cfahamilton

I have a 97 Civic CX Hatchback. Basic, manual everything, no AC.

I recently replaced the engine (spun crank bearings) with a used stock engine from someone who was swapping to a different engine for modding purposes.

I immediately noticed that it had much better acceleration than before but soon realized that the fuel economy was horrible (low 20s). I also noticed a gassy/rich smell from the exhaust. I realized it was running rich and have been trying to figure out why. In addition, the idle is a little rough and it is now throwing the P0420 code for "catalyst system operating below efficiency".

What I have done/checked:
- plug wires - in excellent physical condition

- plugs - no fouling or deposits, verified gaps are correct

- checked air filter - It didn't look too bad but I replaced it anyway.

- reset timing to factory setting which was a little off due to the swap (this had no noticeable effect)

- checked fuel pressure and regulator - Pressure is within specs: 40psi pre-start charge, 34 psi when running. Pressure is higher when engine is revved but returns to 34 when rpms stabilize regardless of engine speed. Pressure rises appropriately when vacuum hose is removed from FPR and pinched. (I don't know if this means anything, but the idle goes up and seems to stabilize when the vacuum hose is removed if I DON'T pinch it off.)

- checked the codes - there are no ghost or pending codes,

I don't want to just throw parts at it, especially an expensive cat. If I do end up replacing the cat, I don't want to do it until the underlying rich mixture problem is solved so I don't kill the new one.

Questions I have:

- Could it just be a bad O2 sensor causing the rich mixture which is then causing the catalyst problems? If so, which sensor, pre or post cat? Also, why isn't it sending a bad O2 sensor code?

- What else could be causing the rich mixture condition?

I have just ordered a bluetooth OBD2 scanner so I can look at live data but it will be a few days before I receive it.

- Are there any other tests or procedures I can do in the meantime?

- When the scanner comes what should I be looking at?
What should I expect to see from the O2 sensor readings?

How do I interpret the short and long term fuel trims?

Sorry for the long post. Just trying to be thorough and not waste people's time. Any input or suggestions would be greatly appreciated.

Thanks.

ezone

Did your new engine come with the cat and both o2 sensors? Did you use them, or reuse your original parts?

cfahamilton

Thanks for the reply ezone.
The engine did not come with cat & o2 sensors, I reused the old ones. Since my original post I tested the fuel pressure regulator, and the resistance of the injectors and everything was normal. I then swapped the fuel rail and injectors with those from the old engine, just in case. There was no change in the symptoms.

oneheadlight

check the voltages from an OBD2 scanner for the o2 sensors. I bet you'll see them giving wrong voltages causing the car to run rich. I had mine go bad and they were running towards the rich side of things. Sure, there was lots of power, but keep that up and it was a good way to burn out a cat! Luckily I didn't. also, when I checked the resistance of the primary o2 sensor, it was infinity! It was definitely "Dead".

ezone

Ok........


What oneheadlight said above.

Get your new reader and plug in and watch O2 sensor values as you drive....And watch fuel trim values at the same time.

Also check stuff like MAP for false load value (low manifold vacuum, valve adjust, cam timing, etc.), and ECT and IAT for accuracy. Operating temperature is important.
Make sure EGR (if equipped) isn't stuck open or has clogged passages.


I'll toss out some random stuff here but I have no way of knowing what's going on with your car yet.....

Pre cat sensor is what I have seen problems with in the past, causing intermittent poor running without any code.
Use your new scanner to watch both O2 sensors and fuel trims....Watch the front sensor during a normal calm test drive.....see if the voltage cease regular fluctuation, voltage gets stuck low, then the computer tries to drive the fuel richer as it tries to elicit a response from the sensor. It maxes out fuel trim as it tries to get a response, sometimes they can go rich enough to puff black smoke and make the engine chug. It may resume working correctly again in a short time, before the computer decides there is a problem.




Rich running can wipe out the cat.

Why did the first engine fail? Lost oil and spun bearings?
Oil consumption can ruin the cat too.

But false reading O2 sensors can make the computer believe the cat is bad when it may not really be.



Does your new scanner toy have graphing ability? Post the graphs?

You can unplug the front O2 sensor, do a battery reset (resets learned fuel trims to 0), and drive it to see if the richness and chugging stops, unplugging it should cause it to revert to following the base fuel table without O2 sensor corrections.



All this may be futile if the cat is truly bad, broken element, holed or hollow.

What are you smelling? Raw exhaust or raw fuel? (Raw exhaust would be exhaust that did not pass through a catalytic converter.)

Is there any exhaust leakage between the cylinder head and the cat? No leakage after the cat for at least a couple feet of pipe?

Sometimes the engineers can't predict all the different ways something can possibly fail.

I still see stuff like this even on new cars. 2014 Civic with 6000 miles....a flaky front AF sensor that's causing strange codes elsewhere, but not setting any code directly indicating the bad sensor.

cfahamilton

Hey guys,
thanks so much for the responses. This is the most specific advice and relevant feedback I've gotten so far.

First, unfortunately I haven't received the scanner yet but I will look for the things you mentioned when I get it.

Second, your other questions:

I never quite figured out why the first engine failed. It inexplicably overheated one day under normal driving conditions and stalled out. It had not lost any oil or coolant. It started again after a while of cooling but after that it would periodically overheat, was hard to keep running, and it made an awful racket. It never burned oil, before or after that. a mechanic told me it had bottom end damage and I confirmed the spent bearings after I swapped the engines and dropped the oil pan off the old one. The best I can figure is that there were clogged oil passages. The thermostat tested normal.

I have no reason to believe there is any physical damage to the cat (no cat codes prior to this, no rattles) but it is hard to check as there are curved pipes before and after.

Today I pulled the upstream O2 sensor and tried to see what kind of readings it would give me with a propane torch bench test. I also tested some other ones that I had previously pulled from a junkyard for comparison. Overall, it performed better than any of the others. The voltage ranged between ~.96 and .00 and the response time seemed pretty fast but I don't really know if it's fast enough.

*** How do I do the battery reset? Do I just take the negative cable off the battery while the O2 sensor is off? ***

There is definitely no exhaust leakage before the cat. The first point where there could be any leakage after the cat would be a couple feet later at the spring-bolted joint just below and behind the engine. I don't hear a leak there but I haven't actually checked it.

I'm really hoping the scanner comes tomorrow so I can get and share real data.

*** Where can I get normal values for MAP, ECT, IAT, and EGR sensors to compare to? ***

ezone

Yes. Leave it off for maybe...10 minutes.
Sounds good enough to me.
All depends on how your scanner interprets the readings. I set mine up to read MAP in volts, temp sensors in degrees F. I'm used to seeing it this way and I don't have to convert in my head to think about it.

MAP is actual manifold vacuum. I like to see less than 0.9v at hot idle with no loads.

ECT and IAT should read exactly what the temps are at each sensor. Obviously they should both read nearly the same as ambient temp after the car sits overnight and change as the engine heats up. Neither should give a false reading, verify with a real thermometer.

EGR.....probably about 1-1.15v closed pintle.

cfahamilton

Okay, I finally got a scanner that works with my car and I now have data.

First, I did the reset with no change in symptoms.

Now on to data. Forgive me, I am new to torque and haven’t figured out how to log data and post graphs yet but I can describe what everything was doing.

I have looked at the data when the vehicle is cold and warming up, when warm at 2500 RPMs in my driveway, at highway speeds, and at idle fully warmed up after a highway drive.

Summary:
ECT: Goes from ambient temp when cold to 183-187 deg F when warmed up.

MAP: Ranges from 23-27 kPa in the driveway, 47-63 kPa at highway speeds (ambient was 98 kPa)

RPMs: cold start RPMs ~1500, after warm up idle is rough and hunting, ranges from ~600-700

Timing Advance: 25-35 deg btdc when revvng (cold start or highway), idles warm at 10.5-13.5 deg btdc but is not steady

O2 pre cat: generally oscillates between .1 and .8 in a normal way based on viewing other people’s “normal” graphs. At highway speeds it settles some and oscillates between .65 and .75 @ ~2500 RPM (70-75 mph)

O2 post cat: when idling after fully warmed up, it reads ~.75-.8 and is fairly steady, at 2500 RPM in the driveway it ranged from ~.4-.9 mostly following the O2 sensor 1, at highway speeds it ranged from ~.3-.8 and oscillated much more widely than O2 sensor 1. (This seems bad to me and I hope my cat isn’t permanently dead)

STFT: at cold startup, stft is 20-25%, at warm idle it is usually between 12-17% oscillating slowly (usually between 13-16, sometimes between 10-13) at highway speeds it settles down and oscillates between -2 and +5%

LTFT: this value has been rock steady at about 2.5%

The large positive STFT at idle combined with low STFT at speed suggests a vacuum leak according to my research. Unfortunately, I have performed the propane test and have not found any leaks. I have presented the propane around all vac hoses, intake to head gasket, tb to intake gasket, etc with no response in idle speed or STFT. The only time I got STFT response was when I fed propane directly into the intake to confirm that I would get a response.

Please let me know if you have any ideas or suggestions based on the data I’m getting. Could this be due to a bad or dirty IACV?

Any help would be greatly appreciated.

ezone

Ok on a stone cold engine, ECT and IAT should closely agree.
MAP and BARO should match with the key on - engine off.

This should fluctuate ALL the time the system is in closed loop, from .1-.9v and anywhere inbetween.

Your scanner may not update fast enough to display the actual O2 sensor values in real time, the sensor can fluctuate several times per second.. Very few scanners can update that fast.

If the rear sensor mirrors the front sensor, the cat is probably not doing its job.
The rear sensor should not change faster than the front, although it may appear that way if your scanner is slow to update.

(You didn't somehow get the sensors swapped, did you?)
Odd.
But it is not coding as lean or rich, the computer thinks it has control of the fuel delivery.


Vacuum leak should cause idle speed to increase on this fuel system (speed-density), not show as a big fuel trim change.

Got a ton of miles on this? What's the spray pattern of the injectors look like?

Can you convert the MAP numbers to volts? Or use a volt meter and backprobe the sensor to see what the volt reading is at hot idle with no load?


Have you unplugged the O2 sensors and driven it?

Have you replaced any O2 sensors?

cfahamilton

Thanks for the reply ezone.

"(You didn't somehow get the sensors swapped, did you?)"

When I saw the weird readings I had the same thought, but NO, I did not. It would be vey difficult to swap the connectors and I did double check to be sure.

I wish I understood better why a vac leak would present differently on this engine, but I'll take your word for it. Especially since I couldn't find one anyway.

The car does have a ton of miles ~260K but I don't actually know how many miles the replacement engine has on it. I think the guy told me more like 150K but I'm not sure. Also not sure about the spray patterns. I do know the prob got slightly worse when I swapped the original injectors back in. I'm going to swap them back but I was thinking of swapping the whole intake assembly from the original motor at the same time and seeing if that helped. Seems like a lot of work for a guess, but I'll do it if I run out of other ideas.

Later today I will try to get a volt reading for MAP.

I have not driven it without O2 sensors. What would this tell me? Am I just looking to see if the symptoms improve?

I have not replaced O2 sensors because the one determining mixture seems to be reading correctly. I am going try swapping them with the ones from my 98 which is running fine. I was hesitant to do it because if something is fouling my sensors I didn't want to end up replacing those too.

I will try to report back later today with more data.

ezone

Speed-density fuel injection strategy is to use the MAP sensor as its main load detector, and if vacuum drops then fuel is increased as it believes that to be a load.
If you open a vacuum leak then MAP drops, and the computer adds fuel to compensate, so the RPM ends up raising.
On these it will raise until it hits the 2000 threshold for decel fuel shutoff, then it drops to 1200 and fuel is turned back on, this cycle repeats and causes the famous surge.
Just to see if symptoms change (you may need to kill battery memory to clear adaptives, check fuel trims on your scan app)

Disabling the sensors can remove compensation factors, computer should run at base injection table without trim corrections. If it runs much better this way, then one might suspect a lying sensor.

If it was running rich enough to chug or lope due to O2 sensor fault, disabling can help find it.

Ricardo2

Hi all,

New poster here. I know this is years later, but I am having the same problem on my 05 outback and this has been extremely helpful and insightful. Have not had much luck with most mechanics.

Hoping that my cat isn't totally dead and wanted to find if you guys ever figured out what was going on here.

Thanks

sdaidoji

iTrader: (1)

flat4 engines tend to leak more oil to the cylinders, since pistons always laying down.
not sure of outback, but forrester is known to have oil loss.