Honda Odyssey Forum banner

1 - 16 of 16 Posts

·
Registered
Joined
·
8 Posts
Discussion Starter #1
04 Odyssey. Starts & idles fine on warmup. After warming, begins to idle rough. If I drive about a mile or two it sometimes stalls altogether if I hit a stoplight. After being driven many miles, then parked for a while, on restart it does not immediately respond to acceleration. Huge flat spot. Have to let up on gas, then tromp down a little, then it kicks in. Has plenty of smooth (mostly) power under hard acceleration, but cruising at 75 or 80, at 1500 to 2000 rpm it will stumble when accelerator is used gradually. Check engine light turns off overnight, but comes back on after a rough running episode. Usually TCS also comes on. I have replaced all the plugs, replaced the EGR valve and cleaned out the maze. The maze was completely carboned up, as were the passages to the EGR valve. Replaced the pcv valve. There was a lot of oil in the intake manifold crevasses. Vacuum check reads 17 lbs, steady, when cold. Codes read misfire on 3 and 5 I think. It is random though, and it clearly is NOT missing at all times, as it sometimes has it's old full power and smoothness. It doesn't have the "hammer" you get when a cylinder simply isn't working. Adjusted the valves after watching a YT vid suggesting that could be causing the problem, and they needed it bad. However it didn't resolve the problem. Sometimes on takeoff from a light, under strong acceleration, it feels like when you take off fast on a paint strip, and the wheels slip, then grab. That might be a tranny issue. Now on the verge of either buying a complete reman engine and having it installed, or replacing every fuel injector and coil on the motor. This is my work vehicle and it simply MUST run properly. It's spend the money to fix this, or bail and go into debt buying another vehicle. I just feel like this is a solveable problem that I just haven't figured out yet. Had a guy put his computer on it, and he called out EGR, ABS and a bunch of other problems that I know aren't really problems. He did suggest maybe a variable cam timing sensor or something. I would greatly appreciate any feedback that might explain all these symptoms before I blow thousands on it.
 

·
Registered
Joined
·
553 Posts
You could replace all of the coils, or when the idle is rough do a coil drop test and verify that 3 & 5 are bad and swap those out.
Cleaning out the carbon in the intake and EGR is common. My bad coil I replaced with a Hitachi that I pulled at the U-Pick yard. For plugs, best to use OEM.
I prefer to diagnose and verify the problem rather than throw $ at parts and hope for the best. Fix one and move to the next.
On my 99, got a code for the MAF sensor. Verified it was getting proper voltage, but resistance was bad. Got free sensor from junk yard. Soon after had rough idle and various misfire codes. Verified #5 coil was bad. Drove to the yard and bought replacement and spare from the yard. Verified both were good and replaced bad coil. Problem resolved.
 

·
Registered
Joined
·
2,983 Posts
It could also be a worn timing belt. Sometime they jump a tooth and you can get similar indications.
As for the tranny, be sure the fluid is clean and pinkish and wiithin the marks on the ATF dipstick. If it hasn't been changed in a LONG time (over 60k miles), doing at least 3 drain and fills with driving time in between can really help your tranny.
Most recommend using ONLY the Honda DW-1 ATF but I have had great luck with using the Valvoline MaxLife Dex/Merc ATF (around $18/gal jug at Walmart).
Also, if the ATF external filter on a 2004 is not too difficult to get to, replace it.
I strongly recommend using the recommended NGK spark plugs.
And YES, those coil over plugs do go bad.
Sounds like you are very capable on doing your own work. :D
I need to do the valve lash adj on my 03 as it has 200k miles on it and it has never been done.
Unfortunately, I don't have a garage to work in, only a parking lot. :eek:

Buffalo4
 

·
Registered
Joined
·
8 Posts
Discussion Starter #6
I bought a compression tester today and pulled out all the plugs. The front bank ran around 130 lbs compression. Rear bank every plug tube had oil pooled around the plug. All of the plugs were gapped correctly and looked like healthy firing cylinders. Rear cylinders read between 150 and 160 lbs. I assume that's because some oil dripped into the cylinder and sealed the rings a little better. Coils all looked ok, no oil in passage up the inside of the coil where it snaps on the top of the plug. Small amount of oil on the tips. Read the codes before wiping the light, and got 8 codes. Calling out misses on every cylinder and a couple of random misfire codes. Wiped off all the oil on the coils and plugs and replaced everything. Looked carefully at the seals at the top of the tubes and could not see evidence of oil entry from there. Wondering if the bottom of the tubes is capable of leaking? Took the van for a test drive with high hopes, but still stumbled on idle after warming up. Can't say about the other symptoms, as I haven't driven it any distance yet. Fixing to buy a used engine and have it swapped out. Remanns has a complete used engine for about $400, they claim about 73,000 miles on it. Shop quotes about $1800 to R&R it. I am baffled and about to spend the money. Not a worn timing belt. Just replaced it. Maybe I screwed it up?
 

·
Registered
Joined
·
8 Posts
Discussion Starter #7
Btw, thank you for all of your ideas. In thinking about it now I'm haunted by the idea maybe the belt is just off a notch. It felt right when I put it on, but wouldn't that be a simple fix if it was the problem? Is there an easy way to check that?
 

·
Registered
Joined
·
1,687 Posts
I'm pretty sure your belt is not off by a notch. If it were, it would run badly at all temperatures. First time I did the TB on my '99, I was off by one tooth on one of the cams (due to some ambiguity in the service manual, and me not realizing there were two sets of timing marks - the real ones, and also the ones for quick inspection through the port hole). With the TB off by one tooth, I got a ton of misfire codes, and it ran like crap from the very beginning. Barely idled. Actually got better at higher RPM.

From your overall description, I would not be surprised if you have a dirty idle air control valve (IACV). That's a common problem on these vans, and they just need removal and cleaning. Search on here for symptoms.

However, I don't think that is the main problem you've got. I've had IACV issues a few times over the years, and never got misfire codes from it.

I'm thinking you've got ignition issues, meaning probably one or more coils are bad. Again, common problem on these cars, so search on here for that for all the good info. Briefly, "coil drop test"; a single bad coil could be causing all of those codes; the oil in the spark plug wells could be making things worse - common failure of the tube seals that is part of the valve cover gasket set.

I really think you'll do better finding and fixing whatever problem(s) you've got vs. replacing the whole engine.

Good luck.
 

·
Registered
Joined
·
8 Posts
Discussion Starter #9
The way this started was a ticking in the engine and a little roughness. I got a code that said EGR, so I bought a new one, tore the intake down to get to it, and thought, what the heck, I'll go ahead and replace all the plugs, since they are easy to get to. When I went to spin out the middle rear plug, it just spun. In short I accidentally found out the plug had backed out, and the threads were burnt off. The ticking was the plug sliding up and down every upstroke of the piston. The tube became part of the combustion chamber, and I had to replace that coil. My only option, as far as I knew, was pull the head and have it fixed. I did so. Bought an entire upper gasket set and put everything back together. Don't remember why I had to go back in, but when I next pulled the plug, it was covered in oil. Pulled the valve cover and found out that in my ignorance I'd just jammed that cover back on, and distorted the seal badly, which caused the leak. So I put 3 new seals in and carefully lubed them and slipped that thing on, making sure the seals slipped nicely over the tubes, which they did. I thought "problem solved". Unfortunately incorrect. I'm having a hard time believing I screwed it up again, and I don't see any oil at the actual top of the tubes where they meet the seals, so I'm trying to find out if it is possible for the oil to penetrate from the bottom of the tubes.

Concerning the coil possibility: Wouldn't the clean, uniform appearance of all the plugs belie the idea that one coil isn't firing? I'd love it to be that simple, but I don't see any evidence for that idea in what I saw when I pulled the plugs. Why would the obd show misses on all cylinders if it was just one coil? Not being argumentative, I am just seeking understanding. I'm way too old to think I know everything!
 

·
Registered
Joined
·
688 Posts
You have to make sure there is no oil inside the plugs. If you still find wet plugs redo the seals. Once you get that corrected you can start checking which coil pack is defective.

Besides changing the egr these vans are notorious in getting the egr passages clogged. You need to inspect and clean that very well. Using a drill bit or something thin and sharp are needed when its badly caked up. It gets really hard. Use lots of carb cleaner-several cans. If you have this problem you will surely get some kind of hesitation when on the freeway.
 

·
Registered
Joined
·
1,687 Posts
The way this started was a ticking in the engine and a little roughness. I got a code that said EGR, so I bought a new one, tore the intake down to get to it, and thought, what the heck, I'll go ahead and replace all the plugs, since they are easy to get to. When I went to spin out the middle rear plug, it just spun. In short I accidentally found out the plug had backed out, and the threads were burnt off. The ticking was the plug sliding up and down every upstroke of the piston. The tube became part of the combustion chamber, and I had to replace that coil. My only option, as far as I knew, was pull the head and have it fixed. I did so. Bought an entire upper gasket set and put everything back together. Don't remember why I had to go back in, but when I next pulled the plug, it was covered in oil. Pulled the valve cover and found out that in my ignorance I'd just jammed that cover back on, and distorted the seal badly, which caused the leak. So I put 3 new seals in and carefully lubed them and slipped that thing on, making sure the seals slipped nicely over the tubes, which they did. I thought "problem solved". Unfortunately incorrect. I'm having a hard time believing I screwed it up again, and I don't see any oil at the actual top of the tubes where they meet the seals, so I'm trying to find out if it is possible for the oil to penetrate from the bottom of the tubes.

Concerning the coil possibility: Wouldn't the clean, uniform appearance of all the plugs belie the idea that one coil isn't firing? I'd love it to be that simple, but I don't see any evidence for that idea in what I saw when I pulled the plugs. Why would the obd show misses on all cylinders if it was just one coil? Not being argumentative, I am just seeking understanding. I'm way too old to think I know everything!
On the coils stuff ...
In my experiece with these coils failing, they often initially do so intermittently. I never noticed any difference in plugs, since the car would not run enough in this intermittently misfiring condition to allow any plug effects.

On the multiple misfire codes due to a single failure - definitely the case. Observed many times, reported on here many times. I belive misfire detection is made by the crank sensor. It measures the exact speed of the crank as it spins past the sensor. In theory, if #3 (for example) misfires, the crank speed will not be as fast as it should immediately following the time when #3 was supposed to fire. So the misfire detection algorithm simply checks for an off-spec timing/speed measurement, and sets the code. Great. But the problem is that on some engines, like these Honda V-6s, that fault detection threshold can be triggered for other cylinders that are firing just fine.

Good new info here on the replaced head. That may explain some of the difference in compression measurements.

Easy mistake to make on the tube seals when doing the VCGs. I did the same thing on the rear bank the first time I replaced my VCGs (when doing a valve adjustment at around 105k miles). Discovered it, perhaps simlarly to you, at around 180k miles while tracking down a code-free intermittent rough idle that I wanted to get fixed.
 

·
Registered
Joined
·
2,983 Posts
The way this started was a ticking in the engine and a little roughness. I got a code that said EGR, so I bought a new one, tore the intake down to get to it, and thought, what the heck, I'll go ahead and replace all the plugs, since they are easy to get to. When I went to spin out the middle rear plug, it just spun. In short I accidentally found out the plug had backed out, and the threads were burnt off. The ticking was the plug sliding up and down every upstroke of the piston. The tube became part of the combustion chamber, and I had to replace that coil. My only option, as far as I knew, was pull the head and have it fixed. I did so. Bought an entire upper gasket set and put everything back together. Don't remember why I had to go back in, but when I next pulled the plug, it was covered in oil. Pulled the valve cover and found out that in my ignorance I'd just jammed that cover back on, and distorted the seal badly, which caused the leak. So I put 3 new seals in and carefully lubed them and slipped that thing on, making sure the seals slipped nicely over the tubes, which they did. I thought "problem solved". Unfortunately incorrect. I'm having a hard time believing I screwed it up again, and I don't see any oil at the actual top of the tubes where they meet the seals, so I'm trying to find out if it is possible for the oil to penetrate from the bottom of the tubes.

Concerning the coil possibility: Wouldn't the clean, uniform appearance of all the plugs belie the idea that one coil isn't firing? I'd love it to be that simple, but I don't see any evidence for that idea in what I saw when I pulled the plugs. Why would the obd show misses on all cylinders if it was just one coil? Not being argumentative, I am just seeking understanding. I'm way too old to think I know everything!
So, when you pulled that plug and it was covered in oil, I take it to mean that the insulator part was covered in oil and not the business end?
"Wouldn't the clean, uniform appearance of all the plugs belie the idea that one coil isn't firing?
So, I take it that the oil is on the insulator part, Is that correct?

Are you also using the recommended NGK spark plugs and not Bosch, etc?" Only pertinent if the bottom end (business end) of the plug is oil covered, or coated in something else.
Ever change or clean your PCV valve? Built up internal pressure can help force oil through the spark plug seals as well as forcing oil out of the rear main seal.

Buffalo4
 

·
Registered
Joined
·
8 Posts
Discussion Starter #13
On the coils stuff ...
In my experiece with these coils failing, they often initially do so intermittently. I never noticed any difference in plugs, since the car would not run enough in this intermittently misfiring condition to allow any plug effects.

On the multiple misfire codes due to a single failure - definitely the case. Observed many times, reported on here many times. I belive misfire detection is made by the crank sensor. It measures the exact speed of the crank as it spins past the sensor. In theory, if #3 (for example) misfires, the crank speed will not be as fast as it should immediately following the time when #3 was supposed to fire. So the misfire detection algorithm simply checks for an off-spec timing/speed measurement, and sets the code. Great. But the problem is that on some engines, like these Honda V-6s, that fault detection threshold can be triggered for other cylinders that are firing just fine.

Good new info here on the replaced head. That may explain some of the difference in compression measurements.

Easy mistake to make on the tube seals when doing the VCGs. I did the same thing on the rear bank the first time I replaced my VCGs (when doing a valve adjustment at around 105k miles). Discovered it, perhaps simlarly to you, at around 180k miles while tracking down a code-free intermittent rough idle that I wanted to get fixed.
I ended up realizing I was simply too emotionally invested in my previous work, and what I thought I knew to objectively diagnose the thing. I took it to two different mechanics (I may have mentioned that earlier) and both said it was time to ditch the van. But my gut told me that that was not really necessary, as the engine is solid. So I finally found a mechanic about 3 hours away that was willing to actually take the time and effort to DIAGNOSE the problem. I took it to him, told him the saga, and left him to it. He called back a few hours later, with a couple of key pieces of info: One: That brand new coil I bought was garbage (not cheap). I bought it from O'reilly, and would never have suspected it myself. He also said it's possible to install those seals around the plug tubes backwards, and they are directional. So he replaced the coil, got a new upper gasket set and replaced the valve cover seals (again). Haven't physically pulled the plugs since he did the work, so can't say for sure the plugs are staying dry, but the miss is gone, and the engine runs smoothly now, except the flat spot still rears it's head occasionally. Same deal: Drive the van for quite a while, shut it off, start it again slightly cooled off (maybe an hour later), and nothing happens when I push the accelerator. A quick pump on the gas and it'll return to normal. Wondering about the idle air control?

I want to thank everyone who contributed their ideas. You were pretty much right. This discussion helped me decide to spend a few hundred to save this van, rather than go into debt for thousands for another vehicle, which probably would have had it's own problems.
 

·
Registered
Joined
·
8 Posts
Discussion Starter #14
So, when you pulled that plug and it was covered in oil, I take it to mean that the insulator part was covered in oil and not the business end?
"Wouldn't the clean, uniform appearance of all the plugs belie the idea that one coil isn't firing?
So, I take it that the oil is on the insulator part, Is that correct?

Are you also using the recommended NGK spark plugs and not Bosch, etc?" Only pertinent if the bottom end (business end) of the plug is oil covered, or coated in something else.
Ever change or clean your PCV valve? Built up internal pressure can help force oil through the spark plug seals as well as forcing oil out of the rear main seal.

Buffalo4
I bought the most expensive set of plugs I could get. Not sure if they were NGK. It was like $80. Anyway, since the guy found the bad coil, everything has been fine, except the flat spot I described. And yes, I changed the pcv valve. Yeah, the oil was on the insulator side. Hoping that got taken care of as well with the changing the seals again. I appreciate your help and suggestions!
 

·
Registered
Joined
·
8 Posts
Discussion Starter #15
You could replace all of the coils, or when the idle is rough do a coil drop test and verify that 3 & 5 are bad and swap those out.
Cleaning out the carbon in the intake and EGR is common. My bad coil I replaced with a Hitachi that I pulled at the U-Pick yard. For plugs, best to use OEM.
I prefer to diagnose and verify the problem rather than throw $ at parts and hope for the best. Fix one and move to the next.
On my 99, got a code for the MAF sensor. Verified it was getting proper voltage, but resistance was bad. Got free sensor from junk yard. Soon after had rough idle and various misfire codes. Verified #5 coil was bad. Drove to the yard and bought replacement and spare from the yard. Verified both were good and replaced bad coil. Problem resolved.
Just out of curiousity: How did you verify the coils were good? Is there a bench test for coils? That's one reason I didn't try to diagnose the coils myself, as I don't have a diagnostic computer, and wasn't sure I could competently conduct a drop test with the motor running.
 

·
Registered
Joined
·
1,687 Posts
It is widely reported on this site (including by me) that an ignition coil on these cars is one item that should not be aftermarket. I don't know why it is impossible for others to succeed, but it just seems to be the case. Genuine Honda coils are made by Hitachi. So if your coil does not say Hitachi on it and come in a Genuine Honda bag, you are taking a chance with high probability of failure. Glad you got it straightened out.

True on the "coil drop test" being sometimes difficult to see when a coil is bad. Made even tougher if there is some other problem meaning the idle is not perfectly smooth. One thing I found (at first, by accident) is that laying a screwdriver (e.g., a 4" long, 1/4" tip slotted) on top of the plastic engine cover while it is idling will amplify any unevenness in idle. So even if the engine slight misfiring is not detectable by sound or feel, by looking at the screwdriver (just laying there on top of the engine) vibrating, it will move a lot more under misfiring conditions. Similar concept from way back in the day, was to detect a misfire by watching for vibrations in the radio antenna at idle.

NGK plugs sounds good. On those (on any spark plugs on any car these days) the main thing is to make sure they're not fake. Price sounds about right.

Nice work on keeping it alive. On the flat spot and your thoughts on IACV, the IACV is a definite problem area on these cars. Generally fixable by cleaning. Lots of good info on this site about that. From your description, I can't tell if it fits, so maybe search on here for reports on symptoms and tests. Briefly, general symptoms of clogged IACV are surging idle, surging RPM while driving, and dying at idle / unable to idle / unable to start unless foot is slightly on the gas.
 
1 - 16 of 16 Posts
Top