Repair-It

Could I get this lucky?

https://www.toyota-4runner.org/3rd-gen- ... -%245.html

Totally passive approach.

https://www.instructables.com/P0420-Oxy ... ator-Hack/

I could also buy one of these: https://www.rockauto.com/en/moreinfo.ph ... 42&jsn=569
And one of these: https://www.rockauto.com/en/moreinfo.ph ... 01&jsn=401

Make my own splice harness. Basically wire up the resistor and cap into the harness, unplug O2, plug the splice into both ends. Fully reversible and if I had to tweak the discrete components I could unplug and do on bench (soldering, etc.). That would be about $48. Plus a couple bucks for some resistors and caps...

I fixed the P0138 high voltage by using 1M and a 4.7uF along with another 1M resistor across the 4.7uF cap (signal to ground right next to cap). I now have a nice smooth line at about .8V and a happy ECM

This other guy struggled and struggled and eventually ended up using a spark plug non-fouling bung to pull the rear O2 out of the exhaust stream. I had no luck removing mine last time and I'm not sure I'd even be able to install the adapter given the clearances. Maybe I need to give that another try as it might work if I could just get it installed...though an inspection might be interesting.

"What's that?"
"That's nothing"
"OK"

I got my sensor and harness plus resistor and capacitor set. Should be able to make a special harness that plugs in. This will be interesting. I hope this works.

Did more reading of those Instructables comments. Apparently a P0138 or P0139 is an issue that some people have.

P0139 (P0159 for bank 2 which is the bank this issue is in) sets if the engine goes to DFCO and the downstream O2 takes too long to register the rich/lean transition (presumably high to low voltage). And it sets if the transition is longer than 1 second. So that's cheeky. Seems like a way to detect the filter circuit that I'm installing here. The recommended components were 1M ohm and 4.7 uF. That gives a time constant of 4.7s which seems long. So I guess I'll do the 1M ohm and 1 uF for a 1 second time constant.

DTC P0139 or P0159
Circuit Description
The heated oxygen sensor (HO2S) produces a voltage that varies between 100-900 mV under normal operating conditions. The engine control module (ECM) produces a bias voltage on the HO2S signal circuit of 420-480 mV. The reference ground for the sensor is provided through the ECM. The ECM monitors the signal voltage to determine if the exhaust is lean or rich. The oxygen sensor voltage is high when the exhaust is rich, and low when the exhaust is lean. The ECM constantly monitors the HO2S signal during the Closed Loop operation. If the ECM detects that the decel fuel cut-off rich-to-lean transition time is too long, DTC P0139 will set for bank 1 sensor 2, or DTC P0159 will set for bank 2 sensor 2.

DTC Descriptors
This diagnostic procedure supports the following DTCs:

• DTC P0139 HO2S Slow Response Bank 1 Sensor 2

• DTC P0159 HO2S Slow Response Bank 2 Sensor 2

Conditions for Running the DTC
• DTCs P0021, P0024, P0030, P0031, P0032, P0036, P0037, P0038, P0050, P0051, P0052, P0056, P0057, P0058, P0101, P0102, P0103, P0106, P0107, P0108, P0116, P0117, P0118, P0121, P0122, P0123, P0125, P0128, P0131, P0132, P0135, P0137, P0138, P0139, P0140, P0141, P0151, P0152, P0155, P0157, P0158, P0159, P0160, P0161, P0201-P0208, P0221, P0222, P0223, P0261, P0262, P0264, P0265, P0267, P0268, P0270, P0271, P0273, P0274, P0276, P0277, P0279, P0280, P0282, P0283, P0300, P0301-P0308, P0335, P0336, P0340, P0341, P0345, P0346, P0351-P0358, P0365, P0366, P0390, P0391, P0442, P0443, P0446, P0449, P0453, P0454, P0455, P0458, P0459, and P0496 are not set.

• The ECT Sensor parameter is more than 66°C (151°F).

• The Vehicle Speed Sensor parameter is between 5-180 km/h (3-112 mph).

• The calculated catalytic converter temperature is more than 520°C (968°F).

• DTC P0139 or P0159 runs continuously once the above conditions are met.

Conditions for Setting the DTC
The ECM detects that the decel fuel cut-off and rich-to-lean transition time has exceeded 1 second.

The P0138 (P0158 for bank 2) is basically the opposite. Instead of taking too long to go low, this sets if the voltage stays too high for too long. And the time parameter is more than 2V for 1.6 seconds.

DTC P0138 or P0158
Circuit Description
The heated oxygen sensor (HO2S) produces a voltage that varies between 100-900 mV under normal operating conditions. The engine control module (ECM) produces a bias voltage on the HO2S signal circuit of 420-480 mV. The reference ground for the sensor is provided through the ECM. The ECM monitors the signal voltage to determine if the exhaust is lean or rich. The oxygen sensor voltage is high when the exhaust is rich, and low when the exhaust is lean. The ECM constantly monitors the HO2S signal during the Closed Loop operation. If the ECM detects an HO2S voltage that stays above a specified value, DTC P0138 will set for bank 1 sensor 2, or DTC P0158 will set for bank 2 sensor 2.

DTC Descriptors
This diagnostic procedure supports the following DTCs:

• DTC P0138 HO2S Circuit High Voltage Bank 1 Sensor 2

• DTC P0158 HO2S Circuit High Voltage Bank 2 Sensor 2

Conditions for Running the DTC
• DTCs P0030, P0031, P0032, P0036, P0037, P0038, P0050, P0051, P0052, P0056, P0057, P0058, P0131, P0132, P0133, P0135, P0137, P0139, P0140, P0141, P0151, P0152, P0153, P0155, P0157, P0159, P0160, P0161, P1137, P1138, P1157, P1158 P2231, P2234, P2237, P2240, P2243, P2247, P2251, P2254, P2626, P2629 are not set.

• The engine is running.

• DTC P0138 or P0158 runs continuously once the above conditions are met.

Conditions for Setting the DTC
The ECM detects that the HO2S parameter is more than 2 volts for longer than 1.6 seconds.

So I guess this "informs" me to keep the time constant to 1s? I'm kind of reaching here but it's worth a try for science.

Here is my circuit diagram. I wired the cap on the wrong side of the resistor the first time.
I did look up the wire positions so I modify the right wires, and I've labeled both ends of my pigtails. The other good news is I am using ceramic caps so they are not polarized for my convenience.
Still doesn't work. Getting P1158 and P0161 as pending codes. The FSM says they can not both set at the same time. I don't understand the heater code. The sensor code I get because I can visually tell voltage is not behaving per the criteria for setting the code (voltage doesn't drop to below 30mV during DFCO for 5 seconds). I can plainly see it goes to 0.390V and sits when it should be 0V. So that code checks out. I think that is because I am using a ceramic cap which is non-polarized. I think it's causing the issue so I need to redo with an electrolytic cap. Argh.

I triple and quadruple checked my wiring. I can't find any issues with the actual pinout setup I am using. Best I can figure is the voltage is getting screwed up because of the ceramic cap and SOMEHOW this also triggers the pending heater code (makes no sense at all though).

Alright a quick update. I stuck a 220 uF cap in there (measured around 250) without the resistor. Still got pending P0430s and honestly no visible change in the waveform. I thought maybe the impedance of the sensor and/or ECM would preclude the need for the resistor. Guess not.

So I stuck a 20k in series to get my RC time constant to around 5s per the Instructables post. Wow SUPER laggy response. You could see the long decay upon DFCO and it never even hit 0.000V as the sensor 1 does almost immediately. And it got a pending P0159 which was predictable. I did a series of resistance decreases and each time I did notice better behavior but still presented with a pending P0159. Here is the actual path that I went down: Ended up with this which you can see is a much shorter decay and a calculated 0.25s time constant. Still too laggy though.
So what I did was order a kit of assorted potentiometers because soldering resistors each time was getting annoying. So I'll throw a 1K potentiometer in there and I guess halve the RC time constant until something changes. Ideally, I am shooting for getting rid of the P0159 while still retaining enough lag (voltage storage is theoretically mimicking the catalyst oxygen storage that isn't working as it should) to make everything happy. There is NO guarantee that there is a combo here that will work because this is a sophisticated ECM with CA emissions. They probably threw the book at it. I think my realistic best bet is to add enough lag that I can squeak through the test but not permanently turn off the CEL because under some conditions my circuit may not cut it. I will continue I guess on Friday or something.

At least I am getting A result. Not the one I wanted yet.

I did an errand and sure enough it set the CEL. This means that pending is basically what you think it is. If you see a pending code, you probably didn't fix the issue...of course, you could have a pending code that comes and goes and if it happens to never hit the ignition cycle/test threshold for the CEL, you'll never have a CEL and be emissions compliant. Kind of living on the edge though and not recommended, or predictable. Unless it's predictable.

I added a resistor in series with the signal wire to go with the cap. This finally got a result. It changed my P0430 (same as before) into a P0159, slow responding O2. This make sense. So I kept reducing the resistance values to knock the RC constant down. I got down to a 1K resistor and it was still doing the same thing. So I decided to order a pot assortment because if the answer was below 1K, I'd rather turn a screw than keep desoldering and soldering.

So that was a couple days ago, and now I got my pot assortment. I picked up a 1K and measured it and it was basically in the middle so I threw it in because I was just using a 1K so this followed my progression of halving my resistor (and RC constant).

STS CEL-be-gone kit.
Anyway I started the car to drive on ramps and then shut it off to remove the jumper harness. I installed my pot at around 500 ohms, installed the harness and went for a drive. Only a few miles down the road and I got this as pending.
P1158 is what I got with the 1M ohm and 1 uF cap. Except unlike that time, the signal waveform did not appear to be meeting the criteria to set that code. In fact it hadn't since I had dropped the resistance (I think I started around 20K?. Basically it means it saw over 30mV during 5 seconds of DFCO. That was happening with big resistors because even in DFCO when the sensor should be 0.000V, it was stuck much higher than that. My interpretation is the large resistor was not allowing the O2 to "ground" the reference voltage from the ECM.

But anyway, I have no clue why this set as a pending code. The signal was dropping quicker than ever upon DFCO and took WAAY less than 5 seconds to drop. More like less than 1 second. So I simply reset it AGAIN, and kept driving. Here is my tally so far, since I began tracking:
More driving revealed.....no pending codes. In fact I just started it after it had been sitting most of the afternoon and took it on the same few mile drive that originally set the P1158 pending and I saw no issues there, or on the way back. In fact, all systems are ready except evap (which I could get ready if I had more fuel in the tank - I burned through an entire tank testing this I think). So I could theoretically bring the car in now and it would pass. Maybe I'll make my appt for May 1st. And maybe try not to drive the car in the meantime....

Anyway I'm cautiously optimistic here.

Brought it in. Had to buy a rear marker bulb but it passed.
I think the plan now is to just run it for a tank or so and see how it does. If it's still good I will make my wiring harness more permanent. Not sure if I will swap out the pot for a resistor or just tape it up and loom it.

And then what, sell it?

Picked Ian up. No pending codes. It would DEFINITELY have had at least a pending code from this drive.

Another day of driving later, still no pending codes. In fact it got even more compliant: evap became ready. Now every system is ready and no pending codes.

I have to say this is going better than I expected. I think the only other test would be a long highway drive.

Haven't been driving the car much but I have driven it. No CEL. I'll check for pending codes next time.