I6 4.0 1991 bucks/surges on med/hard acceleration

Installed a new O2 sensor, new exhaust headers, cleaned the intake manifold and intake and exhaust ports on the block.
Problem persists.
I posted a more complete problem on another forum here
http://www.edmontonjeep.com/forum/viewtopic.php?t=2818&postdays=0&postorder=asc&star t=0

In summary:
With everything plugged in correctly, under normal environmental conditions, the engine runs excellent, when first started for the day.
Once it has time to warm up a bit (~5 minutes) it starts to buck and surge during hard acceleration. When i unplug the MAP sensor electrical connector, the problem goes away until >4000 RPM.
Fuel mileage is very poor. ~2-4 KM per Litre.
When it starts bucking and surging, the check engine light flashes on and off about 3x a second, which is about how many times it bucks and surges a second. (i think the ECM makes 3 adjustments a second.)
It is difficult to push through the bucking, but eventually I do get up to highway speed, where I can cruise normally, but the problem comes back every time i hit the pedal.

When cold, with MAP plugged in, or when hot, with MAP unplugged, the engine seems to have lots of torque, and feels fairly close to the stock ~195 HP.

No trouble codes other than the begin sequence, ECM reset within 50 starts, and end sequence.

I've replaced the O2 sensor, MAP sensor, exhaust manifold (swapped for performance header), exhaust/intake gasket.
I also cleaned the intake manifold, and intake and exhaust ports on the block. Ran through a can of sea foam, checked plugs, wires, distributor.
Have not checked stator.

All sensors report within spec, other than the TPS variance is a bit low ~800mV closed to 3.9V @ WOT.

All the tests outlined in the haines manual, report that all sensors are at least not completely broken. Brand new O2 and MAP sensor installed.
No more engine codes under normal conditions. Codes are reported whenever i disconnect something on purpose. They all clear gracefully when I plug things back in.

The problem ONLY happens during Closed Loop (with ECM using sensor data) operation, and always recurs when the engine is warmed up. The problem always goes away when I disconnect the MAP sensor.
If I let the engine cool off overnight, I can plug everything back in, and it goes again, for about 5 minutes with no problems.

The only test I haven't tried yet, is hooking the ECM up to a scanner (can't get ahold of one in town).

Any ideas?

how old are the plug wires. is the distributor indexed properly?

check plug wires and distributor?

Sounds like a dirty/plugged fuel filter.

All those answers would make sense if the problem happened regardless of hte MAP sensor being plugged in.

The only thing the MAP sensor can directly affect though, is timing. and given the map sensor is doing its job correctly when cold (engine runs great cold, with map plugged in, but poor cold with map unplugged) and then makes the engine run like crap when warm (again, once warm, unplug the map sensor, and the problem goes away)

I've tested spark leads, and plugs, distributor is advancing and stator has to work or i don't think the engine can start.

Timing is deadly when the engine is cold, at any rpm, and deadly when hot, but only at idle.

Timing goes to hell when hot, with map plugged in. looks like lots of misfires on the scope.

I'm suspecting the CPS right now, but only because i really can't possibly spend enough money on this problem lately
Goona pick one up tonight.
Any other suggestions?

Sounds like you got a handle on it, let us know what you find out.

Sorry, I didn't intend any offence. I've just already tried all that, to no avail.
The previous owner had this problem as well, and 6 shops he took it to, all said the same things, but none could actually fix the problem.
The fuel pressure has been verified as acceptable, the distributor cap and rotor were replaced, spark leads and plugs replaced, O2 sensor replaced, MAP replaced, ECM replaced, TPS verified, swapped TPS with a known good part, and no change, IAC verified, disconnected CAT from pipe, no change.

I've tried all the mechanical routes, although i haven't yet checked valve clearances to verify the cam is in good condition, but again, truck runs like a top until it goes into closed loop, then runs like crap, unplug the MAP, and it runs great again (ie runs exactly like it does when its cold, which is good other than poor fuel mileage).

What I'd really like to know, is how the ECM responds (ideally that is) when it gets no input from the MAP sensor, because knowing that, would very likely point me directly to the actual point of failure.

Can anyone say, with any certainty, what is SUPPOSED to happen if the MAP gets unplugged when the truck is in closed loop?
Is there some vacuum line that opens or closes when the engine is warm, that is directly tied to the intake manifold (perhaps a vacuum switch that isn't switching when it should?)
Does anyone know if say, i unplug a single sensor, will the ECM stay in closed loop mode, and use a predefined default value for the unplugged sensor, or does the ECM just say fuckit, and use predefined defaults for ALL sensor values?

When I was building my MAP adjuster I noticed a few things. On my 1990 4.0 Renix if the MAP is unplugged (engine running) it will idle fine but as soon as you attempt to rev the motor it stalls. I've repeated this several times, and the result is the same.

To clarify when I say unplug, I mean the vacuum line.

Ah, and to clarify on my end, I'm unplugging the electrical connector from the MAP. When i unplug the vaccum line from it, so long as both ends are sealed, it runs, but poorly. hot or cold.
If i unplug the vaccum, and don't plug it, then mine stalls as well, although i can coax it into running with some fancy pedal work.

has the distributor ever been removed? if so it may not be indexed properly. its worth checking before you spend any money needlessly

1991s have a MAP sensor?

EDIT: Nevermind....I'm retarded....

Carry on ...

blistovmhz said:

Ah, and to clarify on my end, I'm unplugging the electrical connector from the MAP. When i unplug the vaccum line from it, so long as both ends are sealed, it runs, but poorly. hot or cold.
If i unplug the vaccum, and don't plug it, then mine stalls as well, although i can coax it into running with some fancy pedal work.

Click to expand...

If the new TPS I suggested earlier does not solve your problem, you might check the MAP sensor and TPS wiring, especially the ground wires. I had to replace the ground wires on My TPS and on my TCU (Transmission Control unit) recently. I mention this because when you disconnect the MAP sensor wires the ECU should see an open circuit, and thus ignor the MAP sensor but when you disconnect the vacuum line it still sees an electrical signal, like 14.7 psi atmospheric instead of the normal vacuum, which makes the A/F ratio extremly lean with the vacuum line disconnected.

Lastly, I would check the O2 sensor wires for a short or poor ground (with a multi meter, volt-ohm meter) which could cause the closed loop to run poorly even with a new O2 sensor.

fuel pressure fuel volume ,map sensor low vacuum high volts close to 5 .high vacuum low volts 1.trans shifting ok try accelarting up through .check dist weights.check vacuum leaks with propane torch..

Ok, I replaced the TPS as well. No difference.
I'm going to recheck the fuel pressure again, then start rewiring if no one can think of anything else to try.
I'll start with rewiring the MAP, TPS, and O2.

Manifold air temperature sensor (MAT or IAT) could be the problem. It may be sending bad data to the ECU when hot. ON the Renix the MAP and MAT are in series IIRC, if the OBD-I is the same way, disconnecting the MAP electrical is also disconnecting the MAT input to the ECU!!!! Check a wiring diagram on the OBD-I ECU sensors and check the MAT/IAT sensor resistance at several temperatures! When mine went bad it would not start when cold, but the effect depends on which way the sensor went bad, i.e. is it sending too hot or too cold data to the ECU.

I unplugged the MAT, no effect. The problem very definitively goes away till 4000 RPM with the MAP unplugged. Cannot reproduce with anything else. I think this brings me down either to three possible conclusions. (ECOMIKE, i'm hoping you'll tell me I'm mistaken and that one of these three isn't the problem )
1. MAP data is correct, and is sending spiky data to the ECM in response to a vacuum leak.
2. MAP data is incorrect, and is not consistant with actual vacuum, indicating a wiring fault.
3. Engine can't handle advanced timing, indicating... *sigh*
a. Valves clearance is incorrect, or I have a broken valve that is only symptomatic at a certain timing threshhold (i've heard of it, never seen it).
b. Worn cam lobes?
4. Distributor gear is indexed one tooth too advanced, but the ECM is compensating by retarding the timing as far as possible, which isn't far enough at high RPM or under load.

A buddy is supposed to bring me a propane torch tonight to check for vacuum leaks again.
The problem didn't seem as severe right after I did the exhaust manifold swap, (new gasket). Perhaps there really is a vacuum leak, and stuff just needs to be torqued down again.
If I don't find a vacuum leak, I'll start replacing wires.
This'll be a "pita" as its fairly consistantly -40 C lately and I don't have room in the shop for the Jeep .

Ah, a development.
I believe this problem is only apparent when the transmission is in forward gear. I threw it into reverse and punched the gas, and it went just fine. Threw it into 4L reverse for good measure, same deal, no problem.
What can possibly show itself only in forward gear, and disappear with the map unplugged?

Flexplate/CPS orientation? Motormounts might move differently, maybe affecting engine harness connections, vac lines...don't think anything would be different on intake or exhaust system as far as leaks...

I suppose if the engine is lifting under torque, it could screw with the CPS. I can picture it, but I don't know if its possible to throw out the orientation enough to mess with it. How do we throw the unplugged MAP making that problem go away? Would the CPS have less/no effect on ECM if the MAP's not plugged in?