Diagnosing a P0420 code
 

I keep getting the p0420 code that indicates either low catalyst efficiency or a faulty downstream o2 sensor. Obviously, the CAT is possibly the problem. But I suspect the CAT efficiency could be undermined by another problem, such as burned oil or unburned gas, so I want to find and fix any such issues. The car still drives fine, so I keep driving it, but I have a SMOG test due soon.

Symptoms:

The P0420 code has lit up the CEL 8-10 times but in one situation only: steady cruise, light throttle, freeway speed, after the car is pretty thoroughly warmed up (at least 8 or 10 miles of driving). It has not yet come on when doing any of my typical hypermiling, such as P&G, which involves high load and short runs.

Diagnosis (pictures!)

[I will edit this to upload pictures in a few minutes]

Blackened upstream O2 sensor:
http://ecomodder.com/forum/attachmen...1&d=1529895123
This sensor was cleaned throughly to near new appearance about 2500 miles ago.

Slightly plugged CAT honeycomb:
http://ecomodder.com/forum/attachmen...1&d=1529895126
The dark spots are probably plugged honeycomb, and across the entire surface it looked like this. I cleaned this CAT with the acid wash method in 2016 and the honeycomb was all uniform light grey and clear afterward.

Fuel fouled plug #4:
http://ecomodder.com/forum/attachmen...1&d=1529895126
All four plugs looked about the same, like this one, but this one (#4), was wet and smelled of gas, too. The plugs were all used but clean and gaped properly about 2500 miles ago.

Possible causes.

As steady cruise, I have also noticed the longterm fuel trim gets lean by a percent or three. So I have been wondering about a rich fuel condition. Obviously, the fuel fouled spark plugs suggest that.

I have wondered about a partially plugged up CAT, which obviously kinda seems to be the case in the photo.

Next steps

I already tried cleaning the O2 sensors. That solved nothing. So...

1) Test spark plugs and wires and replace the plugs if indicated.
2) Maybe wet compression test to look for oil blow-by.
3) Maybe injector test to look for leaking
4) Clean catalytic converter and O2 sensors.

YOUR THOUGHTS?

P0420 is almost always a dead cat. The description of your test conditions when the code sets are part of the drive cycle. That is when the cat efficiency test runs. What happens is the lean rich swings match on both sensors and this tells the pcm that the cat is not working. If the cat was working the two sensors would not have the same readings at the same time. I would try cleaning the cat that is between the two sensors.

What I might try is get it tested, if it passes by a lot it probably is the O2 sensor.
If not the code is correct and catalyst is below efficiency threshold, try to clean it.
Converters have been using a lot of palladium over the years, palladium is near a 15 year high, so you don't want to buy one if you can help it.

Buy a spark plug anti fouler and drill it out to 1/2 inch and install it on the downstream sensor and just be done with it. Worry about it next time you have to smog it.

Thanks for the reference to the testing cycle. It never occurred to me that the testing might be intermittent and not constant. I have been thinking it is constant and that it sets a CEL when a dip below say 80 or 85% efficiency is detected. But this website says the threshold is 50%. So according to the website, my car sometimes detects below 50% efficiency across two drive cycles and other times does not for weeks. What could cause intermittent efficiency drops like that? Maybe the gasoline of spark plug #4 is a hint?

I would rather identify the problem in advance, not least because it appears there are other problems besides the CAT. What do you guys make of the gasoline on spark plug #4? Seems the spark is not firing completely sometimes, which might mean the gasoline/air mix is being sent straight into the CAT. If that happens when the ECU is running a test of the CAT and boosting the amount of fuel in the mix, then a gasoline/air mix burning in the CAT could possibly send O2 across the downstream sensor sometimes, no? That might trick the ECU into thinking the CAT has failed when actually spark is weak. No? Look at my picture of the CAT surface in my car. That's pretty clean: no melting, no cracking, no black spots, no flaky powder. Just some off-coloration.

As I mentioned in the OP, I have a SMOG test soon. Gotta get it done in July.

Fuel trim reading 1-3% lean wouldn't raise my eyebrow at all. Have you looked at the spark plug again since driving it more? Wet and gasoline smell could well be because the engine was shut off on the intake stroke.

P0420, in my experience, rarely lies. I'm sorry I forget what instrumentation you have, if you have an Ultragauge I know you can watch O2 sensor voltages. The data resolution sucks but should be passable for seeing if your downstream sensor is switching or holding fairly steady. Other instruments may have this function as well, I just don't know.

Do check the spark, just to be on the safe side, but if you had a serious enough spark issue to burn up the cat, I'd think you would see it on the fuel trims, and under 3% correction for your long terms is great; the rule of thumb I've heard is that more than a 10% correction is when you need to start worrying.

An interesting thing then is that short term trims will read very lean, sometimes way over -10% and often under high steady load (climbing a 1.5 mile steep hill on my commute) I have seen short term trims of -26%. My long term trim might be deceptive because P&G has always produced modestly positive longterm trim, such as +0.78 or +1.56. So what happens on the freeway now that I have taller tires and am steady cruising at light throttle is the longterm trim gradually moves into negative territory. If I had a longer ride it might get -10% or higher eventually. That make sense? If true, my car has been possibly feeding unburned gas into the CAT for a VERY long time.

I resistance tested my ignition wires just now on a lunch break: the #4 with the unburned gas on the plug has oddly low resistance in the wire. I got a bunch of reading near zero on the 20k scale. See attached notes:

http://ecomodder.com/forum/attachmen...1&d=1529954275

Number 4 wire is faulty, I gotta think.

[EDIT: AND I TOTALLY WILL CHECK THE OBD2 DATA OUTPUT FOR THE DOWNSTREAM SENSOR. THX.]

If you have a misfire or leaky injector that could clog up the converter if ran long enough.
If you fix the mission fire and give the car an Italian tune up it may go away on its own.

If one spark plug wires is not like the others, replace it.
Normally they develop really high resistance, not low.

Hmm ok, strange the long terms aren't adjusting (they usually follow short terms fairly quickly) but with short term that far out, then yes I would definitely want to check out fuel and spark. If you have an INCANDESCENT test light (say NO to LED for this test), ground the claw and hold the other side out from the boot while cranking or running. It's a quick and dirty test but you want probably at least a 1/2" of spark (preferably 3/4") reaching out to your test light from the bottom of wherever the conductor is buried in that boot, and with the resistance of the incandescent bulb you both put extra resistance in the circuit and avoid damage to the ignition and/or any computer drivers.

Low resistance is strange, but remember that you're checking that wire unloaded, and unloaded circuits are strange. Put said incandescent test light in-line with your multimeter and see what the resistance does, compare to the other wires.

Can you see both O2 sensor patterns at the same time? When they match it trips the P0420 code. There needs to be a lag in the second pattern for the code to go away. This is best seen on a graphing scanner like a Modis or a laptop with the correct software. We have mostly Fords at work and we have the same computer programs the dealers have. This has been great being able to flash computers and diagnose drivability issues.

Well do the anti fouler trick and see if it will pass smog? The cat may be on its way out, but it might be good enough to pass. You could also play with the timing a little to lower emissions to help it.

So I decided that the resistance test on the ignition wires were bad enough to justify replacing the wires. By Ohms law, low resistance will mean low voltage, which might mean weaker spark and incomplete burn, even if there is no misfire. Over a long period, that could pollute the CAT with unburned fuel. The new ignition wires showed about 10,000 ohms per foot (25K is the Honda FSM service limit). Seems to have changed functionality only modestly.

I can now. I had to do a freeway errand tonight and forgot my scantool. On the way to, the P0420 code popped again. I set my Ultra Gauge to show both O2 sensors side-by-side before coming home. At idle the downstream mostly stayed at about 0.9v (1.25 max and rich, 0.1 min and lean). The upstream moved around pretty normally. It was on the freeway that the fun started. At steady cruise I caught the downstream sensor dip into the 0.1 voltage area once and a couple times into 0.2v. That means a lot of oxygen suddenly crossing the sensor. Clearly I have a CAT problem. But nearly all the time, the downstream stayed in the 0.8v or 0.7v range and did not seem to swing widely. So I think it is contamination and something else...

The short term fuel trim is going moderately to very negative, sometimes -25%+ but often only -7% or -9%. Longterm trim is not at -3.9%. SO I still think part of my problem might be unburned fuel getting into the CAT.

I will acid wash the CAT, but first I will test some other items looking for the cause of the rich condition.

1) Test spark plug itself.
2) Test the MAP.
3) Test the IACV.
4) Test the injectors for leaking(?)

I guess. But if I do stuff like that the underlying problems will still be there (a failing CAT and a persistent rich condition). I want to car to run well, clean, and durably. So I gotta diagnose it right, if I can.

Lunch break testing of the spark plugs.

The plugs in the car are shot. I did not realize how badly the contamination they show might errode function. With great difficulty getting any resistance readings, I eventualy managed to get these: at the 20k setting on my meter I got #1 @ 8.60, #2 8.65, #3 8.75, #4 9.17... but they all failed the ground test and the continuity test because of corrosion. They were always a variant of the stock plug anyway.

I have a lightly used set of the stock plugs on hand for smog testing. They passed continuity and ground tests and I got these resistance readings with relative ease when they were still a little dirty:
#1 8.75, #2 9.59, #3 9.80, #4 9.66

So, I cleaned them to shine and bare metal and gapped them to 1.0mm to 1.1mm.

Will see how they do on my errand with my kid tonight. I fully expect to get the P0420 again. But solving all this in the fuel/air systems seems important before I clean the CAT and see if I can regain acceptable emissions efficiencies.

Thanks for all the input guys. The freeway errand tonight lit the P0420 again on the way out. This time I was watching the upstream and downstream 02 data on my UltraGauge and when the light went off I was actively expecting it because the two readings began to move almost in unison for a few seconds. On the way home on the freeway, there was a prolonged period of 5 seconds, 10 seconds, or more when I could see the two readings moving almost in unison. I got lean readings from the downstream sensor as lean as 0.15 volts repeatedly.

I am gonna start a new thread on the persistent rich condition. Once I get a better understanding of the rich condition, I will clean the CAT with an acid wash.

Thx!

Surprise!
 

Got a nice surprise.

I don't know if it will last, but suddenly the rich condition was gone on my return freeway run last night.

To recap, I tested and replaced failing ignition wires and spark plugs but the very high STFT readings (as high as -29) continued at idle freeway steady cruise across two drive cycles. I did my freeway cruising using P&G last night (third drive cycle) on the way out to appointment and the LTFT, as expected, improved from -9.4 to -5.4 as I expected. That's normal LTFT behavior for P&G. On the way home I wanted to see the negative fuel trims again and observe, so I steady cruised it, but all STFT readings were normal.

It seems possible that the ECU has "learned" it faces normal conditions now. But I am gonna wait a few more drive cycles and use the sensors for diagnostics. Maybe an intermittent failure has gone away temporarily.

The downstream O2 sensor reading is still fluctuating a lot, though, so I still think I need to clean or replace the CAT.

Solved?
 

Tonight's two drive cycles at steady cruise freeway speeds makes three in a row without a hint of the rich condition, super lean LTFT. In fact LTFT is at 0.00 exactly now. Better yet, I have not seen the P0420 return, either. This exact drive has set that code off at least 6 or 8 times this month. I wonder if that will continue.

Not solved
 

On fifth drive cycle since fuel trim fix, the P0420 came back today. Diagnosis testing done. I have an inefficient CAT. I will clean it to see if that clears the code. I'll post those results in my CAT restoration thread.

At least this process taught me new testing techniques and revealed a simple cause for the bad fuel trims.