Idle rougher when warm?

Vehicle in question is a '91 Acura Integra. It has a manual transmission and 327,000 miles. Driveability is about as good as I can ever remember it being (bought new in '91).

No hesitations, flat spots, sags, surging, hard starting, stalling, nothing at all like that. Idle has always been very smooth, at least as smooth as, or smoother than, the 4-cylinder engines in other and newer cars I've been in. Idle speed remains normal at a very steady 750 rpm.

However... in the last half-year or so, the engine has started to have brief periods of warm idle where it idled noticeably slightly rougher than it normally did . These episodes were few and far between and seemed to have no connection to any obvious environmental circumstances.

The episodes of slightly increased roughness have increased in frequency to the point where smooth idling is the exception. It also seems to me that fast idle (when cold) is a bit rough now as well. Fast idle was always completely glass-smooth. Part-warm idle has been noticeably rough for several years, but that always cleared up once operating temp was reached.

I have noticed that idle roughness seems to be tied to ambient temperature. If the ambient is under about 70, idle smoothness is usually what I'm used to. The warmer it is, the greater the roughness.

I can see nothing obviously wrong in the engine bay. The car is, IMHO, exceptionally well-maintained. Gas mileage is perfectly normal, and I do record that very diligently.

I do not have a DSO or other equipment (other than a VOM) that would help me see what the many sensors are doing, so I'm asking here for ideas before spending money or contacting my guy, who is 50 miles away..

"Paul Hovnanian P.E." wrote in news: snipped-for-privacy@hovnanian.com:

Nope. Check Engine light works properly, too.

That's only a few years old and checks out fine according to the Honda shop manual. I've studied everything as far as I can with no success.

The intake air temperature sensor /does/ use the very same element as the coolant sensor though (it tests OK too). And it is original to the car. It's only $40 to replace, but I'm not a big fan of the shotgun approach...

Any "Check Engine" lights?

I'd suspect a failed coolant temp sensor. Ususlly, that would store an error code and turn on that light. But if the sensor was just starting to go out of spec, the controller might not catch it.

If you can find it and its specs (resistance at various temp points), this might be something you could check with a VOM meter, a thermometer and a pot of water on the stove.

Hachiroku $B%O%A%m%/(B wrote in news: snipped-for-privacy@e86.GTS:

Yep. What do I know about this sort of thing? Little. That's why I don't try to answer similar questions from others.

I wondered that myself. One thing I thought I'd try to do is to take the car in for a Motorvac service, which is supposed to help with such injector problems. But that costs $110, so...

If I had a DSO, I could hook that up and see for myself, but I don't.

My money is on a leaky injector.

Have you "read" the plugs? The rich cylinder should look "different" than the rest.

Al

Tegger wrote in news:Xns9C67BA31C6A22tegger@208.90.168.18:

Update from this morning: I now wonder if I may have a head gasket starting to fail.

I pulled all four spark plugs. All looked the same (whitish) except for #2, which was noticeably yellowish-brown. That one was also slightly wet, with what I couldn't tell. It did not smell like gas. The deposits were soft and easily wiped off. The plug also dried off once exposed to the air.

The piston tops all looked the same (dry black), except for #3, adjacent to the one with the wet plug. #3 had slightly wet-looking black carbon.

In previous years all plugs and piston tops always looked identical.

I've noticed the coolant level in the reservoir goes down about 1/8" every few weeks, which is not normal for this engine. There are no external coolant leaks that I can find. The rad remains fully topped up, right up to the bottom of the rad cap.

The last time the head gasket failed, it was at about 200,000 miles. This gasket has been in for 127,000. The last failure failure manifested as bubbles in the coolant and a greatly elevated reservoir level.

I'm just going to continue to monitor things and see if there are any further changes.

At the risk of sounding foolish (to which I am no stranger, nor ever very scared of), I seem to remember in my lifetime running into a few engines that did this because of a valve leak which was not as obvious at cooler temperatures.

Has anyone else ever seen this?

If it's intermittent and not setting any codes, I'd put 95% probability it is a connector issue. Corroded connector flexing varying the connection resistance without going out of range enough to set a code. A wire chafing and shorting to ground or breaking and going open should go far enough out of range to set a code. The vehicle is certainly old enough to have failing connectors.

Oh, yeah. I remember. Acura with THREE HUNDRED AND TWENTY SEVEN THOUSAND MILES?!?!?!?!

Drive it, then say good bye, unless you're like me and can't bear to part with the car; if so, drive something else until you do the gasket!

What kind of shape is it in? Got pix?

Anumber1 wrote in news:IxUhm.101103$ snipped-for-privacy@newsfe10.iad:

See my update, sent this morning.

"hls" wrote in news:i_

I did my annual valve clearance check this morning (see my update, sent this AM). Four had changed since last year by about .001". Most years there's no detectable change to any of them.

I don't think the valves are leaking. That one wet plug is a definite concern though. If I'm getting coolant into that cylinder, why isn't there exhaust being pumped into the coolant?

Well it could be a crack above the valve seat up in the port wall.

Or it could be that the leak is only under vacuum.

Might even be something like an intake leak between a coolant passage and the intake port. I think the 1.8 has a port between 2&3, but it's been a while since I've had one off.

"Steve W." wrote in news:h69l6a$f9h$ snipped-for-privacy@aioe.org:

Either way, sounds like bad news.

I'll wait a few weeks and see if things get worse.

Hachiroku $B%O%A%m%/(B wrote in news: snipped-for-privacy@e86.GTS:

Yeah, that one. The one with the ultra low-geared 5-speed that sends the engine to 4,000 rpm at 80 mph. And I spend most of my time on the highway.

How long can an engine last when subject to such treatment?

It's my hobby. I don't golf, have a boat or frequent hookers, so this is my hobby. And unlike hookers, I know this car's history, which is what I like about it.

Not sure it's the gasket. See Steve W's response.

Not recent ones, no. But there is zero rust and no body damage. The paint isn't so great, but that's the original coverage which shows touch-up repairs and sun damage.

I have heard/read of cars going over a million miles. cuhulin

snipped-for-privacy@webtv.net wrote in news:19936-4A889455-1951@storefull-

This one won't. Not without a lot of help, anyway.

A few things spring to mind... Are all the grounds good? Throttle plate gummed up? EGR sticking (if equipped) PCV valve wide open?

HTH, Ben

ben91932 wrote in news: snipped-for-privacy@e27g2000yqm.googlegroups.com:

Yep.

Nope. Cleaned that a month ago (it was clean anyway).

No EGR on this car.

PCV valve rattles when shaken. Pinch hose shut, idle drops a bit then recovers.

Voltage drop grounds?