[Rich]

What do some of your other specs look like? For example, fuel pressure. What fuel pressure do you get with the regulator vacuum hose connected and disconnected (as per service manual)?

Something is causing the lean condition. This means that the computer thinks that too little fuel (or too much air) is being received. Thinking beyond vacuum leaks about things that cause lean conditions:

– restricted fuel flow (regulator, pump, leak)
– vacuum leak
– defective throttle body

Some of these fit better than others given what you describe. So far, the best explanation I can think of is a vacuum leak of some kind.

[Rich]

This may be a long shot, but you didn’t mention whether or not the wheels had ever been balanced.

I imagine this would be done for any tire replacement (as you note the tires were replaced), however, I wouldn’t assume the wheels had been properly balanced.

You might consider taking the car to a shop specifically to get the wheel balance looked at.

[Rich]

With a OBDll reader hooked up after the car has warmed up, my STFT is at +18% and LTFT is at about +17%. When is rev the engine to about 2,500 rpm the STFT drops to a normal range of -.2%-+3%.

This is the classic pattern for a vacuum leak. Your title indicates “rich at idle” but positive fuel trim actually indicates a “lean” condition. This excellent two-part video series explains the principles in detail:

Later you note:

I have checked with propane all over for vacuum leaks and checked timing with a light. No luck finding a cause so far

This is a good way to detect vacuum leaks. Carefully point an open, unlit propane torch in the vicinity of various parts of the engine compartment. When you reach the point of a vacuum leak, you’ll notice short term fuel trim leveling out toward 0% (and possibly RPM changes).

You also note:

new intake/exhaust manifold gaskets

The full contact area of the intake manifold isn’t easy to reach in the D16Y7 because this gasket sits underneath the throttle body. Ditto for throttle body gasket. I wonder if you may have missed this area in your vacuum leak tests.

A lean condition under load but not idle can also be caused by a vacuum leak. In this case, however, the engine adds fuel despite there being no need to do so. The air excess air is introduced after combustion and so fuel is added unnecessarily. This may be the reason you’re seeing sooty exhaust. For this reason, it seems worth another try to find a vacuum leak in the vicinity of the manifold gasket (before the O2 sensor).

Take your time with the propane and be thorough. I can’t think of any other cause for what you’re seeing.

[Rich]

What do your long/short fuel trims look like at idle and under load?

If you don’t have a scanner that will give you that information, get one – it’s an essential diagnostic tool. Here’s the theory behind fuel trim:

[Rich]

> Recently it was diagnosed (professionally) as a small leak in the head gasket.

On what basis was that diagnosis made? For example, one method could be a leak down test.

Also, when was the last time your water pump was changed? It’s usually done with a timing belt.

[Rich]

You’re seeing negative fuel trims (rich condition) on both banks at idle. Whatever it is, the problem is unlikely to be bank-specific such as leaky injectors.

What happens to fuel trim at higher engine speed? Do long term fuel trims go back toward zero? If so, this would be consistent with a bad fuel pressure regulator. You can check this.

You mentioned fuel pressure was within spec. Was this while the engine was running or with key-on-engine-off? If the latter I would try getting a reading with the engine running.

Specifically, I’d check fuel pressure while the engine is running, both with the regulator vacuum hose attached and without. There should be specs for both states.

Are your readings for both within spec?

[Rich]

You’ll need to systematically eliminate possibilities. It doesn’t appear you’re doing that.

Your car needs seven things (and only these seven things to start):

1. Fuel
2. Air
3. Spark
4. Mechanical Timing
5. Electrical Timing
6. Exhaust
7. Compression

Start with any one of these categories, then systematically eliminate it by performing tests.

Often, the order is chosen by difficulty in doing the diagnosis – there are no clear indications of what is the trouble (or it’s your first time).

For example…

Fuel is easy to get a quick read on. Check your fuel pressure with key-on-engine-off. Are you within spec? Don’t guess, and don’ just assume that because you see fuel when disconnecting the line that pressure is sufficient. If you don’t know the spec, look it up.

You say you checked “fuel pump and injectors.” That’s not the same as measuring your fuel pressure.

Air is often overlooked. Is there any blockage in your air intake? If you crank while holding down the accelerator, does the car start (bypassing the IAC valve)?

And so on.

This video will teach you a system that works every time for diagnosing these problems yourself systematically:

[Rich]

Had a very similar issue on a Civic. The IAC valve was really dirty. Cleaned it and the problem went away.

But… I cleaned it really well.

One of the other leads I was chasing was the engine coolant temperature sensor. The idea being that the engine needs more fuel on a cold start. When the sensor fails, it fails open – resulting in a false low temperature reading. This simulates a cold start even at high temps.

The cylinders end up getting flooded with fuel on a hot start.

I tested my ECT and it turned out to be fine. IAC valve was the last thing on the list of possible culprits if memory serves. Main relay was also examined without finding anything wrong.

[Rich]

What does your fuel pressure look like at idle and under load?

How does that compare with what your service manual says?

You should do the test both with the fuel pressure regulator connected and disconnected as per your service manual. What does that show?

[Rich]

… cylinder 2 caused the coolant to rise and flow out the top, but not to bubbles, and cylinder 1 cause the coolant to spurt everywhere.

Cylinders 1 and 2 are connected to your coolant system somehow. You also noted:

Did the engine block test with the chemical and it showed there was exhaust in the coolant …

Taken together, these observations are consistent with the following causes:

– damaged head gasket
– warped block
– warped head
– cracked block

However, it’s hard to reconcile any of these against this:

Checked where the oil is added and the oil looked good, no milky froth.

Have you recently changed the oil? If so, did you notice any coolant mixing? If not, what does the oil from the drain plug and dipstick look like?

Is there any sign of oil in the coolant?

[Rich]

If I continue to turn the key in the ignition of and on again and again, the car will eventually start. On the times it doesn’t start I hear the starter click.

Have you tried removing the starter and bench testing it?

[Rich]

You might consider the ignition switch. Here’s how to diagnose it the next time the no-start occurs:

Civics of that vintage are well known for ignition switch issues and there was a recall for the problem in the early 2000s.

[Rich]

Here’s a thorough tutorial on how to diagnose cooling problems on a civic – you could probably find an equally detailed procedure for 95 Accord, but I don’t think that would be necessary.

[Rich]

There’s no need to wring your hands about this. The next time the symptom manifests itself, do the diagnostic procedure in Eric’s video.

[Rich]

It’s not clear from your post whether the cooling system has ever been properly burped since the overheating issue began.

Air bubbles in the cooling system can wreak havoc and cause overheating. Bubble near a temperature sensor are the worst because they will cause the ECU to react to the wrong engine temperature.

So… have you yourself – or a different mechanic than the one you initially took the car to – ever properly burped the coolant system? Lisle funnels are extremely helpful for this procedure. More here:

[Author]

Replies