Baffling Problem After Engine Warms Up (Honda)

[Barry Shea]

1997 Honda Civic DX
D16Y7 – 167,000 miles
No “Check Engine” light.

When the engine is cold, it starts right up, idles smoothly, and runs great on the highway (65-70 MPH). But, as soon as it reaches operating temperature, it starts bogging down.

If I don’t move the accelerator pedal, the car will slow down to 40-45 MPH, then the engine will kick back in, but will run rough.

If I slowly press on the accelerator, there will be no response for about 1/8″ to 1/4″ of movement, then the engine will kick back in, but at a higher RPM, so I’ll have to let off the gas to slow down, and the problem will repeat.

When this problem starts, I can smell fumes seeping into the cabin. They resemble what the inside of a valve cover smells like. Kind of a mixture of oil and exhaust.

Sometimes, at random, the engine will start running smooth again. But, the problem soon returns.

If I get off the highway while it’s running crappy and come to a stop, the idle will be real rough, almost to the stall point, and I’ll have to give it more gas to keep it going. Sometimes the engine will suddenly start idling smoothly again, but will eventually go back to running rough.

It has gotten so bad that the other day I drove to the store and when I came out the engine wouldn’t start. I had to sit there for an hour and let the engine cool down before I could get it started, and it ran so rough that if I let off the accelerator it would stall. I barely got it home.

This problem first started about 5 years ago. At first it was seldom, gradually became more frequent, then was assured, and now I can’t even drive the car anymore. Although, again, when it’s cold, it starts right up and runs great.

I’ve taken it to 3 repair shops, who said no codes showed up on the diagnostics, so they started throwing parts at it. I’ve been hosed for over $2,000 for a catalytic convertor, distributor assembly, fuel filter, air filter, spark plugs, spark plug wires, and a PCV valve. And nothing has fixed the problem.

I need your help. It’s definitely something related to the warming of the engine. But, why and what?

Topic

[Mike]

There is something in the throttle body on a lot of Hondas where coolant flows thru it and when it warms up, something changes airflow or mixture or something. I’ve fixed this twice before, due to really wacky idling once it warmed up. I’m sorry I don’t know anything more technical about it, I think I need to study this today (and check to see if my Accord has this setup).

The problem was that where coolant flowed thru the throttle, it was clogged with a kind of rust gunk/gel. The coolant pipes and hoses around the throttle are barely any bigger than vacuum lines, and one of the metal pipes commonly has a second metal vacuum pipe built onto it.

You basically take that apart, blow the pipes and throttle coolant passage out with compressed air and rinse it clean, then reassemble and fill the coolant. If that is the problem, you should be able to tell as you start to blow it out. You should get some sort of clog that comes out. Even if this is not the problem, at least it is very cheap and takes only about an hour to do it very thoroughly.

[Barry Shea]

UPDATE:

Cold engine starts right up and idles smoothly. I let it idle until it warms up to operating temperature. Still runs fine. I turn the engine off, try to restart it…it won’t start. It just turns and turns and turns.

Removed and cleaned the Idle Air Control Valve – a real pain in the ass. Waste of time. Nothing improved.

Also, a friend pulled the dip stick and said it smelled like the oil had gas in it. What the hell?!

HEEEEEEEEEEEEEEEEEELP!

[Mike]

Sorry that advice didn’t work out, but at least it was less of a pain in the ass waste if time as earning that $2000 was. There are some other smart fellas around here too who will hopefully have some good ideas to chime in with.

It sounds like the computer thinks the engine is always cold. The fuel would have gotten in the oil during the extended cranking trying to start it warm. If too much fuel is injected during starting, it will not start and much of the fuel will leak around the pistons and fall down into the oil pan. It’s fairly safe to say that too much fuel was injected during the warm start. More fuel than normal is required to start a cold engine, and that extra fuel will be injected if the computer thinks it’s cold.

So the thing to investigate at this point is the coolant temp sensor and it’s wiring. The car is old enough at this point that corroded wiring or connector terminals could be an issue. Corrosion increases resistance in the wire, and the sensor tells the computer what the temp is by changing it’s internal resistance. Obviously having corrosion present adding artificial resistance to the coolant temp circuit will cause problems even if the sensor is ok. Of course the sensor could just be bad. It happens.

Try unplugging the coolant temp sensor and see if that changes anything (you can start it cold and then unplug it before it gets warm). It won’t run quite right without it and will make a code, but won’t run as bad as you’ve described so far (with everything else being ok). The point being that if it’s better without it plugged in, it’s very likely that’s the problem area.

A code reader that displays live data will also show what the computer is seeing. If the car is cold and it’s 50 degrees out, the coolant temp should be 50. At full warm it should show around 180-190. Any difference from actual shown on live data would also indicate this as being the problem area.

[Ingvar]

You know, 1st thing that came to my mind when you mentioned fumes smell, is PCV. It notoriously goes clogged on aged engines and shitty hose with no clamp they normally have on it cracks. And because it does run sort of that kind of fumes through it, that makes me think you want to take a good look at it. It’s also only few bucks replacement.
Next would have been your EGR valve. At that mileage, it’s destined to be in need of cleaning anyway.

[Gary]

[quote=”Fopeano” post=124088]Sorry that advice didn’t work out, but at least it was less of a pain in the ass waste if time as earning that $2000 was. There are some other smart fellas around here too who will hopefully have some good ideas to chime in with.

It sounds like the computer thinks the engine is always cold. The fuel would have gotten in the oil during the extended cranking trying to start it warm. If too much fuel is injected during starting, it will not start and much of the fuel will leak around the pistons and fall down into the oil pan. It’s fairly safe to say that too much fuel was injected during the warm start. More fuel than normal is required to start a cold engine, and that extra fuel will be injected if the computer thinks it’s cold.

So the thing to investigate at this point is the coolant temp sensor and it’s wiring. The car is old enough at this point that corroded wiring or connector terminals could be an issue. Corrosion increases resistance in the wire, and the sensor tells the computer what the temp is by changing it’s internal resistance. Obviously having corrosion present adding artificial resistance to the coolant temp circuit will cause problems even if the sensor is ok. Of course the sensor could just be bad. It happens.

Try unplugging the coolant temp sensor and see if that changes anything (you can start it cold and then unplug it before it gets warm). It won’t run quite right without it and will make a code, but won’t run as bad as you’ve described so far (with everything else being ok). The point being that if it’s better without it plugged in, it’s very likely that’s the problem area.

A code reader that displays live data will also show what the computer is seeing. If the car is cold and it’s 50 degrees out, the coolant temp should be 50. At full warm it should show around 180-190. Any difference from actual shown on live data would also indicate this as being the problem area.[/quote] Exactly, from the sound of it, hes running too rich. We need some more information here for a proper diagnosis.

[Andrew Button]

That temp sender is a variable resistor. Find the ohms value for the sensor online (google it) when car is at running temp, get a resistor of the same spec (it will be in ohms), and put it in the harness where the sensor was. This will tell you whether the sensor is bad or not. Depending on how many wires that sensor has lets just assume 2. If more wires it will take more research. May not run right until the car is warmed up, but this will be the easiest test you can do. It sounds like its loading up and running rich for cold all the time.

[Paul]

I agree with Fopeano, sounds like coolant temp sensor faulty and over fueling the engine when hot, disconnecting the sensor will bring on your engine light but if the problem gets better then its likely the temp sensor at fault.

[Andrew Button]

I don’t know Honda wiring well, although its a two wire from what I saw on the diagram.

Honda ECT thermistor resistance range;
20kΩ @ -20°C (-4°F)
0.1kΩ @ 120°C (248°F)

It appears that the resistance goes down as the temp goes up. In this case, if you were to unplug it the sensor you would be going in the right direction, although possibly throwing the SES light. Think about this though if the connectors are dirty, it will increase resistance, so check the connections on the harness and sensor. Ohm out the sensor with motor hot and cold, and see if values change – if its stuck on a high value, and your case probably pegged high then you know your sensor is no good. Just DON’T ohm anything else, just the sensor, DISCONNECTED from the harness. Don’t under any circumstances ever put the ohm meter in the harness, ever. In fact keep your probes away from the harness at all costs unless you know exactly what you are doing with your diagnostics. And one more thing, if you do conclude your temp sender is bad, wait until the engine is cold to change the sensor.

[Barry Shea]

UPDATE:

First off, mucho thanks to all you guys for offering help. I really appreciate it. You are good and decent men. It restores my faith in mankind.

Forgot to mention that the Check Engine light is NOT on, and there are no trouble codes reported when connected to a scanner.

What I’ve done so far:

Replaced fuel filter
Changed oil & filter
Adjusted valve clearances
Checked timing
Checked PCV valve
Checked fuel pump/pressure
Replaced cap & rotor (rotor screw had come out, rotor was flopping around, loose screw was arcing)
Replaced spark plugs
Replaced Coolant Temperature Sensor
Replaced Coolant Temperature Switch
Cussed multiple times

Problem persists.

But, when the engine is cold, it starts right up, idles smooth, and runs real nice.

F me.

[Barry Shea]

What info do you need?

[Paul]

I have once or twice seen this type of problem with a crank sensor breaking down, the sensor would behave normally when the engine was cold but once the engine got hot the sensor began breaking down under load and causing many of the symptoms you describe. This was overlooked by many mechanics as the engine ECU had not identified the sensor as bad, and after all how could the car run with a bad sensor right? Once it was put on a scope with the engine hot the sensor break down was obvious. Not sure if your crank sensor is inductive pulse or hall effect but if you can get it scoped or even just swap it for a known good one then at least its ruled out. At this point you need the help of a decent tech with good test equipment as all the best will in the world from us is no match for good face to car diagnostics guy that knows his stuff! 🙂

[James O’Hara]

So far everything that I have thought of has been said except cam sensor can do same thing as crank in the older cars. So I would check that as well.

Another thing that has me is you say you smell exhaust and oil fumes. Well I know when an engine gets hot things expand and if you have a bad enough exhaust leak you can lack proper back pressure but, that is a stretch to make it not start. You more then likely do have an exhaust leak somewhere because you shouldn’t be smelling that. Also are you smelling it with windows down or up.

Another thing is how are your fluids. Specifically oil but, how are the rest of them. Do you have to add oil. Did you look all over especially the exhaust side of the engine for an oil leak? If you are smelling burnt oil either its being burnt and put into the exhaust or its falling onto the exhaust.

When was the last time the fluids were changed? Time frame for each.

Does your engine fan kick on?
Does your belt squeal?

[Barry Shea]

UPDATE (response to new comments):

1. The smell comes into the cabin with the windows down, and resembles if you took a sniff of the inside of the valve cover through the oil fill cap – part oil, part fuel, part exhaust. And I only smell it when the problem starts.

2. I’ve been super diligent with oil and filter changes, and just changed them (Castrol GTX 10w-30, Fram)

3. Coolant was changed in July 2011 with new radiator hoses.

4. Radiator fan does kick on and cycles on/off.

5. No squealing belts.

6. No check engine light

7. No diagnostic trouble codes reported to the scanner

Guy at O’Reilly’s mentioned that it might be the Throttle Position Sensor. What do you guys think about that?

Also, as for the extremely rough running, stalling when I left off the gas, and not starting until it cooled down, this was fixed with a new cap & rotor (rotor screw had come out and rotor was flopping around inside).

So, I’m back to the original problem I’ve had for the past 4 years: When cold, it starts right up, idles smooth, and runs great. But, on the highway at 65-70 MPH, as soon as it warms up it starts bogging down and running rough. And if I don’t move the gas pedal, it slows down to 40-45 MPH, and then kicks back in. If I slowly press on the gas pedal, there’s no response for about 1/4″, then it kicks back in, but then I’m going too fast, so I let off the gas to slow down to 65-70 MPH, and the problem repeats.

[James O’Hara]

Just because the scanner doesn’t say a sensor is bad doesn’t mean it isn’t giving poor feedback i would check the cam and crank sensors. On top of that it is possible that once your engine heats up you are getting a vacuum leak. Was the intake ever smoke tested.

[Barry Shea]

MDK22:

– Intake has never been smoke tested.
– I’ll check the cam & crank sensors

I did check the Throttle Position Sensor tonight, and all the voltage and resistance values were good when closed and wide open.

Again, this bogging down problem occurs once the engine warms up, cruising at 65-70 MPH. When I slowly press on the gas, there’s no response for about 1/4″, and then the engine kicks back in, and runs OK, but I’m quickly up to 80 MPH, and I don’t know if it will continue to run OK at that speed, because I slow down – who needs a speeding ticket?

Thank you for your continued help.

[Author]

Replies