Ok, my 95 accord at 211K miles has been getting terrible gas mileage. I get around 200 miles per 13 gallon fillup, or ~15 mpg in town. When I am on the interstate, I get excellent gas mileage, near 450 miles for a whole tank.

I have tried to figure out the problem, but I haven't been able to pinpoint it yet. I am determined to figure it out. I will lay out my plan below, and I am up for any other suggestions too.

Here is what I know.

- I am starting to get a consistent engine stumble in the 1500-2500 rpm range. It will stumble in other ranges too.
- It seems like an EGR problem, but the valve and ports are clean.
- I do not smell any gas leaks near the injectors.
- I periodically get a code 43, especially when it is cold out, but it hasn't popped up in over a month. No other ECU codes are coming up.
- I checked and cleaned all connections for the O2 sensor. I even cleaned and checked the proper connector near the engine bay fuse box.
- The O2 sensor is new, but I will call TAS Auto to get a replacement under their warranty, since I am getting a code 43.
- My fuel pressure is within spec. I will retest and post.
- My tires are about 3 months old and inflated properly. Alignment was done at that time too.
- All the routine stuff has been done, plugs, dist. cap, rotor, etc...

Here is what I need to check and I will post my results:

- Compression test on the cylinders. I forgot to open the throttle body when I did this the first time and they all were consistently low.
- Make sure that the brakes are not locking up. I don't hear any strange noises.
- Replace O2 sensor.
- I am going to replace and inspect the exhaust from the engine to the catalytic converter.
- I might replace the gaskets in there too. Honda sells an "anti-rattle" kit. Is there anything special about this, or are they new gaskets, springs, bolts for the intermediate pipe?

 

Have you checked the vacuum going to the EGR port? It is vacuum-operated, isn't it? If so, maybe the valve looks okay visually, but it's stuck shut or not getting enough vacuum to open.

 

PA, you've been around here long enough to know you should use NGK plugs, right?? (just checking...)

Dragging brakes can be noticed by coasting the last foot before stopping. When you coast that slowly, a dragging brake will grab at the last moment before stopping.

Maybe the O2 sensor? Any reason to think silicone compounds (curing RTV rubber) got into the combustion? That stuff is sudden death for O2 sensors.

 

Yeah, I used the OEM NGK ZFR5F-11 plugs.

I'll try the dragging brake test you mentioned.

I only used RTV on the corners of the valve cover gasket and on the corner of the oil pan gasket.

 

I've used Permatex Grey (substitute for HondaBond) in those areas & didn't give me any trouble. I assume you didn't use some generic window caulking? For automotive sealantsthey apparently do something to the formulation so it doesn't give off silanes as it cures.

 

I used the Permatex RTV Blue ultra to seal those gaskets.

So I tested the fuel pressure today.

32 psi with FPR shop manual range 30-37 psi
42 psi without FPR shop manual range 38-46 psi

While driving the pressure would go up when I accelerated and would drop as my rpms leveled out. Stumbling didn't occur with any noticable change in fuel pressure.

I shut off the car and the pressure didn't drop significantly for about ten minutes. I guess that the injectors are not leaking.

I think that this can eliminate fuel supply as a culprit.

 

Yeah, looks like fuel supply & FPR are OK.

Behavior when driving it good. The fuel pressure is supposed to go up& down to follow intake manifold absolute pressure.

 

I thought sure some of that fuel had to be leaking by injector-fuel rail orings, but that's ruled out by your test,...except for a fuel tank leak. This type of leak would be consistent w/ your pattern, good road mileage and poor suburban driving. Reason is the longer time to use a tank in city, usually 1-2 wks. This would allow time for a substantial leak from tank or fuel line to cause your poor mileage. Leak would have be on fuel return line to tank or tank itself.

This would not account for your mid-range stumble which has all the characteristics of EGR stumble, but you say valve and ports are clear. Have you checked EGR system by applying vacuum at idle (bypass solenoid switching valve)? This should cause engine to stumble badly when EGR opens and overleans the engine. You could have a blocked passage from EGR valve to EGR manifold.

good luck

 

I was hoping that leaking fuel injectors would be the cause too. I wish it was that easy.

So TAS Auto warrantied my O2 sensor, and I will get it by early next week. I also ordered new gaskets for the exhaust system. I will have to see if this will fix the problem.

The stumble feels like an EGR problem. I will repeat the vacuum test on the EGR system. I will have to look up those tests, but if my memory serves me right...
- Apply vacuum to the valve to see if it opens and causes the engine to stumble.
- Also hook up vacuum gage to the line feeding the EGR valve and open throttle plate. At a certain RPM, there should be vacuum.

I'm trying to figure out other items that I can easily test. There are no OBDI DTCs for the EVAP system. I wonder if there were any faults in the EVAP system that would cause that much of a fuel loss?

 

Code 43 can be a faulty in O2 sensor heater circuit. Worth checking resistance of heater circuit.

See my previous post. Rule out leaks before getting into evap system. It's supposed to scavenge vapors off the gas tank.

Good Luck