Car runs better with check engine light on??

[nitsujvg]

89 Nissan 240sx S13 SOHC (so other people can find this searching.)

Forgive me if this has been answered before, I tried searching but it's sort of a weird phrase that makes searching hard.

So after having the battery disconnected (any time after working on the car, I usually unplug it if it's electrical) the first time I start the car again, my check engine light will come on. What's weird though is that the car will run flawlessly. Starts up quickly, revs smoothly, no misfiring, idles at 700-750 constantly (after warm up), delivers power nicely... etc.

I'll take it for a few laps around the neighborhood, and the check engine light will turn off. (Not sure if it "learns" or something.)

Once the check engine light turns off, the car goes to hell! The idle raises to about 1800, tapping/feathering the throttle will bog the engine (keep feathering and it'll almost die), it has issues starting, sounds like it's misfiring.. etc.

When this first started happening, I pulled the ECU codes and got TPS, and MAF. I've since replaced both. For kicks I took both the old and the new to Autozone (they've got a really great diagnostic mechanic there.. this dude knows his stuff) and tested both old and new (not Junkyard new, but new new) and they both looked to be fine -- so I don't know what's up with that.

Any ideas guys? I'd love to just force the check engine light to stay on. I'm thinking possibly a bad ECU? Considering it runs fine when the check engine light is on, I doubt its my sensors..

Thanks in advance

[BCNR13-specD]

it still could be your sensors or how theyre wired, meaning bad/intermitent conection. the reason i say this is cause the obd II that you have in that car will go into "limp home mode" if its getting a bad reading from the maf or tps and will run off of a factory set memory of what it should be reading on one sensor if the other goes bad. this will trigger the light and a slightly rich mixture wich makes your car run great. but if your codes are telling you those things either one is bad/ bad connection or a bad adjustment on one of them, because you cant slap either one of those parts on and just go. you have to get out your multimeter and set it to Ohms and set to the proper adjustment. and the tps is very touchy, you can set it to what looks like the same position as the old one but be way off. I would start with the adjustments then if its still hapenning check the wireing. its most likely not the ecu because it has a self diagnostic and will throw a code saying its got something wrong.

[nitsujvg]

well the 240, being an 89 S13 is not obd ii -- whether there's a limp mode or not for my ecu, I can't tell you, but its definitely not running in limp mode. the sensors are also definitely good, and wired correctly. everything that had an even slightly questionable ground has been regrounded to rule that out as well. the tps is running at exactly 0.45 V 0% (when warm, 0.44V) and around 4.5V if I recall correctly (whatever it's supposed to be, it was at that a few weeks back when I looked it up). Both the new TPS and old TPS ran that way. They also both held steady on my 10 minute "test loop" I run around the neighborhood (through a neighborhood, a mile or so down a 45 mph road, on a freeway for two miles, then back home on another 45 mph road. -- just enough to warm up the car and check for any obvious issues.) by running some wires from the sensor to a DMM i've got in the cabin.

Also, the car is bone stock the obvious replacement parts over the years had all been to OEM specs

[BCNR13-specD]

your right about the obd 2, for some reason i thought that it was newer than 96 when the obd2 came out. but thats all i can think of so good luck, and if your computer is saying its the tps or maf then something is wrong there. i know you say that cant be it but it runs bad and the computer says thats it, so thats it.

[nitsujvg]

thanks for the help.

had it run on the sniffer, when the check engine light is on, it actually had better smog scores than before I had these issues a couple years ago. so I don't think it's running rich.

and like I said, if its running at seemingly full power (remember, it's stock, so a loss in HP is going to be pretty noticeable.) and the sensors, both OEM new and old, are running fine, then there is obviously not something wrong with the sensors in my book.

i understand that the ECU says there's something wrong, but the reason I've posted for help is because I don't see how it can run better under conditions the ECU thinks are bad, and now with data to prove it.

i'm sure it's something simple, but TPS and MAF it is definitely not. I think if the sensors were really busted, then there would at least be intermittent problems, or have the problems manifest when the check engine light is on.

[rioredstang]

If the check light is on then it should have stored some codes. What are the codes? If the cel is on the computer operates in open loop. It looks at all the sensors except the 02 sensor. It defaults to .45mv on the o2 sensor. You can back probe the o2 and watch to see if the o2 is working. When its warmed up it should swing from .25mv to .85mv. If the o2 is dead it will never get above .100mv.
Get someone to pull codes.

[nitsujvg]

If the check light is on then it should have stored some codes. What are the codes? If the cel is on the computer operates in open loop. It looks at all the sensors except the 02 sensor. It defaults to .45mv on the o2 sensor. You can back probe the o2 and watch to see if the o2 is working. When its warmed up it should swing from .25mv to .85mv. If the o2 is dead it will never get above .100mv.
Get someone to pull codes.

i can check the codes, I can do that myself no problem.
i'll check the O2 sensor -- thanks for the help!