Misfire on Start cold.... fine at Normal Temp.

[LunchBox04V]

Hey, I thought I would post to get a few conformations of my theories. lol Here is what I got going on. ('99 BK with the Gen 2. EJ22) Occasionally when I start the car from cold (generally overnight, and maybe 50% of the time) the car will stumble and misfire when under light load. The warmer the car gets, the greater the load it can take before it starts misfiring. Once at normal operating temperature, the car operates normally.

So.... I pulled the codes and got 2. (No misfire code....)
P1133 - Heater Circuit O2 sensor (Essentially need a new pair of O2 sensors anyway.....)
P0420 - Cat. efficiency. (Been getting this one, but thought it was for a bad O2 sensor)

I checked through my Haynes manual and they didn't cover things very well. I was planning on changing the O2 sensors (yes both, just get them both done, then I don't have to mess around with it again), and now the Engine Temperature Sensor (the one that goes to the ECU). I did a little searching on here and found this data on the P0420....(lol never learned to use the "Quote" function...)
================================================================
A code P0420 may mean that one or more of the following has happened:
* Leaded fuel was used where unleaded was called for
* An oxygen sensor is not reading (functioning) properly
* The engine coolant temperature sensor is not working properly
* Damaged or leaking exhaust manifold / catalytic converter / exhaust pipe
* Retarded spark timing
* The oxygen sensors in front and behind the converter are reporting too similar of readings
================================================================

Ok.... so here is the actual question.... Do you folks think that changing the O2 sensors (which was already planned) and the Engine Coolant Temperature Sensor with solve the problem? How could I check the ECT Sensor? Thanks for all the help. Oh, and I guess the ECT Sensor is a super cheap unit to buy.... I might just buy one and replace it... I heard like $15.oo. Thanks again.

[Huffer]

Change your O2 sensors... they will help greatly in proper running. You'll find that the PO420 will disappear after you replace the O2 sensors.
Reset your ECU also.

[LunchBox04V]

Ok, that was the plan. It was just odd when I did a little looking that I found that the Engine Coolant Temp Sensor might also cause this..... I was told by a subaru tech to do the ECT Sensor as well. Well good. Thanks Huffer, and I do remember that we were looking at the same car. lol She is still running strong, except for this little thing... But after 100k, timing belt, front main seal, and a few sensors, I think it was a good deal. haha thanks again.

[ouch1011]

Why are you replacing the ECT sensor? The part may be cheap but as I remember it sits under the intake manifold and is not easy to access. And you have a code for an O2 sensor heater and cat code? How does that become an ECT sensor prob?

I'd start with the O2 sensors. Should fix the o2 heater code, will probably fix the hesitation, might fix the cat code if you are super lucky, will save you from wasting money/time on the ect sensor that you don't know is bad.

[LunchBox04V]

The main reason I have been thinking there is a problem with the temp sensor is due to the timing when the misfire occurs. (engine is not throwing a misfire code, which is odd... But it is 100% a misfire, not hesitation). Again, when it is cold, the misfire is greater, and as the car warms it can take more load before the misfire begins. And once at normal operating temp it is gone. I agree with the fact that the ECT sensor is a bit hard to get too. But the car had been running with the other two codes for 2-4 weeks before this misfire problem started.

I was going to try and ohm/resistance test the sensor, but there is no data on this in my Haynes manual... Lol I guess if it has no resistance when checked, it is toast. Either way, the O2 sensors (yes both) will be done in the next 7 days. I will test, then change the ECT sensor only if a) it tests bad, and b) the Problem was not fixed by replacing the O2 sensors, and resetting the ECU.

[jey]

You could plug in a good OBD-II scanner and see what kind of actual voltages you are getting out of your O2 sensors. You're right it doesn't quite make sense - car is in open loop mode anyway when it is cold and that is when your problem is happening. It could be your ECT is bad, causing the ECU to not start out in open-loop like it should when it's cold, and then using bad/cold O2 sensor readings and then you get a misfire.

[LunchBox04V]

Lol exactly Jay! That was what I couldn't wrap my mind around. I should be in open loop and running off the map anyway. I "think" the ECT is faulty, making the computer think it is at or above operating temp, causing the ECU to run off the cold/bad O2 reading, further confusing the ECU. And I "believe" the problem comes up more when the temp out side is colder and more humid. Just trying to remember.... Thanks again though.

[ouch1011]

Have you checked your spark plugs and wires recently? Does it seem like only 1 cylinder is misfiring or more? Misfires that happen more cold or under load are often caused by the ignition system.