I have an 87 Duke auto 65,000 miles that stumbles upon light acceleration when it is cold. After 2-3 minutes of driving it is fine. It has no cat and in the last year I have put a new head gasket on it, rebuilt the throttle body, replaced the fuel pump, IAT, TPS, ECM, MAP, O2 sensor, plugs (which are gapped properly and are NGK coppers), wires, CTS, and all the vacuum lines. I have tested the coils and the ICM, the injector has a nice spray, and I am going to change the fuel filter just for fun. I am basically out of ideas. Please help!
cap and rotor? might be a mechanical issue in the head or something where the valves arent seating properly until the engine gets to temperature? or maybe check your tbi for how it is spraying when cold? maybe something is going on with that.
It has a DIS so there is no distributor, I had the head reconditioned when I did the head gasket, and I've watched the injector both on cold startup and when it is warm there are no issues there..
It could possibly be the ECM. I had an ECM in my 88 coupe (duke engine) that would not work cold. On a cold day, if you took a hair dryer and warmed the ECM up (center console removed to allow access), it would run fine. If not warmed up, it ran badly, or wouldn't start at all.
It has a DIS so there is no distributor, I had the head reconditioned when I did the head gasket, and I've watched the injector both on cold startup and when it is warm there are no issues there..
I bypassed the EGR I replaced the wires a few thousand miles ago, for some reason I have a hard time believing the ground would only be faulty when the engine is cold...
My vote is the ECM not operating correctly in OPEN loop. The O2 sensor sets ECM in OPEN loop when cold and CLOSED loop when warm. If you can exchange the ECM with a known good one that would verify this. Go to Auto Zone, Advance Auto Parts or Oreilly Ayto Parts, they will scan the ECM for free. O2 sensor could also be sedning unstable signal when cold causing ECM to bonce around between OPEN and CLOSED loop until it warms up. Check the O2 sensor connector also. I am new to the Fiero so this is just my 2 cents, hope it helps. Let us know what you find when it is fixed. MIKE
I have 2 ecm's it hesitates with either one. Just for fun I unplugged the o2 sensor it acted the same way except worse, which leads me to believe it only does it in open loop. are there any other sensors the computer only reads during open loop?
You mentioned the stumble is in light acceleration when cold. A possible reason it might not stumble under heavy acceleration is because that would cause the ECM to enter "power enrichment" mode, where it's programmed to go extra rich.
quote
Just for fun I unplugged the o2 sensor it acted the same way except worse
If you do this it will stay in open loop permanently. When you do that, does it still stop stumbling or at least improve significantly when it warms up? If it does, then maybe closed loop isn't really what fixes it, and everything I wrote after this doesn't apply.
My guess is it's running a bad fuel mixture in open loop, then it corrects itself when it gets into closed loop and starts seeing feedback from the O2 sensor.
quote
I have 2 ecm's it hesitates with either one. Just for fun I unplugged the o2 sensor it acted the same way except worse, which leads me to believe it only does it in open loop. are there any other sensors the computer only reads during open loop?
In open loop it still reads the same sensors as usual, but the difference is it can't verify the A/F ratio due to not having a working O2 sensor. So during open loop, the fueling behavior is somewhat blind. It relies on preprogrammed assumptions about the engine's airflow characteristics and uses some sensors at the intake side, but it can't see the actual results in the exhaust until the O2 sensor comes up to temperature. So if the calibration has been thrown off, it won't figure that out until it warms up.
Why would the cold calibration be off? That's the hard part. Outside of mechanical issues, I'd try verifying proper operation of the temperature, MAP, and throttle position sensors. I know you've already replaced them though, so this is probably not it but I'd feel compelled to check.
The CTS and IAT sensor can be checked by measuring resistance across their terminals. I'm assuming 4cyl is same as the V6 here.
TPS can be checked by watching the resistance (or voltage, if connected) gradually change as you move the lever.
The MAP sensor is hard to test, GM gives basically no guidance on that at all. Make sure you don't have a vacuum leak messing up the MAP's readings. Since there's no check engine light, you don't have a dead sensor but they could be "off" a bit and the ECM wouldn't know.
Also check the +5V power feeding into the MAP and TPS. With the engine running you should get a steady 5.0V, and it should be very close. Mine was 5.00-5.01 when I last checked it. This is a regulated output from the ECM and it shouldn't jump around.
With all the stuff you replaced it's hard to think of much, but it seems that something is throwing off the tune until the O2 sensor feedback corrects the situation. I don't know if it's possible to datalog an 87 Duke but if it is, it might help you.
quote
I have a hard time believing the ground would only be faulty when the engine is cold...
I'm not really sold on the grounding theory, but it wouldn't be unthinkable. The actual problem might be very hard to imagine, it could potentially be anything that messes with the mixture. Once the O2 comes online, the computer gains a better ability to compensate for it.
Wow +1 to you thank you for the detailed description! With the O2 unplugged it still ran like crap warmed up.
quote
My guess is it's running a bad fuel mixture in open loop, then it corrects itself when it gets into closed loop and starts seeing feedback from the O2 sensor.
This sounds exactly right. The Map sensor I swapped from my GT (which runs fine) so I know it's good and it also has a new vacuum line on it. The TPS had 4.99v on one wire and I forget what the reading was on the other wire but it was in mv and went up as I increased the throttle (does anybody know what this reading should be at idle?). I also checked the resistance with the engine off and did not find any dead spots or get an OL reading. I checked the resistance of the original CTS and IAT and they were good but I replaced them anyway when the new ones didn't fix it, I verified their resistance as well. I will find out what the output range for the TPS is and check that again.
Yeah, it sounds like the TPS is fine. So based on what you tested with the O2 unplugged, it sounds like closed loop feedback from the O2 is what's making it start to run better. If it was some other unrelated heat-related issue, then it should have still improved after heating up even without the O2.
I wish I could think of something significant that you haven't already looked at or replaced. Does that 5.00v (4.99v) power supply to the sensors stay steady even with revving of the engine? I'm grasping at straws there, probably not it. But if that voltage isn't steady, the readings that come back from the sensors won't be steady either.
Open loop normally runs somewhat rich. Since heavy throttle is okay, it sounds more likely that your problem is running lean. That would mean that for whatever reason, it's needing a longer fuel pulse than the ECM thinks it should. Somewhat low fuel pressure or flow rate might cause this. Another possibility is it's not getting the "pump shot" ie "acceleration enrich" that it should when you move the throttle, but I don't see why that would be a problem as long as the TPS is working. If possible, I'd maybe try to hook up a fuel pressure gauge that you can see while driving, and see if anything happens to the pressure when it's acting up.
When you say you bypassed the EGR, do you mean it just doesn't have a vacuum supply, or is it completely out of contact with the exhaust stream? I don't remember the 4cyl setup real well. But if you had a constant slight leak at the EGR valve, it might confuse things. In that case I think it would run rich, not lean, because the unexpected exhaust would displace oxygen so it would need less fuel than expected.
[This message has been edited by armos (edited 06-23-2012).]
The 5v is constant, I made a block off plate for the EGR so there are 0 exhaust gasses going into the intake. What is the best way to hook up a gauge? I have a tool but since there is no schrader valve, and I don't want to cut the steel line, the only thing I can think of is to replace the filter with fittings somehow...
