parking lot to pull the codes the other day. I'll look into how expensive of a scanner I'd need to do this... but my truck just decided to eat some money in the form of a stuck A/C clutch on a long drive, and a failing power steering hose. Not shaping up to be a great week...

Thanks, and keep the ideas coming if you got 'em!

Tried a detailed reply with my test data, but it too disappeared into the ether. Long story short, the MAP sensor checks out on a bench test. Pulling 22.5 inHg vac at 3000', I get 0.38V out. At atmosphere, it's 2.61V. Linear within error throughout the range.

As for the scanner, I don't have one. I borrowed one in the auto parts parking lot the other day to pull the codes. I know I should get one, but I'm sinking too much money into truck repairs right now. Not a great week for my vehicles!

Thanks for the ideas, and more are welcome!

 

try warming the car up and disconnecting the oxygen sensor connector so it throws a code and see if it still does it

 

No, the post wasn't approved yet (just like the prior thread you tried to make). You shouldn't be experiencing issues, going forward.

 

Great minds think alike! I just got back from a test drive without the O2 sensors. CEL on, but it drove like it is supposed to. So the question is, does that mean bad O2 sensor, or does it mean the car went into open loop so the problem went away?

toyomatt84, sorry for the ignorance. I was having browser issues as well, so I didn't realize my post was pending approval.

 

Pretty sure it's the front O2 sensor. Both are on order and should be here late next week, I'll post back with results in case someone else has the same problem.

 

its very important to ensure there is no exhaust leak before installing the new sensor

seafoam it while its on a cold start, idling high and look for white smoke billowing out anywhere from the exhaust manifolds to the oxygen sensor

 

Thanks for all the help guys, looks like the upstream O2 sensor was the culprit. I got my hands on a code reader that does live data, and the old O2 sensor was swinging about 0.35V to 0.70V, whereas the new one is 0.1 to 0.8. Wasn't throwing a code, but it must have been on its way out. The car is idling nicely now and has smooth power delivery on the test drives. I swapped the alternator too (along with the belt and tensioner... a bit of a bastard for big hands to get in there to a couple bolts!), and the electrical system seems to have settled down, so I probably avoided a nasty over-voltage situation there.

Now I just need to track down a couple oil leaks so the new O2 sensors don't get doused...

Thanks again everyone, I appreciate it!
-Thomas