plat

I have struggled with a stumble on acceleration for months, and more recently bad fuel trims (+ and -) and intermittent pinging in hot weather. Never any codes until today. The pinging simply stopped two week ago for no apparent reason and recently fuel trims have been closer to normal. Today out of the blue the car started running horribly, stalling, unstable idle, and threw P0300 random misfire code. It did this for about 5 minutes and went back to running fine for no apparent reason. I haven't been able to make it do it again with the scanner connected yet.

The freezeframe data from P0300 had fuel trim pegged at -25 on both banks, nothing else out of the ordinary. RPM was 2350.

I have recently noticed that the stumble almost always occurs at EXACTLY 2700 RPM, regardless of load or temperature or anything else. Sometimes it happens lower and then again at 2700.

Fuel trims are high or low consistently across RPM and load range, and are not dependent on engine temperature, only ambient temperature.


Now to my testing.

When I unplug the MAF, STFT pegs at -25, LTFT sits at 0.0, same on both banks. Idle hunts badly and the car is totally undrivable, stalls when you try to pull away or do anything. Then, when I plug the MAF back in, it stalls instantly.

I have read that if the MAF is bad, the car will run the same or even better with it unplugged. Mine is indescribably worse. Stalling when MAF is plugged has happened with the last two MAFs, no difference.



This is the current state of things:

• low fuel trims when cool outside (-15% to -5%)
• high fuel trims when hot outside (5% to 20%)
• stumble, like engine simply cuts for 1/2 second at 2700 RPM
• fuel trim pegged at -25 with MAF unplugged
• stalls when MAF plugged back in
• P0300 is only code it has ever had, from today

All of the basics I think have been checked or not suspect - fuel pressure, coils, injectors, plugs. This is something that it affecting the whole engine equally, not just one cylinder or bank.

It has had three MAFs in it the last few months and while none are the OEM part except the old one, I'm getting more convinced that I can rule out the MAF. It seems like something else must be wrong with the car for fuel trims to be so bad with MAF unplugged.

JvG

I'm going to guess that you might have a vacuum leak somewhere.

Does your scanner have a setting for engine vaccuum, or do you have a vaccuum guage.

If so, I'd like to know how many pounds of vaccuum you have at idle, and also your altitude above sea level.

This reminds me of vaccuum leaks in some of the old school cars I've had with carburetors.

plat

I'll check everything I can think of again for vacuum leaks and hook up vacuum gauge to it. I wasn't thinking that was likely since the trims also go way rich. Fuel trims correlate directly with ambient temperature.

Just to illustrate what's going on, these are typical trims when hot outside (top, 100*F) and when cool (bottom, 60*F). These readings were taken at 2500 rpm in neutral but they are about the same at idle and while driving.

LTFT is zero because the computer has been reset recently before these images were taken. If it runs long enough to start learning LTFT, then the total trims (STFT + LTFT) sum to the same STFT values that you see here.


Common fuel trim when hot outside, 100*F

Common fuel trim when cool outside, 65*F

plat

I'm getting 20" of vacuum at idle (650 rpm, a/c off). My elevation is 1000 feet. I plugged the gauge into the line to the boost sensor. Also, anyone know why my car would have a boost sensor? It is totally stock.








Also I found out what happened last week with the P0300 code. One or more injectors were sticking open. The same thing happened again two days ago, I had huge plumes of white smoke out the tailpipe that was raw fuel, trims pegged at -25, and I could actually hear the injectors clicking irradically - sounded like electricity arcing almost. Car ran horrible and stalled repeatedly. Half hour later it was fine again.

Does it possibly make sense that injectors in both banks would simultaneously stick?

JvG

20 inches of vacuum at 1000 ft indicates a healthy engine and no major vacuum leaks.

Seems that the injectors are operating normally , and are responding to commands by the ecu. The ecu in turn is responding to lies or inaccurate information from one or more of the sensors.

Does the car run differently if the air is low humidity vs after it rains? Sometimes a bit of corrosion on a sensor or a wiring junction can create electrical resistance. This can send inaccurate info to the ecu

plat

Unfortunately I haven't been able to test different humidites yet. I doesn't rain at all here from about april to november normally.

I have just put a junkyard MAF from a '99 and fuel trims are identical, so MAF is ruled out. O2 sensors were new less than 30k ago and I have tried unplugging them one at a time to see what affect had. I have heard that they can affect each other so if one is bad you could get weird readings/trims on both banks. But trim stays the same when I unplug the opposite bank's O2 and the downstream O2.


I have also concluded that IAT sensor input has no effect on how the engine runs, trims are the same with it unplugged/plugged in.

I also have tried heating the even bank injectors with a hair dryer to see if that causes the trims to change, like if they were sticking closed somewhat when hot and sticking open somewhat when cold. It has no effect.

What other sensors/factors could possibly affect fuel trims? I can only think of the crank position sensor but that really seems unlikely to behave like this.

JvG

I'm also out of guesses.

Seems that you have described the problem well enough.

plat

This is the same car that I had 14 ignition coils fail on, which turned out to be bad diodes in the alternator causing it to put out ac. With everything else ruled out, I can only think that the ECU was damaged by that in some weird way. But I still don't understand how temperature could affect that.


I am open to replacing the ECU if its cheap but I haven't been able to determine with certainty what the requirements are in terms of the year and codes matching, and then if it requires reprogramming by the dealer, by any shop, or not at all.

JvG

Thanks for reminding me of the coils failig due to too high voltage.

The reason I asked about rain and humidity is an experience I had with my 1992 Nissan pickup.
It ran quite well most of the time. It had no issues when there was no rain. But would like to idle at over 2000 rpm after rain or in high humidity conditions.
I'm in the Pacific Northwet.

I could not identify the problem, let alone the solution. It took quite some time to notice that the high idle only happened once the rainy season started. Our summers are bone dry.

I finally searched and found a factory service bulliton about this. The factory combined all the grounds to the sensors and attached them with a cheap copper staple. This would save a few cents or yen.

The copper staple would corrode over time. That would affect electrical resistance readings involving the sensors. The sensors would report proper information. But the staple would cause the readings to change before they got to the ecu. The ecu in turn reacted properly to the false readings the sensors told. The ecu believed that I was starting my truck in Frostbite Falls Minnesota during a cold snap. Which would cause super high idle.

Your car does not have that issue.

I'm just saying that something quite simple and strange could cause a lot of problems.

plat

What a strange issue, yeah like the kind of thing that’s simple and easy but you just don’t think of it because it’s so weird.

I think I may have something (a simple issue) on my hands as well.

I put in a junkyard MAF because I wanted to rule out the possibility that the non-oem MAF from the shop was no compatible/working right with the car.

I wanted to do this because the MAF readings correlate strongly with ambient temperature (high when cold, low when hot) and it seemed to be directly causing my trims to be both high and low, also varying with temp.

When I put in the junkyard MAF, it was really hot and the trims and flow were the same as with the shop’s MAF when hot, so i figured that wasn’t the issue.

But today i drove the car early with it cold outside and the trims are now still high, identical to when it was 100F the other day. And MAF reading is also identical to when it was hot.

So, the shop’s MAF was absolutely causing the trims to go negative in the cold. They may be high now because the junkyard MAF is a little dirty, so i’ll clean it and see what happens.

I just hope that the shop’s MAF was also causing what first appeared to be sticking injectors by sending really bad data intermittently.

DizzyEdge

It's used to measure the ambient pressure which can let the ECM know what the atmospheric pressure is outside. Not sure why it's labeled that way, maybe the same sensor is used to detect turbo boost on other models But it's a normal stock sensor. Also known as a MAP sensor.

JvG

One possibility we haven't considered is a malfunctioning ects. Engine control temperature sensor. It and it's corroded contact terminal will report a different reading than it should. The computer would adjust for the wrong temperature while it's getting conflicting information from other sensors.

You might want to check the contact, and perhaps remove the sensor to test it with a ohm meter at certain temperatures.