2000 Subaru Impreza 2.5RS running negative 30% and worse fuel trims

Hello,
I am looking for some additional ideas and guidance on a seriously stubborn rich condition I am having on my Subaru Impreza.. I have been following ScannerDanner's material for the last couple years, tried a lot of things and learned a lot, but just can't manage to pinpoint the issue i'm having. 
My engine
 is a 2000 Subaru EJ25 SOHC NA California Emission (Map engine without EGR) 
Current trouble codes include
Cat Efficiency P0420,) (Multiple Misfire P0301, P0302, P0303,)
Previous recently resolved trouble codes:
P0113-IAT Resolved Replaced IAT/Map Sensor
P0340 & P0335 Cam and Crank Sensors Replaced in roadside repair. Towed home and diagnosed ECM Fault. Replaced ECM
P0442 Evap Leak Replaced Fuel Filler-neck and associated lines
Drivability Concerns;
The car drives well despite fuel trims and has good power but horrible fuel consumption, about half the distance I would normally get from a tank of gas. The main symptom is an intermittent low lumpy idle, causing multiple misfires as it nears stalling but does not stall. Followed by erratic hanging idle of 3000rpm very rarely(once a month). I have had to park the car for over a year because of its fuel consumption. 
Data Collection;
Monitor; Car will get stuck full rich in open loop when left to warm up at idle, and switches to closed loop-drive after touching throttle, applying negative trims but never fully balancing 02's) 
Data @Idle in Closed Loop
Engine Load 3.1%
RPM 701
Fuel Trims; (Long term -29.7%) (Short Term -12% or similar)
Oxygen Sensor Data: Upstream A/F Sensor 1.3v active or 0.91 Lambda, 1.1Amp HeaterDownstream heated 02 790mv active, 1.02Amp Heater
Air Assist Solenoid; Off
Map Sensor Rate 3.92grams/sec
Intake Air Temp 87.8F
Atmospheric Pressure 743mmHg
Manifold Absolute Pressure 278mmHg
Manifold Relative Pressure -472mmHg
Throttle Angle 0.4%
Absolute Throttle Position 0.392%
Idle Speed Control 17 step
Ignition Timing 10 degrees
Knock Correction 0 degree
Fuel Injection Pulse 2.56ms
Fuel Temp 48.2F
Fuel Tank Pressure 0 mmHg
Camshaft Pos Signal; On/Off
Crankshaft Pos Signal On/Off
Coolant Temp 167F
PCV Solenoid Valve Off
CPC Valve Duty Ratio 0%
Evap Vent Solenoid Off
Fuel Pressure; 38psi (41psi KOEOff, 48psi w/ vacuum line disconnected)

Please kindly post your ideas and I will promise to give you all the accurate data I can.
Thanks

 Blair

Before I share my ideas, I did want to clarify a couple things:

The car drives well despite fuel trims and has good power but horrible fuel consumption, about half the distance I would normally get from a tank of gas. The main symptom is an intermittent low lumpy idle, causing multiple misfires as it nears stalling but does not stall.

So, to be clear, the vehicle does NOT lack power? As in, whenever you floor it, you get the acceleration you think you should?

For giggles, if you disconnect the battery cables, leave them off for a minute and reconnect them, how long does it take the fuel trims to go badly negative once it reaches closed loop? Seconds? Or does it require a test drive?

Onto ideas. For whatever it's worth, I think you're upstream A/F reading and the downstream O2 reading agree enough to suggest that this is a true rich condition, and not a false one. As in, the upstream A/F is showing rich, and the downstream agrees. If the upstream sensor said rich, but the downstream was flat lean, that'd suggest a false rich. That's why I asked about the power - a false rich will cause a severe lack of power due to underfueling.

What stands out to me is the "Map Sensor Rate 3.92grams/sec" reading. Should that have read MAF instead of Map? Nearly 4 g/s might be on the high side for a 2.5L engine at 700 RPM. A MAF that's overreporting airflow will cause a rich condition. Does that reading ever go lower than 3.92 g/s when the engine gets to operating temperature with no electrical loads (A/C, blower, wipers)?

Thanks Tyler,

To your points;
I can confirm that the car does not lack any power.

After hard resetting the ECM the car will pull fuel within seconds after entering closed loop.

I also am believing the o2 sensors as other tests I have performed have confirmed good function.

The engine is a MAP only engine and the 3.92g/s is a theoretical calculation from the ECM. I have seen lower values at idle. Usually around 3.37 to 3.5g/s. I too wondered if this value was too high and replaced the sensor with no improvement to symptoms.

Sincerely
Blair

A MAP sensor reports air pressure. A MAF reports air speed. Can't have a g/sec from a MAP. Even if it's theoretical, it's to high. MAF g/sec should be close to engine Liters. A high g/sec would mean more air going in so ECM would add fuel. Googled 2000 Subaru Impreza 2.5 shows it has a MAF??? I've been wrong before on Hoopdedoos.

Thanks scannerjohn,

This is helpful to hear that the grams per second must be closer to engine size in litres than it is. I was unsure how much deviation was acceptable. I figure it is very likely I got a bad replacement part. One thing is certain though; my engine absolutely does not have a MAF sensor and is strictly a speed density engine. Yet the scan tool reports a value.

Blair

Update;

I tried reordering a replacement MAP sensor and saw no improvement of the ~4grams/sec value at idle or any change in fuel trims. I have been over the wiring harness more than once to check for continuity and resistance values all of which were ok. Do you guys think I have the misfortune of multiple deffective new parts? Or is there something else that could explain this rich condition and mis-calculated air flow rate?

I was doing some reading and found this formula;
RPM x Liters x MAP x IAT Air Density = Mass Air in g/sec

My thinking is that I'm being decieved somehow. The car has had Cam and Crank correlation faults before, which soon after fixing led me to this rich condition.
Is it possible that the recently replaced cam and crank sensors could be messing up the RPM calculation? The RPM value is believable, but I've been noticing a strange cue on my scan-tool. I wonder if anybody knows what it's telling me when cam and crank live data at idle keeps alternating between "on" and "off"? in irregular fashion. Could this hint at intermittent cam or crank signal failures? Is this a common sight for a running car?

From my reasoning and comparing with another car in my stable; a 2009 impreza 2.5i (MAF Engine), the known good car has both the desired 2.5g/s air flow rate and stable 'ON' cam and crank signals 100% of the time. Given this, where do you guys think my problem lies?

Thanks
Blair

"MAF g/sec should be close to engine Liters."

According to one of Paul's videos, that statement is generally true for engines over 3 liters. Under 3 it may not be accurate. 

My understanding of crank/cam position sensor relationship is that usually a crank sensor will trigger once for every "X" amount of times the cam sensor does. This would be verified by scoping hall effect (or other) signal wire on channel 1 and 2 and looking for the synchronization when it should be happening. It's helpful, of course, to know what the crank to cam trigger ratio is for that particular engine when reviewing the data, but it's not always super essential, For example, If you never see a point on the graph where crank and cam signals line up that would be a tip off. Another would be if you can't see any repeating pattern in the data. If the ratio info is important for you to have, I have heard some people having luck with www.iatn.net , though there may be a fee involved. 

When is the last time you changed the upstream O2? You might want to try unplugging its connector and see if any of the driveabilty issues you described (lumpy, erratic idle, bad fuel consumption) improve. Just a thought.  

The original MAP reading you provided (278mmHg or 10.9inHg) at idle looks normal to me, if you had any doubts about it. 


 

Thanks for your input Ksat22,

I had gone down the road of educating myself on the topic, but ultimately do not have access to a 2 channel scope and am reluctant to purchase one on a budget. I only got to borrow a single channel analog scope from my friend which showed nothing remarkable.
Since I have no more leads here. Does anyone else have any new ideas I could chase?

Sincerely
Blair