Request for diagnostic advice to pinpoint least expensive solution to resolve persistent P0421 OBDII DTC

Okay. You're getting an error which is saying that the ratio between the forward O2 sensor and the rear O2 sensor is not what it should be.

And you are -not- getting an error that is saying that the forward O2 sensor is out of range, which is also important. The fact that you are

-not- getting that error tells you that the exhaust coming out of the engine isn't THAT far off.

Now, it could still be off. It's still worth doing all the basic sanity checks: change the plugs, make sure the cables are good and there are no signs of arcing, make sure fuel pressure is good and that the exhaust smells okay. The idle RPM should be right on the mark and should not be drifting.

If you're spitting unburnt fuel down the tailpipe, the converter is going to be unhappy.

Now, look at the data for the two O2 sensors and look at what happens when the engine starts up. The first sensor should climb up into the normal range once it warms up. The second sensor isn't going to get into the normal range until the converter warms up. The shape of the time-domain curves will be different from one model to another, but just watch. The second O2 sensor might be bad... the converter might be bad..... and it's -possible- that the converter is just overwhelmed with junk (although that it is least likely of the three since you aren't seeing any other errors). But watch the numbers.

And you want to look at it for as long as you can... because you can have a loose O2 sensor cable, or a bad engine ground, and see weird spikes on those graphs that occur very very seldom but which can lead the computer to set an error.

If this engine has two banks, you can compare the two and that is very very useful information. Since you don't give the engine or model it's hard to tell if that will help you or not though.

--scott

This is easy to diagnose. You just sell that POS and get a newer one. You know better, Arlen dear. Dumb kid won't graduate anyway. That is why I married you.

Does your scanner give real time data?

Most scanners, all cheaper ones, only give the scan code, which you have.

Look it up online, see what it means, and see what remedies are suggested for it.

That will probably go faster than using the scanner even if it does give real-time data.

Diagnostic trouble code (DTC) P0421 stands for “Warm Up Catalyst Efficiency Below Threshold (Bank 1).” It means that the oxygen sensor downstream of your vehicle's catalytic converter on bank 1 is detecting that the converter is not working as efficiently as it should.

And the cat is expensivie to replace, especially for nothing.

Here are some of the common triggers of the error code P0421:

Faulty Bank 1 catalytic converter Faulty oxygen sensor Issues with the oxygen sensor circuit, such as damaged wires or poor connections Exhaust leak An issue with the PCM, such as software in need of an update

Hi Scott, I appreciate your advice as I'm not quite sure what readings to look at. o Let me know what information you need to help me figure out what to do.

OK. That's a good start as I was unaware that was what it tells us. o The main problem is what should that front-to-back ratio be? And under what conditions?

It's a 3-month old forward sensor. o I have no idea about the age of the rear sensor though.

In the daylight, if it's dry outside, I will ask to look under the vehicle. o There is only one bank (AFAIK).

From what I've gathered, I agree. It goes hours, sometimes days between the registers being set and the DTC. However I watched the dash light go on the instant the CAT register set. So it's "very" close, I think.

Note there is no evidence whatsoever of a misfire. Nor of rough idle. No smoke. No smells. But it's still good to replace the plugs (but that's throwing parts at it).

Yup. The goal, always, is to not destroy the cat.

OK. Thanks. I'll look for that first. a. Start cold (overnight) b. Start & idle & watch the two O2 sensors (pre cat and post cat) c. I'm looking for the pre-cat to get into range sooner than the post cat

Are these the two voltage readings for the two oxygen sensors? o O2SLOC B1S12--B2S---- o O2B1S1(V) 0.120 o O2B1S2(V) 0.140

When do I take the readings once the engine is warmed up?

Here's a reading, for example, taken at a steady 50mph: o O2SLOC B1S12--B2S---- o O2B1S1(V) 0.400 o O2B1S2(V) 0.160

I don't have the necessary equipment to save or see graphs. o I can only see the live data as it happens.

Here is a live shot of the data at a steady 60 mph: o O2B1S1(V) 0.120 o O2B1S2(V) 0.180

And another stead 60 mph reading: o O2B1S1(V) 0.740 o O2B1S2(V) 0.640

I'm not sure the speed for this one: o O2B1S1(V) 0.460 o O2B1S2(V) 0.120

Here's another reading, at a steady 20 mph: o O2B1S1(V) 0.700 o O2B1S2(V) 0.400

And another at an unrecorded speed: o O2B1S1(V) 0.820 o O2B1S2(V) 0.600

And another at an unrecorded speed: o O2B1S1(V) 0.500 o O2B1S2(V) 0.800

Here's one taken at 50 mph: o O2B1S1(V) 0.140 o O2B1S2(V) 0.760

Another, albeit in a sharp turn, at 50 mph: o O2B1S1(V) 0.800 o O2B1S2(V) 0.780 <ttps://i.postimg.cc/nz0VqjBs/obd20.jpg>

Another, on the highway but at an unrecorded speed: o O2B1S1(V) 0.780 o O2B1S2(V) 0.140

Taken shortly thereafter at 55 mph: o O2B1S1(V) 0.820 o O2B1S2(V) 0.700

Another on the highway at an unrecorded speed (2500 RPM though): o O2B1S1(V) 0.140 o O2B1S2(V) 0.220

Another on the highway at 60 mph: o O2B1S1(V) 0.400 o O2B1S2(V) 0.160

Another on the highway at 60 mph moments later: o O2B1S1(V) 0.620 o O2B1S2(V) 0.300

Another on the highway at 60 mph: o O2B1S1(V) 0.120 o O2B1S2(V) 0.180

Another on the highway at 60 mph: o O2B1S1(V) 0.440 o O2B1S2(V) 0.240

This one at 1,000 rpm (probably idling): o O2B1S1(V) 0.220 o O2B1S2(V) 0.840

Still at 1000 rpm shortly thereafter: o O2B1S1(V) 0.060 o O2B1S2(V) 0.760

Another at 1000 rpm: o O2B1S1(V) 0.600 o O2B1S2(V) 0.480

Yet another at 1000 rpm: o O2B1S1(V) 0.240 o O2B1S2(V) 0.140

Here's one at 10mph and 2300 rpm: o O2B1S1(V) 0.820 o O2B1S2(V) 0.780

I'm not sure if it's normal for these readings. o I do wish I had a graph - but I only seem to have live data.

Only one bank, thank God. :)

BTW, I just checked. It's a 2004 Mitsubishi Lancer ES, 2.0liter SOHC 4G94 engine

Given that you're in a State that mandates emission inspections...

Step 1: use your scanner to _clear_ the error code. If you're very lucky, it won't show back up.

Alternatively, if it shows up again every 500 miles, just live with it [a].

If it comes right back after 30 nominal miles (that is, once the system decides to do the rechecks), then you've got to figure out what's wrong

In my case, with a different car and a slighrly different OBD code, it turned out (we discovered after _lots_ of misdirection) that the EGR was pouring plenty of garbage into the exhaust, coating the catalytic converter, and messing things up.

Once we replaced the EGR this went from getting the error light/message 20 miles after reset to.. the aforemention very roughly every 500 miles.

Presumption was/is that there's caked on gunk which is interfering with the converter, but... it's not enough to set it off right away.

And, alas in my case, no easy way to pull the converter and clean it. (That's an option with some other cars).

For good measure, which might or might not have made a difference but which made me feel better, I

good luck

For the record, here's a description of what the computer checks for:

1. The computer looks to see if the heated oxygen sensor voltage rises above 600mV and falls below 300mV and switches in less than 100ms.

1. When the cat is working properly there is almost no switching action from the rear oxygen sensor.

2. To test the rear oxygen sensor, the computer forces a rich or lean mixture that the cat can't compensate for, and then monitors the oxygen sensor voltage. a. Rich to lean threshold voltage b. Lean to rich threshold voltage c. Low sensor voltage for switch time calculation d. High sensor voltage for switch time calculation e. Rich to lean sensor switch time f. Lean to rich sensor switch time g. Minimum sensor voltage for test cycle h. Maximum sensor voltage for test cycle i. Time between sensor transitions

According to this discussion on the oxygen sensor reading:

1. The post-cat oxygen sensor should be more steady at around 500mV since it's reading "cleaned up" exhaust.

1. However, if the pre-cat sensor is receiving too rich a fuel mixture, it will read a fairly steady 1.0V while too lean will read near 0V.

This explains the pressure waves:

1. The oxygen sensor monitors the content of the pressure waves as they pass the sensor. With each pressure wave the sensor sends a rich or lean mixture voltage to the engine controller.

1. At temperature, the sensor changes resistance as the air/fuel ratio changes. But instead of a gradual change, it switches very quickly from low resistance (less than 10 K ohms when the mixture is rich), to more than 20 K (ohms when the mixture is lean).

2. The engine controller supplies a base reference voltage of one volt to the sensor.

1. Lambda1 is the reference voltage of 450 mV, which represents an air-fuel ratio of 14.7:1 or stoichiometric.

2. The sensor is capable of generating a one-volt signal when there is no oxygen in the exhaust stream.

Oh Arlen the cat does not work It just sits on my lap all day. But you is so smart anyway.

This is a pretty good video (IMHO), on how the rear sensor works: o

No wonder my sensor readings jump around; it's a narrow band sensor! o Rich will read around 900mV; lean will read around 100mV

And it's tested for seven things a. Rich to lean switch time b. Lean to rich switch time c. Frequency of sensor transitions d. Rich to lean sensor threshold voltage e. Lean to rich sensor threshold voltage f. Low sensor voltage for switch time calculation g. High sensor voltage for switch time calculation

Under acceleration, the mixture will be rich (reading around 900mV). o When we slow down, the mixture will be lean (reading around 100mV).

While idling or coasting, the ECU will constantly change the air:fuel ratio between slightly rich & slightly lean to analyze how the rear sensor responds.

The rear sensor shouldn't switch frequently even if the front sensor does, which indicates healthy high oxygen capacity of the catalytic converter.

Low switching speed is greater than 24 seconds, whereas high switching speed is less than 1 second (which would be indicating that the catalytic converter has a low oxygen storage capacity).

The oxygen sensor needs to be at around 315 degree Celsius to operate correctly, so during a cold start the rear sensor will be at around 300mV flat, but when it heats up (after about 15 to 20 seconds), it will start to switch between lean and rich.

While driving at a constant speed (constant load), the rear oxygen sensor should switch slowly (which is a sign of a healthy catalytic converter with high oxygen capacity).

When coasting, the rear sensor should drop down to a steady 100mV; when accelerating it should go rich at around 900mV.

If you stab the gas pedal four times, you should see four spikes from lean (100mV) to rich (900mV) of the rear oxygen sensor voltage.

If you have access to a laser temp device or can borrow one... they seem to be everywhere nowadays... if the cat is doing its catty thing then the outlet temp will be a lot higher then the inlet temp. Run it for a

If it were my car I would also take the O2 sensors off and bench test them.

Oh Arlen, you get me so hot when you talk about stabbing four times but you cum in two.

Read the message over again. What you want to know is in the graphs for the two O2 sensors and how they track.

Don't worry so much about the ratio itself, worry about how they track.

But you will not figure the problem out just by looking at the plots, you will figure the problem out by thinking about how the emission control system works.

All the time. You take them before it warms up, you take them after it warms up, you take them over the course of driving around. What you are looking for is trends and correlation more than specific values.

Well, that's useless. So you're going to have to plot it on semilog paper by hand. Your scanner does not provide ANY plotting functions?

--scott

And the vehicle is long past the age where it should need a new converter.

But that doesn't mean you shouldn't make sure there is no possibility of something else damaging the converter. Because you don't want to swap out the converter and find the problem is fixed but returns.

--scott

Hi Scott,

I'm learning a lot, as I didn't even know the voltage range of the lambda sensors, nor what low versus high voltage meant when I opened this thread.

I appreciate that, so far, you're the only one who posted who knows anything, which I very much admire and hope to learn what you can impart.

As you know, you can't diagnose without understanding, so I've been boning up on how the system works, where I have the same attitude as you do, which is it could be any number of things, but replacing them without first confirming for sure they're definitely bad, is idiotic.

A _lot_ of people fix a lot of things that way; but it's still idiotic. o I see now that a _graph_ is almost essential, for proper diagnosis.

Given the moving nature of a graph, I learned from this video which shows, in real time, what the voltage graph should look like for the rear sensor: o Understanding & Live Graphing of the Rear O2 Sensor #19, by OBD4Everyone

Given I don't (yet) have a way to graph the results, this video shows how to diagnose using just the live data, which is the redneck way to do it: o P0420 How To Diagnose A Bad Catalytic Converter, by EricTheCarGuy

Speaking of redneck diagnostics, this video shows how a guy taps into the wires, which I think is too invasive given wire coatings, stainless steel wires, etc., but what's interesting is _how_ he forces a "lean" reading and a "rich" reading out of the oxygen sensor to look at fuel trim reactions: o Test O2 Sensor.mov, by realfixesrealfast

Literally, he grounds the sensor to simulate lean conditions (to watch the fuel trim respond by adding fuel); and he puts himself as the voltage divider resistor to force the oxygen sensor to output rich conditions.

While his approach (e.g., feeding propane into a vacuum port to simulate rich conditions) is a bit too much for me, we can learn a lot from redneck methods because they distil the problem down to the essential components.

In summary, when I opened this thread, I had no idea what to look for; I still am not sure; but I'm learning to associate low voltage (about ~100mV to about ~300mV) as a lean condition reading - & higher voltage (about ~700mV to ~900mV) as a rich condition - neither of which did I know prior.

I'm also slowly starting to understand that the cat acts sort of like a capacitor for oxygen, where the fluctuations of the leading sensor are damped in the readings of the trailing sensor, such that the rear lambda sensor should be steadier at around a stoichiometric ~500mV even as the front lambda sensor wildly fluctuates between lean & rich.

If that cat capacitor isn't functioning well, then the voltage fluctuations in the leading sensor will be more closely mirrored in the trailing sensor.

A key distinction I need to wrap my head around though, which you seem to have caught upon in the very beginning, is that this is a "warmup" error: o What Does P0421 Mean?

This makes the problem harder to diagnose since, as you seem to have realized instantly, it will only show up during the warmup process.

Worse, it happens only "most" of the time, so it's intermittent (i.e., it can go seconds, hours, minutes, or even days between showing up - but it eventually will show up and keep the CEL lit, interestingly, forever).

That's right Arlen, those other people don't know anything and just waste your time. Only one person knows and gives you a hard on. Just rub it now like you usually do

That is the factory method for testing, not "redneck" as you seem to think. As for being invasive, not even close as long as you seal the holes you make or use a backprobe method.

In simple terms it means that the converter isn't starting to operate as fast as the table in the ECM thinks it should given the engines run time, fuel and air states. The ECM looks at the time the engine started, it's loading, coolant temp, air temp, rpm timing, injector flow, air flow and a couple other sensors and using the tables programmed in it has decided the converter isn't starting to do it's job fast enough based on the numbers they used in the programming.

Sometimes this can simply be a bar sensor, IE the rear sensor is sluggish due to contamination or age and the ECM thinks the cat is the issue when it's the sensor. That is the whole point of the graph, you want to see how long it takes on a cold start for the sensors to start working and what the voltage numbers are. The direct testing at the sensor is also the best testing because everything that comes through a scan tool is processed data, it has been known to be wrong multiple times, while direct testing shows the raw data.

Hi Steve W.,

Thanks for your advice, as cat "warmup" is the critical parameter for sure!

I agree that the P0421 indicates a catalytic "warmup" fault, where it must be close since I've seen three situations occur multiple times so far: o In most cases, the P0421 is set the instant readiness monitors are full o In other cases, the P0421 is set days after readiness monitors are full o Oddly, once set, it doesn't ever reset (e.g., after 3 good drive cycles)

Of course. o The difficult question to answer is what to test to debug during warmup.

The lambda sensor is heated, so it only takes ~20 seconds to warm up. o But what's expected warmup time for a cat (& how exactly is it measured?)

Yes. Indeed.

I certainly didn't realize this when I originally opened this thread, but I do realize that measuring how long it takes to get to steady state after an overnight rest & morning cold start seems to be a critical diagnostic.

Understood. I'm very familiar with flying leads. o I simply prefer not to do that to other people's cars if I don't have to.

What I need at the moment is to borrow the vehicle and see what I can determine in the first crucial minute. o I know it only takes about ~20 seconds for the sensors to warm up

But I'm not yet sure how to test the time it takes for the cat to warm up. o Especially when it appears to be a borderline condition on this vehicle.