Bad O2 Sensor or Cat Problem?

Hi all! I have a 2004 Hyundai Elantra that is having problems getting the Cat test to go complete. It is stuck as incomplete and has a pending P0420 code. The code has been pending for the past two weeks, maybe longer as that’s when I first started checking because I wanted to get inspected. This is my first car so I’ve only owned it for a year, I have no idea if the O2 sensors have ever been replaced. I bought a “more fancy” scan tool than what I was borrowing from my dad (OBD Scan tool from the early 2000s to a cheap phone one since I’m going to college soon.) I used the new tool to graph the O2 sensor data and would like some opinions on it. Do you guys think it’s a lazy upstream O2 sensor or could it actually be the Cat? The O2 sensor and heater monitors both show ok but something just seems off. Thanks in advance!

If the upstream O2 sensor is lazy but still (sort of) working, how come the car is able to enter a closed loop? That’s my only concern with my hypothesis with the O2 sensor being bad.

Detailed Upstream
Detailed Downstream

Here’s what the O2 sensor signals should look like.


So what you’re saying is it could be both (Upstream O2 and cat going) or are you leaning towards just the cat starting to go?

Both lines on your graph are labeled downstream.

The computer enters the closed loop mode when coolant reaches operating temperature.

It doesn’t matter if the sensors are working or not.


Oh okay, that explains the fuel system loop question. I thought that O2S1 was upstream and O2S2 were downstream, since they’re both bank 1 if I’m reading it right.

You would have had some driveability issues like lacking power and bugging down if it was the cat.

So you’re leaning towards it being the O2 Sensor if I’m correct? Do you think it’s the O2S1 sensor like I was alluding to above? I don’t have any driveability issues, just the CAT monitor staying at incomplete with the code staying at pending, it won’t go away or set permanently. If I erase it it just comes back, same deal.

If the green line is the downstream O2 sensor, the cat is bad.



O2s1 is upstream

O2s2 is downstream

you’ve got way too much activity on the downstream sensor . . . that’s why you’re getting P0420 as a pending code

Conversely, there’s not much going on, as far as the upstream sensor goes. It’s skewed too high, and there’s not much fluctuation at all.

Before you condemn anything, make sure you don’t have any exhaust leaks

How many miles are on this car?

And was the exaust really hot when you took that snapshot?

I mean smoking hot, as in several miles of spirited driving

And by the way, there’s nothing in that snapshot proving the system actually entered closed loop

Unless you want to post a different snapshot which literally spells it out . . . that wouldn’t be in a graph form, though

You’ll also want to investigate the fuel trims. That might provide some clue as to why there’s almost nothing going on with the upstream sensor

1 Like

Don’t look like you had a hot enough engine when the graph was captured. Sensors don’t work or work correctly until the exhaust is at around 640 degrees.

According to this, I would change the downstream sensor. $30 if you do it yourself.

More ideas . . .

With everything warmed up . . . after those several miles of spirited driving I mentioned . . . create an extremely large vacuum leak at idle

How does the upstream sensor react?

It should very rapidly show a very low reading, such as 50 or 100 millivolts

If it can’t do that, the upstream sensor is shot

also google “snap throttle test” . . . this is a quick and dirty way to evaluate an oxygen sensor

Curious as to why you would change downstream sensor. It is active and reading. The readings indicate a bad cat, but the sensor is doing it’s job.

OP has no other symptoms than an error code. True, it could be a going bad cat, but there have been no tests performed to determine that. I.e. bad acceleration, funny exhaust smell, running hot, performance issues, etc.

That’s why I recommended to simply switch the sensor around. That can be one way to determine a bad sensor. A bad sensor can give unpredictable values and that is how people replace cats that are perfectly fine only to find out later that cause is something other.

The main symptom to the driver is the illumination of the MIL (malfunction indicator lamp). You will likely not notice any drivability problems, although there may be symptoms. For example, if the substance inside the catalytic converter is broken or failed, it may be restricting the exhaust which will result in a feeling of reduced power output from the vehicle.

Read more at: P0420 Code: Catalyst System Efficiency Below Threshold (Bank 1)


I think we both said the same @Tester and we don’t know if any of that is happening. With the limited info provided it make me crinch when I read suggestions that appear to be definite and final.

Did you look at the OP’s reading of the downstream O2 sensor?

Isn’t that enough proof that the cat is bad?

That’s how I diagnose if the cat is bad.


Hello everyone! The readings were in fact not from too hot of an exhaust. I’ll drive around tomorrow and take some readings at the very end at idle after driving around for a while. I like the idea of switching both O2 sensors around to see how they react. I’ll do that as soon as I have time, sometime this week. I’ll upload more graphs tomorrow after driving and getting data on the sensor when everything is very warm. The car has 108,000 miles. I got it last year with about 102.3k on it, I have no idea when or if the O2 sensors have ever been replaced, as that would’ve been before I bought it. I just want to re-iterate, the code is only pending, the MIL has never turned on for it. Just so everyone knows, it does light during startup, so the light does work. It’s stayed as pending for the last 2 or more weeks, I just notice because I need to get inspected. Thanks for all the info so far everyone!

The engine must have been hot enough when you took those readings.

If it wasn’t, the computer would be in the open loop mode and the computer wouldn’t be sending O2 sensor signals.

Be careful switching the sensors around.

The upstream sensor might not be an O2 sensor. But instead a wide band, A/F ratio sensor.


1 Like

Okay, I’ll make sure that both sensors are regular O2 sensors before I swap them around. The sensors were most likely up to temperature but the exhaust and coolant definitely wasn’t

Upstream and downstream oxygen sensors have different connectors to keep them from getting mixed up.