Hello everybody,

Last December we bought a used Accord '98 3.0L V6 with automatic transmission; the car ran fine and seemed clean, well maintained, and had 165K miles. After about a week the engine started to stutter and hesitate out of the blue especially between RPMs of 1000 and 2500, very similar to other posts. An initial inspection showed that the rear O2 sensor was bad so we had it replaced after which the car ran fine again. However after a few days the problems reoccurred out of the blue.

Same symptoms that is pressing the gas pedal would result in no acceleration and the engine would almost die. Pressing the pedal harder would at one point make the accord lurk forward, however, and pumping the pedal would at least let us drive home. We also noticed that sometimes, just stopping, e.g., at a stop sign (without shutting of he engine) would re-enable the engine and everything was back to normal.

After more reading, changing the front O2 sensor and having the EGR cleaned and the EGR kit replaced (P1491) and the problems not going away I connected an OBD scanner to the interface to maybe see some irregularities. Unfortunately I am no expert with these scans, I did however notice that the short term fuel trim went always deep into the negative numbers -27% when the symptoms occurred. Below is a link to a spreadsheet with some freeze frame data that I took when there was a problem. The PDF has a graph with some normal operation followed by a sudden drop of the short term fuel trim value, which seems to be the culprit, at least indirectly.

The other very strange observation (looking at the PDF) is that there seem to be two modes of operation for normal operation: One where the sensor data fluctuates wildly (good operation) and another one where the sensor data is pretty constant (good operation). The problems occurred in between (abnormal operation when the STFT was low).

When reading other posts where the problems seem to have gone away by cleaning ports, changing plugs, wires etc. it seems a bit different with our Accord because the symptoms occur not consistently but rather randomly.

Could it be that the ECU is bad? Or what exactly is the driving element that manipulates the short term fuel trim? - I was going to reset the ECU (disconnect the battery for a couple of minutes, right?) tomorrow to see if this helps, but I am sort of at a loss of figuring out what else the the remedy could be.

The Accord is otherwise a great car but without that "reliability" factor it is of not much use, especially because we bought it primarily for our sons to learn how to drive and make their way (safe!) to school.

Any help or suggestion are appreciated and I surely will post any results back to this forum.

Thanks a lot!
rcabl

http://www.rcmodeltools.com/pub/Documents/Honda98V6.xls
http://www.rcmodeltools.com/pub/Docu...ODBIIChart.pdf

 

Maybe a fuel pressure test?

Then an injector leak down test?

 

Thanks a lot for your quick reply HondaDude. - I can certainly try to check the fuel system over the weekend and had been thinking about the fuel pump being as suspect too. However what baffles me is that sometimes it completely works and then out of the blue (often after a full stop) the problems start and then again out of the blue the car begins working again.

Maybe I haven't used the best heading "Fuel Starvation" because "Intermittent Hesitation" would have worked as well.

Do you know what controls the short term fuel trim? Is this something that is directly controlled by an ECU or is the value just measured by a sensor?

Thanks again!

 

Short Term Fuel Trim (STFT) is managed by the ECU and a negative value is a response to a measured (O2 sensor input) rich condition. Hence HondaDudes suggestions to check fuel pressure and injectors. Check stealth316.com/misc/obdii_fuel_trim for more info on fuel trim.

One check that is quick and easy is to remove the vacuum reference line from the fuel pressure regulator and sniff. If you smell gasoline, the fuel pressure regulator diaphragm is ruptured and allowing unmetered fuel to enter the engine. This seems unlikely since problem is intermittent, but its easy enough to check.

good luck

 

Cool, thanks a lot for your explanation of this TexasHonda - I am going to check out the fuel supply over the weekend and go over the OS2 sensor - STFT relation and then and get back to you with any results.

Thanks for your pointers!

 

Since my last post it has been a while, and like in most cases "no news is good news", that is, our Honda Accord works again like advertized. - As promised here are the steps that finally solved the mystery.

After following all advice given from friends, in the previous posts, gathered from the web, and seeing only random and temporary improvements, I finally gave up and sought help with the local Honda dealership. After describing the problems and symptoms they came to conclusions similar to those in the posts above. The first suspects were the O2 sensors, which were OEM and apparently not as good as Honda brand name sensors. So these were changed and for two or so days it looked like everything was fine. But then the symptoms returned. I waited another three weeks or so to verify the problems and to find a pattern. I found some sort of pattern but it was very random. There were stretches of two to four days where everything just worked and then there were times when the problems started about when the engine reached operating temperature.

Another trip to the Honda dealer. I left the car there for a week so the techs could drive it around. They had the computer hooked up for 400Miles but couldn't find a suspicious reading. After that we finally decided to put the car's computer on the suspects list and although a new computer is not cheap (~$750), this seemed the only alternative left. Because these ECUs are very specific it has to be ordered from somewhere in the Midwest I believe.

The dealership was very fair and professional and did not charge me for any of the labor involved in testing the car but only for the ECU and installation (They were probably also happy to see me driving away with a working car :-)). Anyway, since changing the ECU our Accord drove without a glitch. All the symptoms are gone; idle is at times rather low but that is just an observation but most important of all, the hesitation and starvation symptoms are gone.

I hope this helps others who have similar symptom with their Accords. Also thanks again TexasHonda and HondaDude for your help, I really appreciate your taking the time to answer all those threads and questions.

Cheers!
Andreas

 

There have been several incidents of ECU failure recently, thankfully fairly rare event.

Thanks for responding w/ the fix.

good luck

 

I also have a 1998 V6 with similar issues. It seems to happen as soon as the car reaches operation temperature. I do have a bad wire on my catalytic converter o2 sensor harness but no issues with that for a long while. I am also experiencing symptoms of engine bogging at or around each shift rpm or when taking turns. Everything you've replaced has been replaced on my car plus many other things since my car's mileage is almost 270k. I have not tried to replace my ECU but I do have an ECU from a 2002 lying around and I will have to give that a try seeing that seems to have helped you out. Wish me luck.

 

The fuel filter should be replaced 1) every 4 years or 60,000 miles, whichever comes first, or 2) whenever the fuel pressure drops below 40-47 psi, with the fuel pressure regulator vacuum hose disconnected and pinched, after making sure that the fuel pump and the fuel pressure regulator are OK. If you need any advice or help please contact lemon law attorney..

Thanks,
Indiana Lemon Law

 

I have been having the same problem with my 93 Accord ever since I bought it. I was hoping that I wouldn't have to change the ECU. Did you also have a problem with the tach jumping? Mine has this problem very randomly. Any info that you could pass my way would help me tremendously.