Runs bad until 3800 RPM.

[liflod]

This is my first post. I've been lurking and reading here for the past 6 months. Here's the history:
1990 Festiva LX, EFI, Auto, 118,000 miles.
Sitting since 1996 in garage.
Found burned exhaust valve #4 cylinder.
Replaced all valves,timing belt,water pump,gaskets etc....
Starts and idles fine.
Smokes from the exhaust(sometimes full throttle,partial throttle,letting off throttle)
Lacks power up until engine hits 3800 RPM, 1st or 2nd gear. Engine slowly surges but does not have power until magic number is hit on the tach.
Replaced coolant temp sensor, checked vane airflow sensor + inlet air temp sensor. TPS sensor switches are working.Checked ditributor for crud in the crank position sensor, checked compression 140 psi-150 psi,checked fuel pressure and regulator at fuel rail.

It runs very good with oxygen sensor unplugged. New oxygen sensor installed, no change in performance. Still runs like crap with it hooked up.

[rotinaj13]

if you disable the o2 from the system then you are disabling the computers ability to dynamically tune the fuel emissions and it runs on a general mapping. Now the last time i had issues with the car and it had symptoms like this I had a problem with the vane airflow sensor. Try it with the old one if you still have it. It might also be a defective o2 sensor, does it have a warranty?

[liflod]

I have not changed the airflow sensor. I removed the cover, checked the resistance of the temp sensor and verified the voltage changed with the amount of air flow. I also moved the vane by hand and was able to get the engine to bog down either way indicating the air flow sensor was commanding the computer more or less fuel. There was no difference between the factory O2 sensor and the new universal from Advance.

I'm wondering the significance of the magic 3800 RPM. Its like hitting the "run correctly now switch"

[Banana Bonanza]

I am not the person to comment on fuel injection. But computerized systems with O2 disconnected tend to run very rich in "limp home" mode. So since yours runs good in this mode, I would suggest you are running very lean in normal mode. When you get to high rpm, computer has to provide more fuel thus it runs better.

Always looking for physical rather than electronic reasons, I'd first look for major air leaks. Spray some starting fluid around perimeters of air intake manifold where it connects to any other surface with a gasket or seal and see if any significant rpm change. Again I am not fuel injection person and am just luddite carb guy, preferably old all mechanical carb without feedback system. Those I understand and can tune.

[liflod]

No major vacuum leaks or none detectable with carb spray. Seems to run better with more timing advance. Disconnected fuel pressure regulator and no change noticed. Also disconnected and plugged idle control hoses and no change.
Does anybody have a detailed test procedure for the airflow sensor? I have the top off and can see the contacts and put a meter on it to see the change in voltage with air flow. I have the Haynes manual, but it doesn't go into detail.

[liflod]

I replaced the smoking engine with a lower mileage engine. The engine runs fine, no noise, no smoke and it finally passed the emissions test.
I still have the drivablility problem.
It does not have much power below 3000 RPM and it gains power about 3300 and surges until it hits the magic 3800 RPM and it runs great.
I'm thinking the timing advance is not working correctly. It is a 1990 with fuel injection and the computer controls the advance. I will check it tonight with a timing light and see if it is advancing.
Does anybody know when the engine goes to max timing advance? Could it be at 3800RPM where mine runs great?
I'm going to put a scope on the distibutor at lunch to see if the cam position sensor output looks good.
Any other suggestions?

Lance

[FestYboy]

double check your cam timming, i'd be willing to bet it's a tooth or 2 retarded, hence the high RPM power.

[qroger]

Any chance the ecu is toast?
qroger

[FestYboy]

^unlikely due to the fact that it still runs... (ecu controls fuel timimg/durration and spark timing.)

[liflod]

I runs the same with the original engine that I fixed the head and the engine that I just installed.
Its a little bit colder this morning and it ran very good on the way to work. I checked the ditributor signal and the timing advance last night. I get a very smooth advance up to about 25 degrees and the RPM signal from the distributor looks good.
I'm going to see if it runs worse when it is warmer.

[liflod]

It was a tad warmer this morning and seemed to surge as usual.
Does any one have some specs on the air flow sensor? I'm going to check all of my sensors/inputs again to see if I notice anything odd.

Who knows the oxygen sensor system in detail? I get nice steady DC voltages with the sensor disonnected and they vary when the wire is connected. The DVM jumps around alot but the value falls within the .1 to .9 volts called out by the Haynes manual.

[FestYboy]

the O2 generates it's own voltage in responce to the amount of o2 present in the exhaust. the varying voltage is there to keep the cat happy (3 way cat requires the car to cycle between rich and lean to provide the catalyst with enough oxygen to do it's job and enough fuel to keep the temp up after light off) i'm pretty sure that's how it goes, but wikki might explain it better.

the steady DC is in responce to the ECU not getting any feedback from the O2 adn in turn going rich to keep things safe.

[doylerl]

Could a clogged cat. conv. cause this? Higher RPM send more exhaust pressure to blow past/through the clog? If both engines do it, my guess would be either the computer, or another part used on both engines, such as the exhaust system.

[Team Lightning]

When the Cat Con on my Dakota clogged it would rev like crazy till about 2500 then fall on it's face. Had much more low-end torque but no top end.
Jerry

[liflod]

The clogged cat was brought up by a friend of mine. I'm not looking forward to droping the exhaust to take a look inside. When he had a cat fail, the car would not start.
Fuel mileage is about 28 MPG but it starts easy, idles great and runs pretty well on the highway, its the lugging up a hill about 40 MPH that really makes it surge.