acceleration flat spots

[203Cree]

Cat is fine. It's new 10k miles ago and I took it appart as my first guess. I also know it's not my ignitor. I've switched them between 2 other running hondas and never had a problem. My problem HAS to be with the O2 and its wiring. The car has problems with known good sensors. However, as soon as I put in a known bad sensor (has a steady voltage out of .139V on 2 other cars, I get a steady voltage reading of 1.153ish volts in the prelude) the car no longer has issues. This is because (my theory anyways) the ECU stops looking at the O2 for a reading and subs in its assumed values. These values let it run well enough but doesn't get any of these weird flat spots.
I've got to have a voltage leak somewhere in my harness. The suspect bad sensor that I had in my prelude would give weird readings (up to 10V), but as soon as it's put in another car the voltage readings are perfectly normal. Once more, known bad sensors will give different steady readings in my car. WTF is wrong with that? I don't know, but the car is driveable right now.

Well, sounds like you have a plan of action. I've never heard of a computer doing what you've described, but then again I haven't seen it all by any means, lol. There's deff. something not right and it sounds like the 02 is about the last thing to do. It really should default to open loop mode tho, I'm not sure why I wouldn't. You should be fouling plugs and puffing smoke at WOT, but maybe Honda ECU's don't work that way, lol.

[preludeman92]

Nope, no fouling, no smoking. It'll backfire sometimes or maybe it's some nasty knock?!?! I've never really heard true knock, just some random pinging that one of our old cars used to do unless we used Shell gasoline.

I'm waiting for my JAW from 14point7.com to show. I'm thinking it'll be here thursday/friday. I'll post a review when I get it. It looks like a very promising alternative to expensive stuff.

[19accord92]

If you're 02 sensor or wiring was bad it would automatically default the computer to open loop mode and make the car run on the rich side as a fail safe measure. So unless you're puffing black smoke like it's going out of style, I doubt it's that. Not to mention you'd get a code for it if it wasn't sending the signal it should.

Tell ya what, I have two theories that maybe you haven't looked at. Number one is the ingnitor box in the dizzy, number two is a possibly clogged catalytic converter. Normally if the cat is clogged it will produce too much back preasure in the higher rpm bands and cause the motor to fall over on it's face. You've got access to a spare car, try swapping the mid pipe. As for checking the ignitor, I'm not sure how to do that. You may also check your IAC valve. That could cause a lot of the issues at idle you mentioned. Without a PCV delete these things dump oil into the intake track like it's free.....

I'm interested to see what you find.

well when i took it to get emissions tested, it was runnign really rich... no ecu code though.

[preludeman92]

Kinda of update.

I visited my prelude friends this weekend and we went drag racing. I didn't race mine but I did use a 98 prelude to go dragging. On the way back down the return lane on my return run my passenger (the owner) asked if I knew what that weird jerk was. It did the exact same thing as my car does. It would do it most of the night when we weren't on it full throttle. We also extracted a code from the computer (thank you OBDII for giving me more information than my OBDI computer). The code was for an excessively rich mixture. The car has also thrown them for excessively lean too.
Now, keep in mind the car has thrown both codes...
The car is on it's 7th engine. The other ones have spun bearings because the oil got too thin. The oil got too thin because of how much fuel the car was dumping into the motor trying to get it to burn richer (this is when it was throwing codes for lean). The fuel would degrade the oil of all lubricating abilities. The first thing we checked for when motor #4 went was the O2 sensor. We bench tested it and it was fine. The same sensor is still in that car 3 motors later, and they've all died of the same symptoms. I bench tested it again yesterday and got "normal" readings. I tested it in the car at the computer and got similar readings to the readings from my car. We tested it a little bit directly at the sensor without the wiring/heater hooked up and we got normal readings. That tells me there is something wrong in either the heater circuit or the wiring harness. Since mine is/has run 2 other hondas perfectly fine for the last 2000 miles, my guess is there is wiring messed up somewhere.
Conclusion, the car is messed up, not the engine, sensor, or computer. More later.

[HondaFan81]

Kinda of update.

I visited my prelude friends this weekend and we went drag racing. I didn't race mine but I did use a 98 prelude to go dragging. On the way back down the return lane on my return run my passenger (the owner) asked if I knew what that weird jerk was. It did the exact same thing as my car does. It would do it most of the night when we weren't on it full throttle. We also extracted a code from the computer (thank you OBDII for giving me more information than my OBDI computer). The code was for an excessively rich mixture. The car has also thrown them for excessively lean too.
Now, keep in mind the car has thrown both codes...
The car is on it's 7th engine. The other ones have spun bearings because the oil got too thin. The oil got too thin because of how much fuel the car was dumping into the motor trying to get it to burn richer (this is when it was throwing codes for lean). The fuel would degrade the oil of all lubricating abilities. The first thing we checked for when motor #4 went was the O2 sensor. We bench tested it and it was fine. The same sensor is still in that car 3 motors later, and they've all died of the same symptoms. I bench tested it again yesterday and got "normal" readings. I tested it in the car at the computer and got similar readings to the readings from my car. We tested it a little bit directly at the sensor without the wiring/heater hooked up and we got normal readings. That tells me there is something wrong in either the heater circuit or the wiring harness. Since mine is/has run 2 other hondas perfectly fine for the last 2000 miles, my guess is there is wiring messed up somewhere.
Conclusion, the car is messed up, not the engine, sensor, or computer. More later.

Interesting. Keep us posted, very good, way to be persistent.

[19accord92]

i knew it had to do something with running lean or rich.... mine used to run extrmely lean ( wrong fuel injectors ) then i changed them to the right ones, and ran excessively rich ( dont know why. ) then i added a fuel pressure regulator, which has eliminated most of the problem. i keep it at a 32 psi range. I am afraid any lower will make it to run lean...

[preludeman92]

Another kinda update.

Beautiful thing about OBD2. You can plug it into the computer at honda and see everything the computer is doing and is trying. The other car I talked about just got another new motor like 2 weeks ago, so it's owner is trying to make sure it doesn't die. He took it to work (he's a honda mechanic) and pluged it into the car. The car showed a primary O2 voltage of 0.000V at idle and 3000 rpm. The ECU didn't think this was a problem though, it didn't even have a pending code. The ECU also has flexable fuel map with an enrichment table you can read with the right software from honda. My buddy says hondanet says it should be 10% +/-5%. When he reset the ECU right before the diagnostics hookup the computer started at 5%, in the 5 minutes it he monitored the car the enrichment was up to 29%. He tested it the next day at work and found the enrichment was over %50 at idle. The injectors were running at about 70% duty cycle at idle! We're thinking that the car is trying to use the secondary O2 sensor and MAP sensor to attempt to run the car ok. It unfortunately isn't working so it runs way rich to protect itself from detonation and lean conditions.

He is going to go unplug the sensor tomorrow while it's on the diagnostics to see what the ECU does. We're hoping it goes back to the safe maps honda programed in for the case of a CEL.

[19accord92]

any updates?

[preludeman92]

Not really much of an update.
I'm still waiting for my lambda sensor to show. The other car went to the tuner before my buddy could take it back to his shop. Actually, his girlfriend took it over one day without him knowing (they have 5 cars to share, so they can go weeks without driving some of them).
When the tuner heard what it was doing he knew it was O2 sensor related, but had no idea what the problem actually was. So the car got a tune that ignores the oxygen sensor readings. I'm mad that he tuned the problem away instead of actually fixing it. Whatever, it's his car to blow up again.

[19accord92]

so o2 sensor problem is the flat spot? if so i should probably change mine out...

[preludeman92]

It's appearing that the O2 sensor is the root cause of the problem. My friend's car has not had an issue since it's been tuned away, and it went to the track this weekend. My sensor should be here thursday and I'll put it in with my JAW on friday afternoon so I can monitor the actual A/F ratio when the hesitation occurs. The JAW also has a narrow band (stock) O2 output, so I'll just use that if it comes down to it.

[preludeman92]

Confirmed the O2 sensor is the problem. I used the JAW to datalog the stock O2 reading in volts vs the actual AFR. The reading the computer saw was always over 2V when the stuttering would happen. The computer doesn't reconize outside of the 0-1V range the sensor is supposed to produce. The JAW would read between 21.4 and 22.0:1 AFR. No wonder why it looses all power and starts to knock and backfire! It's detonation from super lean conditions!
From what I've read, 0V is lean, 1 V is rich (.2-.8V is the range that is supposed to be produced).

If I had to guess, my friend's prelude with a constand 0.000V reading was confusing the car. 0V is lean, 1 V is rich (.2-.8V is the range that is supposed to be produced) The car thought is was super lean, so it kept richening the mixture to try to get the reading to budge. It would get so rich the thing would end up killing itself. I have a feeling that the car eventually would read something and would go super lean because of the sudden information. I don't know. I went to use the JAW as the NB sensor but saw the readings are backwards from everything I've read. Going to contact the designer tomorrow.

[F20A CB7]

Whats the compression of this motor in your lude

[HondaFan81]

FYI: I replaced the TW coolant temp sensor and stock MAP sensor in my red Accord and it appears to have fixed the problem. Once I have driven a week with no problems, I'll consider it fixed and will swap back the original MAP sensor to see which of two sensors fixed the problem. I'm leaning toward me having a faulty MAP sensor. Both sensors have 255,XXX original miles.

Do you think mounting stock MAP sensor on firewall with a car that idles at 1200 rpm and rougher due to an aftermarket camshaft will damage it from vibrations?

[preludeman92]

I wouldn't think the MAP would be damaged from vibrations. However, the difference in vacuum from stock may have something to do with a shortened life. I've tried 3 different MAP sensors on my prelude, and they've not changed anything.
Don't you love having extra parts just laying around so you can swap things around?
I had the brake MC, alternator, and PS pump all go on my accord within a week, and I never went to the parts store for anything other than fluids.


Oh, the compression on my lude is the stock 8.8:1 found on every f22a1 out there. Nothing special in the car... yet.