Hello everyone,

I'm obviously a forum newb with a persistent problem I'm curious to see if anyone can help me out with. We purchased the car with around 198k on the odo.

Around 4 months ago my GF's 1992 Accord automatic starting throwing an engine light. The car ran fine before, during, and after the light would come on. This was with a full tank of gas, an almost empty tank of gas, if the vehicle was hot or cold, going up hill, or downhill, cruising at sustained speeds, accelerating or decelerating both aggressively or mildly. Basically if the light wasn't one, we wouldn't know anything was wrong. It would show up after about 5 miles of driving, roughly.

I jumped out the diagnostic plug behind the passenger side footwell panel and it reported back code 43. I dug around the interwebz and found the most common issue was typically a bad 02 sensor messing with the readings the ECU receives. This was apparently especially common after 150k. This one has around 207k on the odo. I didn't have a OBD-1 scan tool, so I just went ahead and changed the 02 sensor while I was doing an oil change. I believe it is the Bosch sensor, which some people question.

It didn't change the engine light situation, so the next most common issue was actually in the fuel system pressures, typically being affected by an old fuel filter. I realized I'd done a fair bit of maintenance to the car, but had forgotten about the filter. I didn't have a fitting to attach to the fuel rail to read pressures, and since I figured it was old anyway, I went ahead and changed out the filter.

The car continued to drive perfectly normal, and the light stayed off. Great! For about a week or ten days. Not great. The normal driving condition lasted for probably another 3-4 weeks. Then, before the light would come one, while the car was still cold, it would lose power, chug and stutter. Apparently what ever is failing is getting worse. Once the light came back on, the car would run perfectly normal the rest of the trip.

I eventually took my fuel rail plug out, drilled and tapped it for a schrader valve style fitting, and could hook up my fuel pressure gauge to it. Driving around, the fuel pressure hangs out at around 32-33 pounds at idle. It moves to 35-36 pounds under moderate throttle, and tops off at 39-40 under full acceleration. From what I understand of the stock spec fuel pressures, it's supposed to stay between 35-41, so I don't believe this is drastically off the mark.

While driving around with the gauge hooked up, I observed the pressure during the chugging, and power loss issue. When I gave it full throttle thpe system would still read 40 lbs. of pressure while this issue was occurring. The pressures all seemed to remain the same while this issue was occurring. I tried a different fuel pressure regulator just for giggles, but the pressures remained the same.

Once the engine light comes on, it's like flipping a switch. The car runs perfectly normal. It's been behaving like this for around a month and a half now.

I'm trying to think of what in the system could be causing the issue to occur only when the engine is still pretty cold, and then immediately correct itself when the light comes on. It really is like flipping a switch, so it sounds like a sensor or something electronic, and probably related to the open loop/cold running circuit.

I recently checked the system again, and now I'm getting a code 15 (ignition signal) and code 1 (O2 sensor or circuit) along with a code 43.

Other than going back through the 02 sensor circuit, are there any other good ideas people might have? Thanks, and sorry for the long post.

 

Fuel pressure is likely OK and not source of problems based on what you've done.

Since you can't access system performance (LTFT, STFT, O2 voltages), you have to guess what might cause the symptoms. A Honda specialty shop or Honda service dept can probably access these and provide better diagnostics.

A partially blocked Fuel Injector could cause this performance. I had this recently on a Toyota Celica, where cold performance was rough but smoothed out upon warmup. What's happening is the ECM sees a lean condition due to the partially blocked Fuel Iinjector (FI), and compensates by enriching the mixture (+ LTFT) which smooths engine out. LTFT would run 25+%. In Toyota Celica, this produced obd2 code P0171. Cleaning FI's reduced problem and eliminated the P0171, however LTFT was still elevated (+10-15%). True solution is a new set of "balanced" FI's to produce even fuel distribution at same pressure. I paid a shop to perform diagnostics and they determined that FI's were "out-of-balance". They do this by measuring pressure drop over a few secs while rapidly pulsing each FI separately. There's a youtube video on this by scannerdanner.

Note, this is one possibility and difficult to confirm as fault. To "fix" you may need to remove FI's and clean each one individually. Cleaning in place is possible, but takes a special tool. Easiest cleaning is to add FI cleaner to fuel, but also least effective. Check youtube videos on diy FI cleaning, if you decide to try.

Code 1 is likely related to code 43. Code 15 is potentially another issue.

good luck

 

Seems like code 15 would be unrelated. That's the ignition signal from the ECU telling the ignitor to fire the coil. Look for broken wires to the distributor, or else a failed ignitor?

 

Thanks for the responses.

Yeah, I figured the code 15 was probably unrelated. It's not doing anything that I can see as far as missing or hard starting, or cutting off after the vehicle is warmed up like what would occur if it was a really bad connection or a failing ignitor. (Which is kind of like how the other problem started. Code but no noticeable symptoms.) I'm hoping to the fuel issue cleared up before tackling that one since it also seems like it will be a subtle issue as well.

TexHonda, I'd also wondered about the ignitors, but had ruled them out because I was thinking any flow issue would remain even after warming up. I didn't realize a Honda this old would have a computer intelligent enough to compensate for something that specific.

So, I have two other questions. What year was that Celica, out of curiosity? If it turns out it might be the injectors, is there a more modern injector that will work with the car's stock set-up? I ask, because if I can find them used from a newer car in a place like a pull-a-part yard, it might be easier to go that route. Likely low miles, and cheaper.

I'm also going to poke around for a shop manual pdf to see if I can find the testing process for the 02 sensor and circuit. I can check around with the mulit-meter and make sure it's not just that I installed a new 02 that crapped out quickly. I've heard sometimes that can happen.

Thanks again.

 

Part of the problem is the shop manual does not explain how the ECM manages fuel, however all cars that I've known follow the same basic fuel/air mix control strategy and you would expect that same problems as described for toyota celica (2001) could occur in early Accords.

However, there is no way I know of to extract info on the key fuel/air mix control data from obd1 system. Honda had a custom system available, and perhaps a few aftermarket systems had capability. Beyond reach of DIY'r.

That's why code 43 is so mysterious. It basically says you have a fuel/air mix problem but doesn't provide any details. My suspicion is it is very similar to P0171 obd2 code, but I don't know this to be fact.

Regarding injector repair/replacement, there are several possible fixes/states of repair; FI cleaner in fuel, remove/clean separately (see youtube videos), or remove injectors and send off to a facilty that will clean/balance them. They use UT cleaning action and flow test the injectors to insure they meet spec. There are several shops (ebay) that will take your injectors and perform this service for about $120.

good luck

 

I'll probably try to remove them and clean them myself first, to see if I notice any change. IF I do, then I'll press forward with replacement.

I suppose a more straight forward verson of my earlier question is, do you know of any more modern injectors that might give me something like better pricing due to more availablitiy, better spray pattern for more pep or better mileage, etc.

If I wind up having to replace injectors, is there something a little better than what it has now? I'm not planning on doing much more to this car beyond stock, but if it doesn't hurt anything, I don't mind updating some of the 20 year old components with more modern equivalents.

 

The injector is a solenoid-operated valve w/ a fine opening to create a spray pattern. I'm not aware of any special FI's that improve performance, but I'm not much into mods and other moderators may be able to help.

Something like the following should work nicely. Honda Accord LX DX EX 90 93 2 2L Fuel Injector Set Remanufactured w Warranty | eBay

Since no diagnostics are possible, replacing FI's as a fix carries significant uncertainty.

Suggest removal/cleaning of FI's is a good first step and not too difficult. New FI/fuel rail orings should be used w/ pre-lube to avoid damage.

good luck

 

I would suggest checking the harness connector at the shock tower and at the dist. For lose or corroded pins. You could also wiggle the wires in the connector to see if the engine miss happens. Your symptom are seem more of a global issue affecting all cylinders rather than a bad injector. Loose pins or corroded connectors can give you running issues related to temp and would affect all cylinders.

 

Check for an exhaust leak. You can use seafoam to do this. Look for white smoking coming out anywhere from the exhaust manifold to the oxygen sensor.

 

I rechecked the codes earlier, because code 15 just wasn't making much sense. I was right. It's code 14. Whoops. So perhaps something in the IAC system has started to go.

So, I'm still going to pull the injectors and try cleaning them this weekend, but I'm also going to hook up my diagnostic smoker and see if I'm managing to get a leak around anything. I'm not getting any surging idle/high idle issues, but it can't hurt to check.