Wise guys. :D

So I checked the factory manual. I hadn't been able to find the right section before. But I found it. They list 26 separate things to check for a rough idle like mine. The O2 sensor is one of them. This could take a while.

By the way, is that engine fuel-injected, or carbureted? If the latter, maybe it's a jetting issue.

When we were on our last day of driving to Yellowstone, we were about 20 miles out in farmland when a loud hiss arose form the front of the car, Oh Crap!, is it overheating? gauge says not, just the AC wheezing ? then it dies away, weird we drive on.
About ten minutes later it starts again getting louder and softer, sounds right at the front, like it's coming out of the heating vents. After moving my hands around covering things up to see what changes the sound I notice the radio is on, but there are no stations out there it's just the static coming and going
:blush:

It is fuel injected.

So in addition to hooking it up to a code reader, the shop manual says to:
- check oxygen sensor
- check throttle position sensor
- check engine coolant temperature sensor
- check MAP sensor
- check if problem occurs when engine is running Rich or Lean (I think it's lean when it's hot, which it is)
- check evaporative emission (EVAP) control system
- check fuel injectors
- check fuel pressure
- check ignition voltage output
- check spark plugs
- check spark plug cables
- check ignition timing
- check for vacuum leaks
- check the ECM/PCM grounds
- check ECM/PCM for A/C signal
- check EGR
- check battery cables
- check A/C refrigerant pressure
- check PCV valve
- check for broken engine mounts
- check valve timing
- check low compression
- check bent push rods
- check worn rocker arms
- check valve springs
- check camshaft lobes

Holy crap. I can rule out about half this stuff, because it doesn't make the metal clanking sound of worn valves or bent push rods or anything like that. And it has nothing to do with the A/C being turned on. And we just replaced the PCV valve. The spark plugs and distributor cap are not as new as I thought they were (they are about 30K miles old,) but the cables look good and clean on the outside.

Many of those items need not be checked either because they should never fail in the life of the car, or would be obvious from sound and observation.

Never need replacement include spark plugs and wires, rotor and cap, and PCV valve.

Would be obvious for other reasons include: fuel pressure, fuel injectors, battery cables, broken engine mount, and valve timing and those other related entries including camshaft lobes.

These entries should result in a check engine code: oxygen sensor, throttle position, coolant temperature sensor, MAP, most injector problems and low fuel pump pressure, EVAP problem (that will even detect a missing gas cap), ignition timing, and EGR.

I have no idea why refrigerant pressure is on the list unless they are discussing a mechanic who over pressurized the system.

That should seriously shorten a list of top suspects. Having said that, I don't know how 'good' your car's diagnostic software is. But being 1996 or later means a diagnosis is much easier due to that computer diagnostic port.

Have you ever read "Shop class as Soul craft" tw? You might benefit from it. It's about an ivy league college graduate (probably an mba to boot) who's taken up motorcycle repair. He finds a sublime pleasure in diagnosing the trouble in a poorly running machine, kind of like glatt's situation.

I ask if you've read it because there's a thought expressed in the book I think you'd appreciate. The author says that there's an absoluteness to the machine. That despite the theoretical, the documented, the machine doesn't really care. That it is up to the mechanic to observe, to listen to what the machine is telling him, and then respond accordingly, even if it is contrary to "the book".

When you say things like "spark plugs and wires and caps and rotors never need replacement" I just roll my eyes. I don't know anybody who has any actual wrench time that would say such a thing. No one who has spent time actually working on cars would ever entertain such a ridiculous notion in a situation like this. Maybe, maybe in a theoretical discussion of what a perfect world would be like, etc. But come on. That's not what glatt's dealing with. His car's running roughly in some conditions and it is completely reasonable to think that a misfiring cylinder because of a bad spark plug could cause this, as an example.

If you're just ranting, rant on brother. But you're not giving much practical help here with proclamations like that.

My main point is when you say "should never fail" I have to edit it to read "probably isn't the cause of your problem, but ...". It irritates me.

My wife reports that it's not doing the rough idle today. A couple days ago she filled the tank with the super premium gas because she had an 80cents off code from the grocery store. Hmm. Don't know if those two facts are related. I need to continue the observations.

So stop reading so much theory. Learn from those who were edcuatd by reality. So many cars (even in the 1960s) were fixed by learning this stuff the hard way. In the 1960s (due to bad designs), spark plugs and wires had to be replaced constantly. When one who mocks your theoretically trained author, instead, designed an installed an electronic ignition in a 1960s car, then spark plugs never needed changing. How many engine systems did your guru design? How often did he also trace failures back to the source by learning from the hardware?

So many religiously trained mechanics preach, "But the car has a heart". Many are recommending tune ups and wheel alignment only because it increases their income. All cars (even GM) no longer have nor need tune ups. But then I a stating this because I not only know this stuff by doing it. But also designed parts or traced failures to defective designs.

Any car that needs spark plugs, wires, or cap in 200,000 miles was probably something designed by a bean counter in GM. Even GMs parts can not longer fail as frequently as they once did only due to Federal laws.

Listed where the more likely suspects. Only those being scammed by a mechanic need replace spark plugs, wires, or cap even in 100,000 miles. What was once required every 5000 miles or three months was obsoleted decades ago when EPA rules made current technology required in all cars.

TW, I think you need to read Big V's post again.

It may well be that spark plugs should never need replacing, that's the theory. Each car is it's own independent reality, and we've got to allow for the fact that quality control was pared back by MBA managers to save money, and that the car was built and serviced by humans. You know, those frail, irrational bags of mostly water?

It's like with the paperless office discussion. You're not allowing sufficiently for human fallibility.

Human? What's that?

Last time I looked, Limbaugh was being sponsored by spark plug companies. Sandusky recently said to take a defensive position. Replug annually to avoid infant mortality.

The scheduled maintenance for this car makes no mention of changing the spark plugs, but for cars that routinely tow, it says you should change the plugs every 30k miles.

I had read several years ago that plugs today are really good, and that you could go 100k miles on them (or more) but they get to be very difficult to remove if you wait that long to switch them out, so consider switching them out sooner than that.

Did the outside temp change? My money is on the O2 sensor.

We should all bet on it, and whoever is right gets to pay for the repairs.

No.

But you said your money was on the O2 sensor.

Oh, maybe I should have rephrased that. :)

:idea:

Loose wire.

:)

I took it to Track Auto or Autozone or one of those places. And they hooked it up to the code reader, and there were no codes found.
I guess that means the oxygen sensor is OK?

Most likely, yes. But again, described was how to make some code reported. Some codes will not be reported unless a defect is seen in multiple and consecutive restarts.

Most O2 sensor failures I have seen get reported by diagnostics long before it causes engine roughness or lower gasoline mileage. And again, gasoline mileage is an important parameter in locating these intermittent type problems. Some defects will cause mileage reduction. Others will not.

Was it exhibiting the symptoms at the time the test was taken?

Yes and no. It had been rough idling moments before, but it was turned off for the code reading. Key was only turned part way and engine was not running.

part one of two
 

Ok, well then I guess the O2 sensor is not indicating an error.

NO readings means that the problem is coming from some source(s) in the system that don't report back to the computer. Things like ZenGum's "loose wire".

glatt, how much do you know about a 4 cylinder internal combustion engine like the one you've got? I don't want to talk down to you, I'll just think out loud.

When it's running right, two cylinders fire at the same time, like 1 and 4, then two strokes later cylinders 2 and 3 fire (a common firing sequence). Since everything is balanced mass wise and thrust wise, it feels smooth. As soon as something in that little set of sequences doesn't happen (and there's a missing piece, not an extra piece. No phantom fifth cylinder appears to unbalance the weight or the firing. No. Something's NOT happening, like one cylinder not firing unbalancing the engine, manifesting as rough idle). So, what could cause a cylinder to periodically not fire/misfire?

The holy trinity of internal combustion is fuel + air + fire. If you're missing one or more of those you're screwed. And they have to be delivered in precise amounts and precise proportions and at precise times. What could cause an interruption to any of these?

Let's subtract unlikely candidates and see what's left. It's unlikely that you're not getting air to a cylinder. A dirty air filter would restrict airflow to all the cylinders, not just to some, so... I can't think of another likely cause for "missing air" that wouldn't also have some other much more dramatic symptoms. I think your engine's getting the right amount of air.

And, that air is being delivered continuously and at pretty constant pressure and temperature, so it can be treated as a constant, putting the responsibility for fuel/air mixture proportions on the fuel side. You said the engine is fuel injected. Ok, what controls the injection? If it is mechanical injection, it's a masterful clockwork designed to figure out how much fuel to squirt, but your car almost certainly has electronic fuel injection. What electronic fuel injection is about is a system of sensors like throttle position, speed, gear, emissions, etc etc that help the computer decide how much fuel is needed to meet these conditions.

I had an old Buick LeSabre that would sometimes give me grief. It turned out there was a (gah, I hesitate to tell) design weakness that involved the air mass sensor (the MAP sensor in your car). Sometimes the car ran like shit, or wouldn't start at all. I learned to open the hood, and give the plastic intake snorkel a sharp whack, close the hood and carry on, problem solved. It turns out that the sensor was used to determine how much actual air was moving through the engine by putting two different magical detectors very close together, kind of like the rubber hoses you saw in gas stations years ago that rang the bell ding ding when you drove over them? Well the modern version in traffic surveys has a pair of these hoses across the roadway and it can calculate the speed of the cars by measuring the length of time between the tire touches on the two lines. Well this sensor worked like that but the wires were so close together that when a tiny piece of crud was sucked into the airstream, if it caught on the sensor part of the wires (strung directly through the airstream, just like the traffic sensor) it could cause erroneous readings. A smack on the housing usually dislodged the minute particle of junk, the sensor was then able to read the amount of air, and reported back to the main computer and the rest of the calculations would result in the right amount of fuel to be injected; life was good.

Back to your car, something could be interfering with a perfectly good sensor, like some crud on the sensor, and causing it to report bad values, with the downstream effect of a roughness because the cylinder's not getting the right amount of fuel. But which sensor when the computer check says none is ... really tough and expensive.

Maybe you have a tiny piece of dirt in one of the injectors. I have an old house and a very low flow showerhead (2.5 gal / min). Sometimes some rust/crap from the inside of the pipes flakes off and makes it's way to the showerhead. This clogs it and the flow is reduced even further. Not good. I take the showerhead off, clean out the crud, reassemble and carry on. Now, the tub faucet is on the same damn pipes, but I never have this trouble there. The flow/current/pressure (whatever) is just so high that if it even happens, I never notice it. Back to your car. Maybe you have a little piece of dirt in one of the injectors and at idle, low speed, low pressure the dirt can interfere with the squirt of gas, causing a misfire. At road speed, you don't notice because you're pushing a lot more fuel through (like my tub faucet), so problem isn't noticed. To me, this is plausible. Fixing something like this can be pretty hard. The injectors aren't really.... repairable. They're replaceable, but could be bux. The orifice is super tiny, imagine the size of hole that produces a vapor or fine mist, like a perfume mister or spray can. Psssshhhhht. or more like psht psht psht psht psht 800/2 times per minute. You could try commercial injector cleaners that you pour in the gas tank.

What else could cause irregular fuel delivery. It is also plausible that the whole engine is being a little starved for fuel under low pressure situations like this, as might happen if the main fuel filter was moderately dirty. Have you checked your fuel filter? This can be an easy and cheap fix with little downside. Once more, at road speed, this would be less obvious at the higher pressures.

Fuel...fuel... a physical restriction, or bad instructions to cause it to erroneously deliver the wrong amount of fuel... those are all the situations I can think of at this time for fuel. Air, check. Fuel, check. Now fire.

--to be continued--

part two of two
 

--continued--

For fire, you need the spark plug to go zap at precisely the right time. Also you need a big fat spark. Unlike fuel and air, this one is not "quantity" dependent so much, it's really hard to have too big of a spark. You can have too weak of a spark, and no spark is also bad, but there isn't a proportion "rule" unlike fuel and air. But it is definitely timing sensitive. So we're looking for possible faults that could cause a weak spark or a mistimed spark. Ok.

Weak spark could be caused by some deteriorating connection or component working your way backward from the electrodes on the plug to the connector at the top of the plug, to the plug boot, plug wire, other end of the plug wire, to the coil (or whatever is generating the giant pulse of electricity) to the ? and at this point it varies a lot depending on what kind of ignition system you have. This is the same question as what kind of fuel injection, mechanical or electronic. If it is electronic, you only have the ECM (electronic control module(s)) to rely on upstream of the plug wires. There's nothing you can do in there except replace and only on the "instructions" of the code reader (for us shade tree mechanics). You're not getting a code there, so I don't have anything to point to the electronic ignition system, sorry. If you have a mechanical ignition system with a distributor cap (the cap that "distributes" the electrical pulses by rapidly changing which cylinder has a complete electrical circuit) that could be a point of intermittent failure. It also comes with a rotor, same possible fault. There could be other pieces there that are weak at low speed (like points or a condenser). But I don't know if you have a mechanical ignition system. I have a 1996 car (VW Golf) and it has electronic fuel injection and a mechanical ignition system. I have had trouble with the ignition system before.

Trouble's not really fair, it simply needed some regular maintenance which meant replacing some of the components, like the cap and rotor. This was the best $25 I put into the car that whole freakin year. Your car? I can't say. Do you know if you have mechanical or electronic ignition?

Other causes of weak spark could be degraded or damaged plug wires. You could have a bad plug. They operate in severely harsh environments--think Mount Doom--and need to keep performing a very difficult electrical stunt, ionizing air to make a conductive path (fancy ass way of saying spark) a kabillion times. In shitty atmosphere. Sometimes that crud accumulates on the plug and interferes with the spark. Yeah, really. This is an easy thing to check, assuming you can actually get to the plugs and you have a ratchet and an extension and a plug socket. footfootfoot's car doesn't qualify here, since he can't get to all the plugs. He/you certainly could check the ones you can reach though, and if you see one that's bad, replace it. Check them all if you can. There are pics online somewhere I'm sure that show you in living color what the electrode end of a plug in good condition looks like AND what a plug in poor condition looks like. There will be lots of different "bad plug" pictures, since there are lots of reasons why a plug could go bad and be the source of bad symptoms. You can look them up. If you can't find them, report back, we'll find some for you.

So that's weak spark. What about absent spark? That's all the same, just weak to the level of zero. Fine. Oh, just thought that the coil could be weak, that's the part that generates the high voltage needed for a spark. It's like a transformer in your system.

What about mistimed spark? This is less likely, though not exotically rare. You could have a situation where the electricity that would have been your spark is leaking out before it gets to the plug. Like if a plug wire has a fault in the insulation and the spark is grounding out to the engine before it reaches the cylinder. One way to check for this would be to open the hood and look at the engine when it's displaying the symptom of roughness, but do it in a very dark environment. The idea is to have the fugitive spark reveal itself in the darkness. Doable.

Timing... thinking... A car that is mistimed could have a rough idle. What do you know about the car's timing? Have you checked it? Have you changed it? This isn't something likely to change on its own (possible, but pretty unlikely). If *you* have mucked with it or some other mechanic, it is possible that it's been changed from "best" to something less than best. I don't know how else your engine would miss a spark or mistime a spark.

Having thought out loud, it seems the most likely piece is a fuel delivery problem, junk in injector or fuel filter. Next and close behind would be an ignition fault. If it was my car, I'd do these things in this order until I found a likely problem.

1 -- visually check all the ignition wires. disconnect and reseat all these connections. this is free.

2 -- pull the plugs, check them. fouled? cracked? worn? beautiful?

3 -- replace fuel filter.

4 -- run a bottle or three of injector cleaner super duper turbo in a jug or whatever that crap is called through the car a full tank of gas at a time.

5 -- run the car on the TOP HALF of the fuel tank for a while. this probably should be item 1.5 or even item zero.

Tha's a good start. I'm interested to see what happens here glatt. Keep us posted please.

Wow, that's a lot of thought you put in there.

We had the timing belt replaced at the urging of a mechanic a few years ago, and after that happened, I was sure to pay attention to the performance of the engine. I don't remember any changes in performance overnight, so I think they did it right. These problems have been gradual, and I've just noticed it recently because it's reached a certain threshold of being annoying.

We also replaced the fuel filter a couple years ago too at the recommendation of a mechanic. Certainly long enough ago that it could be messed up in the meantime.

I'm leaning towards this engine just being old and dirty as being the most likely culprit, but I'm going to follow your ideas. When I get some time and daylight.

I love it when you can fix a machine by thumping it. Very primal. RAWR! Og mad! Og Smash! [thwack ... vroom!!] Bwahahaha!


Oh, and as well as the above ... you are using the recommended grade of gas, aren't you?

the technical term is "percussive maintenance". It is very satisfying no matter how effective it is.

gas grade is a measure of the octane in the fuel. Octane's role is to change the point at which the fuel / air mixture spontaneously combusts due to the heat from compression. You experience this as "knocking". It is most common when the engine is under a heavy load such as hard acceleration or towing or driving hard uphill. The mixture detonates before the piston reaches top dead center and/or the spark ignites an already lit combustion chamber and the result is a hard knocking sound. I've never heard of octane having much if anything to do with idle loads.

Good points, but I was thinking in terms of it not burning cleanly and being the source of gunk that could cause the problems you mention.

I've learned to check the very basics first. I've got part way through helping a friend remove the door skin from the back door before asking why he was doing this. The inside door handle wasn't working. Oh, does it have a child safety lock? A what? One of [click] these. :smack:

Then there was the friend who's car had broken down. After ten or fifteen minutes of experimenting, then calling the mechanic service, I noticed the fuel guage on E. :smack: Done this twice, about ten years apart. Same friend. :facepalm:

The car doesn't require high octane fuel.

Our regular corner gas station used to be a Texaco. But then Texaco pulled out of Virginia for some reason. So all the Texaco stations became "Liberty" stations. It's more of a discount gas station, but not significantly cheaper. I assumed it was just as good as Texaco. Around the same time, our grocery store began offering significant discounts on gas at Shell stations when you buy lots of groceries. We have a family of 4 and buy a lot of groceries. So most of our gas has been bought at a Shell, but in the last year or two, we have been buying some of our gas from this Liberty gas station on the corner. Maybe Liberty sells crap gas? I don't know. The dates roughly match the symptom dates.

My Beetle is supposed to use regular gas. I guess they've perfected engineering enough that we don't have to fall for the more expensive gas being better. I love that about my car. Not "we recommend the highest grade gas possible" but "USE REGULAR gas, or else."

It's written right inside the door to the gas hole.

Read the many previous posts about this. Add maybe 26 cents per gallon to the price Liberty posts. Discount gasolines are the most expensive as explained previously.

To know more and to fix a rough idle, do gas mileage for every tank. Also posted previously. My experience says it takes two to four tanks of the good stuff to undo damage caused by discount gas.

And again, discount gasolines increase the nation's imports of foreign oil by 8% to 14% (numbers based upon my calculation; increased imports made obvious by your MPG numbers).

Major gasoline manufacturers are selling less gas because so many Americans even believed Saddam had WMDs - by ignoring facts and numbers. How many still do not learn from history? Notice how a majority are so easily decieved even by discount gasoline.

tw, how does recording the mileage the car gets per tank fix a rough idle?

It won't. He is trying to associate the two issues. Additional data points... Also, he is trying to see if the "good stuff " is working - maybe.

A solution only exists when the problem is first defined. Some of those suggestions cause significant mileage decreases. Others (ie oxygen sensor) do not. Just one of many reasons why mileage numbers are important for solutions. Better help means first providing better information. Only relevant fact is roughness in idle. Nothing else identifies meat necessary for a solution.

Random fact:

My subaru *can* run on 91 octane, but the owner's book says to put in a bottle of fuel system cleaner every 6,250 kms (half of a service interval).

I couldn't find any in shops so I rang the Subi dealer. They said "yeah we stock it, but we don't recommend it, just run your car on 95 octane, it will keep it cleaner".

These guys could make money selling me this stuff, so for once I believe them.

So I was wondering if using El Cheapo gas might be dirtying up the system and causing these symptoms. Just checking the basics.

Wait - there is fuel in it, isn't there? ;)

Generic green beans are the same as the brand green beans, generic Wheaties are just as good as real Wheaties, I hear them cry. BS, I say.

Yet there is 'cheap' gas? Gas is gas is gas.

You're all eating a pile of hooky.

I think you need to check your tire pressure. When my tire pressure gets low it decreases my gas mileage. In fact, when my gas mileage is lower than I expect I always check my tire pressure.

Tire pressure is causing a rough idle now?

It's always a good idea to check tire pressure. The car was pulling slightly to one side a while ago, and I inflated the tires to the proper pressure. It fixed that. I go in spurts when it comes to car maintenance. It's probably been a couple of months since I checked the tire pressure. Couldn't hurt to do it again. I suggest all of you go check your tire pressure this week.

Car maintenance - it's complicated!

It's not so complicated, but it's a pain in the ass.

Ah. Now I get it. I had to stop reading between the lines.

yeh, after that ... (shrug)

Find a different target, fellas.

tw tries to participate, and had been making some great funnies lately, but it always seems acceptable to be shitty to him. Why is that?

Anyway, the highlighted text makes no sense. tw doesn't live in Badelaide Baustralia.

I'm not being shitty to him. As for the joke, it went like this: tw (lots of lines of text followed by the words) random fact. I "got it" when I "read" "tw random fact", and not the stuff between those lines. It's a reference to his earlier post where he said to fix a rough idle, measure your gas mileage. That's nonsense.

I am really not being shitty to him, I'd love to solve this rough idling problem glatt's car has. I don't know everything there is about cars, and I love to learn new stuff. But unless tw has some magic trick he's still hiding, I don't see how his explanations achieve the desired result. If they do, that's freakin awesome because 1 -- problem solved and 2 -- I learn a new trick. If they don't, well, ok, fine, I already have a bunch of tricks that don't work.

I want him to explain his idea and show how it works. That's the beauty of working on machines (as I tried to explain several posts before) that machines don't need or are even susceptible to persuasion. I don't want to be persuaded that measuring the gas mileage will fix a rough idle, I want to be *shown how*. Maybe I can understand it, maybe not. But I haven't seen any explanation yet.

I was poking fun at tw's answers because they're freakin random.

I guess I'm overly sensitive on tw's behalf.

Sorry, and carry on.

That is the point. You do not understand the relevance. Do not understand how that information results in solutions.

Solve problems in two steps. First collect facts. Much later solve the problem. Some of that stuff posted fix a problem without first determining what is wrong is called shotgunning. Could even exponentially complicate his problem. Those was some of the most random and irrelevant wild speculation I have read in a long time. Any mechanic that did that would have been quickly unemployed. You should have been mocked for those consecutive and irrelevant recommendations. But then making all those mistakes is how we eventually learn. I just ignored them suspecting that glatt also would probably be doing same.

Embarrassing was a claim that two cylinders fire simultaneously. And a few other bogus claims. Each cylinder fires separately. You should have known that since it was taught even in primary school science. No reason to mention that until now that you decided to entertain your ego. I was not going to say anything then. But now you need to learn some humility. Big time.

We have a list of suspects from the shop manual. And a few others not on that list. Facts such as diagnostic codes, information collected by a portable computer from the dealer, and gasoline mileage all provide relevant information. Facts to significantly reduce a list of suspects to but a few. Even the fact that roughness did not correspond to an engine code was useful information.

Also important are conditions that cause or co-exist when roughness does and does not exist. Whereas tire pressure likely is not relevant, it is still a change that should known. Never short the help of a change only because you consider it random or irrelevant.

Use the oxygen sensor as an example. Some assumed an oxygen sensor could cause rough idle. Why? An oxygen sensor is mostly about operation at high speed; not at idle. An oxygen sensor (in most designs) has no influence on how idle works. Why then did someone suspect an oxygen sensor? An example of implementing a solution long before a defect was even defined.

Gasoline mileage goes a long way to exonerating many suspects. But not everyone would understand that for a same reason why some might automatically suspect an oxygen sensor.

Another example was recommended. Full to the floorboard acceleration repeated after each engine restart. So that diagnostic codes might report some other defects. It should have been done especially when one has no idea what that will discover. What was to you random was actually targeting specific suspects.

You have no idea why gasoline mileage numbers were important. Numbers more useful had those numbers been taken when the car was running better. Numbers that would have even said more if taken for a tank of Liberty and another tank of Shell.

Moving on. Another explained what high octane gas is. It is not cleaner or better. High octane does not burn out crud. It is only different. Some brands include additives that make other differences.

For example, Mobil once contained high detergent additives. Therefore gas was cleaner? No deposits? Nope. High detergent levels in Mobil caused increased carbon deposits on valve stems.

Same applies to high octane. High octane can increase engine wear and failure on some parts. The word high does not mean better (except at the end of the day in a room with a towel covering that hole underneath the door).

I push in the trip odometer every time I fill the tank, and it's usually around 280 miles if I've been driving around town and 310 miles or so if I've been on the highway for the entire tank. I haven't noticed a huge change in those numbers. How big would the change in mpg have to be to be significant? I haven't noticed a change, so I figure the mileage hasn't changed by more than 5%, if it has changed at all. I think it has stayed the same.

The difference between crap Liberty gas and Shell gas would (for example) result in maybe an 8% difference without any noticeable change in idle smoothness. If your numbers are as reported, then Liberty gas should have caused a noticeable mileage difference. The fact that you saw no difference is troubling.

In another event, a tank of Mobil destroyed my fuel injectors. I knew something was wrong due to a slight roughness in idle. But the resulting mileage change was from 32 MPG to about 27 MPG in that tank. And 24 MPG in the next tank. (Honda replaced those injectors for free). In that case, diagnostics never reported a problem.

In another case, a problem with an EGR valve kept causing a check engine light. No mileage change. Symptoms that helped me trace the problem to a hard and not easily diagnosed problem. (BTW, EGR valve is not connected to the idle system; so did not affect idle. And that failure was unique so as to not affect gas mileage.)

Examples of different failures that do and do not result in mileage changes. And that isolated the problem quickly to a shorter list of suspects.

Assuming that 5% coincided with road conditions (heavy traffic, summer open road driving, etc), then is we can assume you have a problem traceable to the idle system and that does not affect the other 'normal operation' system.

BTW, remember Cloud's problem. That was also traceable to the idle system and that would not cause mileage changes. That mechanic had one clear advantage. The Civic's diagnostic system pointed exactly to the defect - which that mechanic refused to address. An example of why some people just refuse to learn good diagnostic procedure.

Now, with Shell, what is happening? What are the latest symptoms?

Well, the roughness seems to be not so bad the last week or so, but I'm not sure. The roof used to buzz from the vibration and that's not happening recently, but there is still some roughness in the idle after it warms up. You can feel it in the steering wheel and the cds in the door pocket rattle. It still goes away when you give it some gas.

I have old records of the mileage from when this car was new and I was better at tracking that stuff, and I can compare that to the mileage the next time I fill the tank, but keep in mind that gas didn't have ethanol in it back then, so it would be comparing apples to oranges.

Is it worth it to buy a bottle of fuel injector cleaner and dumping that in the tank? Would that harm anything other than a few wasted dollars?

Also keep in mind that we don't drive much. I take the subway to work, so it might be a couple weeks before we fill the tank again.

Is that a new symptom? Roughness was when warm; not when cold? Or is roughness same when engine is warm or cold at 1000 or 1200 RPM - the cold idle speed?

I would avoid even trying to fix anything. First establish some trends. You still have some Liberty gas in there. And you really do not have MPG numbers. Only mileage for some approximate gallons of gas.

Stated earlier: first get facts. Dumping that miracle fluid in only adds another variable; can create confusion. One step at a time. You describe something that is not serious. You can live with it for a few months until the problem makes itself obvious.

Or you can take a nuclear option. Spend more. Take it to the dealer. My preference is to learn from an anomaly that is minor - will not leave you stuck. Not just learn how a car works. But how to better diagnose any problem. A memorable moment in life is that "Aha-Ha" moment. When that solution is sudden, obvious, and appears to materialize from nowhere.

I would avoid that additive until current trends stabilize - ie enough tanks to accurately identify gas mileage or to see roughness stabilize – become predictable. Currently roughness is different at different engine temperatures? Or just at different idle RPMs?

right, RIGHT. Exactly. Now, at last, you understand my question. Of what possible relevance is gas mileage measurement fixing a rough idle?
You have some facts. You don't have all the facts. You'll never have all the facts until the problem's solved.

No, it's not called shotgunning. It's called an iterative process, a process of elimination.
Well, he's not taking it to a mechanic, he's asking us, asking me. I think a mechanic that did the things I suggested would be employed far longer than a mechanic that did the things you suggested. Just curious, what part of my recommendations do you consider irrelevant?
Cite. An even firing sequence is pretty common, anything else is unusual for an inline four.
Cite.
Cite.
Buddy, I've been sayin that from the beginning. If you have something to teach me, bring it on. To this point you have not been forthcoming. I've asked you and asked you for your explanations, only to be be met with silence and bloviation in equal measure. You have some mechanical explanations for what's happening here, I'm interested in learning them, humbly. I'm still waiting...
Sure. We have some facts, ok, I agree. Now what do you suggest to do with the facts you have?

I short the help of a change (???) when it a distracting obfuscation. Like gas mileage. You do know that the gas mileage of his car when idling is zero, right?

Really?? For a rough idle? HOW? How? That is my question, how does knowing the mileage fix the rough idle? It doesn't man. You might have posed a better question by asking how the mileage has changed. But even then you would have gotten a small range of answers, same, better, worse, LOTS better, LOTS worse, all over the place, stuff like that. But even knowing which of those answers was the "fact" leaves you with no suggestion as to what to *DO*.
If you do, please tell me.


So, based on what facts we all know, what do you think could be the reason for the rough idle and what do you suggest to resolve that symptom?

What was posted was classic shotgunning. As I said before, a major reason for fixing things is to learn. Eventually you may learn that those long posts were classic examples of shotgunning. You were even trying to fix things that were not relevant to his symptoms. Trying to fix things base upon "it might be this so do that". That is shotgunning.

Some of your suggestions would fix things not even relevant to the symptoms. Understandable because even some basic concepts, essential to auto repair, were unknown. A firing order for a four cylinder engine is typically 1, 3, 2, 4. Two cylinders do not fire simultaneously. That would cause imbalances. Explains many suggested actions that were irrelevant. Shotgunning often happens when underlying concepts are not first learned. Again, cylinders do not fire simultaneously.
Exactly the point. Once never cures symptoms. Symptoms are part of the process of identifying a defect. Fix defects; not symptoms. First identify the problem. Fixing comes much later. Shotgunning is suspecting and then fixing things when one has not a clue.

In an example, you asked how gas mileage fixes rough idle. Again, fixig comes later. Gas mileage is the process of first identifying the defect. Using gas mileage to fix roughness is an example of 'shotgunning'.

Based upon facts, a defect is not yet identified. A long list of suspects exist. Many might be eliminated with better knowledge, labor, or equipment. Ie. remove a suspect to test it on the bench. But that would take time, more posts, and sometimes creates new problems. Instead, use better knowledge, for example, to know an O2 sensor is an unlikely suspect (based upon symptoms provided).

No rush. The car still works. Plenty of time to identify the problem.

Facts provided a reduced suspect list - posted earlier. Still too many. Listed was how to obtain more facts since even mileage numbers could reduce that list. Mileage numbers are still too subjective.

However if you know more, then define what will identify or exonerate each suspect. The list is there. If you know more, then detail how to identify the defect from that suspect list.

Other relevant questions were asked such as behavior at various RPMs and temperatures, changes after Liberty gas has been fully displaced by regular Shell, operation so that engine diagnostics can see some problems, and using an onboard computer to monitor engine parameters when roughness does and does not happen.

In thinking back, I don't remember. Another of many still unanswered questions. Were spark plugs and wires changed a few thousand miles before this roughness started?

Trying to get a straight answer from TW is like trying to nail jelly to a tree, innit?

Someone drove the car until the gas was very low, and/or
the gas cap was not put on tightly so the check engine light came on.

Full tank of fresh gas + tight cap + running engine 30+ minutes => light goes out = problem solved. :rolleyes:
.

I'm tired of arguing with you tw.

I misspoke about two cylinders firing at the same time. What I was trying to communicate is that two cylinders are UP and two cylinders are DOWN at the same time. One of the UP cylinders is on the compression/power stroke and the other UP cylinder is on the exhaust/intake stroke. My point is still valid and that is if there's something that causes one of the cylinders to misfire, that will cause the engine to be unbalanced. This in turn is felt as a vibration, a roughness.

As for the rest... we just disagree on a couple of things. What you call shotgunning, I call eliminating possible problems. What you call gathering information, I call sitting on your thumbs. What you call no rush, I call no clue. That's fine. You seem to want more and more information until you know what to do, and then you want to spend one bullet only to kill the problem. That is definitely one way to fix problems. It's not the only way, and in my opinion it is not a very effective way. But it's your way and that's cool.

The rest of your crap, whatever dude.

Why are your arguing? Did you always argue with your algebra teacher when concepts were too hard?

What you call eliminating problems is classic shotgunning. No way around reality. Since you never learned good diagnostic procedure means it must be wrong?

Many never learn this stuff. Assume it is too complex; therefore must be wrong. Get angry and become argumentative rather than learn. No argument exist. Just facts. Fatigue is you fighting to avoid learning. Only you can choose to learn or remain argumentative. Well proven reality, for some reason, distresses you. As if learning this stuff would somehow hurt you.

OK, so how an engine works was completely contrary to what you accidentally posted. Fine. And still irrelevant to the topic - rough idle.

glatt has options (and outstanding questions) to find an actual defect. Or to whittle that list down to but a few suspects. Apparently you want him to keep replacing parts as a mechanic did to Cloud - four times and how many $hundreds? A new mechanic who did not shotgun meant Cloud's defect was eliminated the first time. He followed the evidence. A concept also defined in Japan as “work smarter; not harder”.

Good diagnostic procedure finds a defect before wildly replacing good parts on speculation. Follow the evidence. They even made TV shows based on the concept.

Actually the fifth. The first mechanic had already tried four things.
Um, isn't that actually the min-max rule?

jus sayin.

A second mechanic was forced to testify before the grand jury?

no, that was the gas station attendant that noticed the wipers malfunctioning.