Catalytic converter unnecessary?

November 4, 2009 5:43 AM

Do I need to replace the catalytic converter and oxygen sensor in a 2000 Honda Civic (125k miles)? Engine suddenly idling low.

My Check Engine light came on night before last. The technician says the codes it gave are for the oxygen sensor and the catalytic converter (which upon inspection, is physically cracked). They said the purpose of the one is only to detect problems with the other.

But in my experience, a faulty catalytic converter makes zero difference in emissions tests (one went bad in a Toyota I had but for years after, the techs said it was as low on emissions as a new car). We just had this one tested.

My bigger concern is the engine is now starting to idle low, as if it wants to stop, but is that coincidental? I asked the tech but their explanation did not sound confident about a relationship between these parts and low idling. It sounded like, "sure, why not."

Are the parts unrelated to low idling, and do I really need them if the car has low emissions?

posted by rahnefan to home & garden (17 comments total)

If the O2 sensor has failed, that could posibly be the cause of your idle issue. Did you ask the mechanic is simply replacing the O2 sensor would work? I'm betting not, if the cat itself is physically damaged.

If you need to replace the cat, there are relatively inexpensive (in comparison to the OEM Honda part) replacement cats on the market.
posted by Thorzdad at 6:05 AM on November 4, 2009

A bad cat can cause major problems with emissions. Or not. Depends on what's wrong with it (clogged vs. used up catalyst).

Catalysts clog up if the engine isn't running right. A bad O2 sensor is probably the most frequent cause of this. Furthermore, the crack could be from overheating of the catalyst rather than physical damage.

A clogged cat can might make the car idle poorly (though it'd more likely cause low power at higher engine speeds) because it's like trying to exhale through a drinking straw.

Anyway, I'd fix that before I tried anything else.
posted by paanta at 6:15 AM on November 4, 2009

Also, if your mechanic thinks the purpose of the O2 sensor is only to detect problems with the catalytic converter, you DEFINITELY need to find a new mechanic.

I don't know hondas too well, but on every (newer) car I've had, the computer figures out the O2 sensor is bad by comparing the values from both the pre-catalyst and post-catalyst sensors. ie, it can't really tell you which of the two (or more) sensors is non-functional. If the one after the cat is bad, the one before it is very likely bad (even if the car doesn't think it is). The one before the cat is the primary feedback for trimming the entire fuel injection system richer or leaner. Without it, the engine control computer is just flying blind.
posted by paanta at 6:24 AM on November 4, 2009

It depends on your region's inspection process. From the sound of your question, I'm assuming you have an upcoming inspection? If so, a lit check engine light is typically enough to fail. If your region does emissions testing, you might have issues passing with that cat, or you might not. Only way to find out is to try.

But in my experience, a faulty catalytic converter makes zero difference in emissions tests...

I've had cars fail, and then simply replacing the cat was enough to make them pass (two different cars). But my state has adopted strict emissions standards, so a lot of older cars initially struggled to meet them.

Are the parts unrelated to low idling...?

I'm pretty much with paanta here. A faulty O2 sensor can cause a variety of problems. This is the great thing about on-board diagnostics — the car tells you what's wrong with it. So fix it.
posted by knave at 6:29 AM on November 4, 2009

Most emissions checks for new cars consist of plugging the emissions station computer into your OBD-II port and reading the codes. So if you're showing codes for a bad O2 sensor, you'll definitely fail.

Also, definitely new mechanic. Try CarTalk forums.
posted by electroboy at 6:37 AM on November 4, 2009

In my state a check engine light is an automatic fail on the emissions test.
posted by COD at 6:51 AM on November 4, 2009

Most emissions checks for new cars consist of plugging the emissions station computer into your OBD-II port and reading the codes.

Cite, please. All emissions checks I know consist of measuring the pollution that your car actually produces, not what the car's ECU thinks it is producing. Perhaps you mean "while checking your emissions they may also scan your ECU fault codes"? I don't know of legislation that can allow an emissions fail if the emissions themselves are actually within tolerance (which you are implying is the case).

OP: Yes, the mechanic doesn't know what an O2 sensor does (unless he didn't make himself clear and was only referring to the rearward O2 sensor, I guess). Both the catalyst and the O2 sensor may be the cause of the low idle (even the crack may cause low idle through incorrect back pressure in the exhaust system - the ECU isn't able to correct for this).

The cheapest fix is the O2 sensor, and I'd replace that and see if the idle issue goes away. If it does, I'd be inclined to ignore the Cat issue until it failed the emissions because they are so damned expensive.
posted by Brockles at 8:01 AM on November 4, 2009

Cite, please.

Cite
posted by electroboy at 8:20 AM on November 4, 2009

Georgia says this about it [pdf] -- sounds iffy though; it expects that the check engine light will cause the test to fail. That's not the same as a technician automatically failing the car if the light is on.

Thanks everybody.
posted by rahnefan at 8:21 AM on November 4, 2009

Cite, please. All emissions checks I know consist of measuring the pollution that your car actually produces, not what the car's ECU thinks it is producing.

With only a few exceptions (like for pre OBD-II cars), that's how it is in Illinois.
posted by iknowizbirfmark at 8:23 AM on November 4, 2009

Well it looks like the OP is in Georgia (from recent questions, but Ontario (and the UK checks when I was living back there) are different to those mentioned here. From The Drive Clean Brochure for Ontario:

The vehicle is warmed up to normal operating temperature. A gas analyzer, connected to the vehicle through a probe in the tailpipe, then measures the concentrations of oxides of nitrogen (NOx), hydrocarbons (HC) and carbon monoxide (CO) while the vehicle is being “driven” on the dynamometer and again when the vehicle is idling at the end of the test.
...
The test results are compared by a computer to emissions standards for the vehicle and its mode year, with an allowance for vehicle age and wear and tear.
...
If your vehicle does not meet emissions standards, you will also receive a fact sheet describing the most common reasons for vehicles failing the test and information on what to do next.
...
If the Malfunction Indicator Light (MIL), also known as the “Check Engine’’ light, on the dashboard remains illuminated in 1998 and newer model year vehicles, the owners will be advised that it may indicate a malfunction of the vehicle emissions system."

So the OBD system is tested and advice as to its effectiveness is considered, but it does not dictate a pass or fail. That seems far more logical than just trusting the OBD system, to me. Odd that some states have dropped actually testing the emissions completely.
posted by Brockles at 8:51 AM on November 4, 2009

The purpose of the O2 sensor is to give your engine computer feedback on whether the fuel / air mixture post-combustion is what it expected, so that it can adjust the combustion parameters (spark timing) to meet its goals.

When an O2 sensor is faulty or missing, the computer lacks correct feedback, and may go into a mode where it continues to run, but does so less efficiently.

There is no way I know of to fix an O2 sensor. The only cure is replacement.

If your check-engine light says the O2 sensor is bad, it's either the sensor itself that's bad, or the sensor's connection to the engine computer. Probably the former, but it doesn't hurt to check the latter while you're in there.

If the mechanic really believes the O2 sensor is just there to monitor the catalytic converter, and wasn't saying that as a gross simplification of "the O2 sensor monitors exhaust," then yes, get another mechanic.
posted by zippy at 9:32 AM on November 4, 2009

That seems far more logical than just trusting the OBD system, to me.

I imagine that in 99% of cases, it's probably sufficiently effective and you gain quite a bit of efficiency. In my state, the non-OBD line takes forever, but with the computer check you're in and out in about 5 minutes. So, occasionally a non-compliant vehicle might slip through, but they can check a lot more vehicles for a lot less money.
posted by electroboy at 9:46 AM on November 4, 2009

It could just be that your car has learnt to compensate for the issues that caused you to need the parts replaced. Now that they are working properly, the engine (which is controlled by a computer) is still using its old information.

Give it a while and it will 'learn' where the new optimum idling point is.
posted by Nufkin at 9:50 AM on November 4, 2009

It could have been the guy was dumbing it down for me to the point of error. Anyway the dealer seemed more interested in the potential moneys heralded by the check engine light than in finding out why it was nearly stalling out. So we took her elsewhere.

Elsewhere replaced the O2 sensor and stuff, and so far, problem solved.

Ontario more sensible than my native Georgia??? Say it ain't so. /sarcasm
posted by rahnefan at 4:17 PM on November 4, 2009

An OBD II car will have at least two Oxygen Sensors. One before the catalyst and one after. The upstream sensor sends a voltage signal to the ECM relative to the amount of oxygen in the exhaust. As the cylinders fire and as the ECM tries to regulate the fuel mixture, the voltage will vary constantly and over a (relatively) wide range. The ECM uses that signal to evaluate whether or not it's leaning or enriching the air/fuel mixture appropriately.
The second sensor is placed downstream of the catalyst. It works more or less the same way and the ECM compares that voltage signal to the upstream sensor. If the catalyst is operating correctly, the signals should be substantially different. The downstream sensor should show a more stable voltage that reflects a higher amount of oxygen. The more the upstream and downstream signals look alike, the less efficient the catalyst is. The ECM will then store "P0420 Catalyst Below Efficiency Bank 1" and the Check Engine Light will come on.
The purpose of the downstream O2 sensor is EXACTLY to detect problems with the catalyst.

Now, if your catalyst has a physical crack, it can cause a fault code to store due to its inability to oxidize or reduce the exhaust gasses. It can also allow extra air into the exhaust and simulate an erroneous lean condition. That is, the O2 sensor will detect the extra air and the ECM will mistakenly try to add extra fuel.
I've seen a few Honda's that have developed a crack in the exhaust manifold, right above the catalyst. That's caused problems similar to what you're describing. It also can make a loud noise when the car's running. I'd be willing to be that's what's happened to your car since the catalytic converter is surrounded by a heat shield, making it very difficult to determine if the body of the converter is actually cracked.

Both a damaged/clogged catalyst and an incorrect fuel mixture can cause rough idle and poor performance.
In many cases, a complete failure of the upstream O2 sensor will cause the ECM to run based on predetermined program or substitute values for the O2 sensor data instead of operating in its usual adaptive way based on O2 sensor input. The unusual idle could be a result of that, as well.

What would be really helpful is if you could find out exactly what faults are causing your Honda's check engine light to come on.

FYI, in PA, there's no tailpipe emissions test for cars 1996 and newer. It's a strictly OBDII test.
posted by Jon-o at 6:45 PM on November 4, 2009 [1 favorite]

Ah, Jon-o is correct about the second O2 sensor. If this is the one causing the check engine light, then your mechanic's explanation, while a simplification, may be correct.

Also, your car may pass a sniffer test with a faulty catalytic converter, but this does not mean that your car is running as cleanly as it would be with a working converter. In other words, yes, the catalytic converter does effect your car's emissions.
posted by zippy at 8:42 PM on November 5, 2009