O2 Sensor Help

[teckersley]

Where is the tune file that was in the PCM when that log was made? The most recent one you posted is the one where every single DTC for everything was turned on.

It runs like crap when you command CL because it's adding more and more fuel trying to get a response from the O2 sensor and it never changes so it doesn't know to do anything but add even more. O2s are still not working - 'working' means that they go above 450mv when rich and below 450mv when lean.

You have a harness or PCM problem, assuming the new sensors are actually good.

Fuel System Status is showing 'OL - Fault' but there are no DTCs set that should be there when that happens.

Unplug both sensors. Key on. In the scanner, do B1S1/B2S1 show something like 450mv?

Do you have good +12v at pin D (heater circuit) for both connectors?

Good ground at pin C for both connectors?

If I used the phrase 'use a test light as a jumper' would you know what that means?

Yes. I have a power probe I can apply power and ground to circuits. Any particular tests?

The heater circuit does work. They do get hot

[blindsquirrel]

No, not a Power Probe. Old fashioned test light with a old fashioned filament bulb inside, to act as a resistor.

Monitor O2 voltage in scanner. With no sensor plugged in voltage should sit around 450mv. Apply ground thru the test light to drive the signal low, then apply power thru the test light to drive the signal high (should be capped at something like 1.2 volts, that's all the PCM circuit will measure/display).

Some people get really scared about applying full 12 volts to the O2 circuit, and recommend using a lower voltage source like a 5vref line from one of the other sensor connectors. I have personally never seen any damage from using 12v but use your own judgement.

If both O2 circuits sit at 450mv when unplugged, and GND makes it go low, and voltage makes it go high, then you have bad sensors.

Don't skip checking the heater +/GND circuits.

[blindsquirrel]

I have also seen folks use a bare finger to act as the jumper - touch a pin in the +12v heater circuit or ground and then touch a pin in the signal line. Skin will still pass enough voltage to drive the signal high/low.

[blindsquirrel]

Note: using the old fashioned test light method, the bulb will not light up. Or, it shouldn't.

If you don't see the 450mv bias voltage with the sensors unplugged do not bother with the other steps. If it's less than 450/near zero with sensors unplugged you have a short to GND. We already know it's not a short to voltage, as your O2 readings are so low with everything plugged in and running.

[teckersley]

Note: using the old fashioned test light method, the bulb will not light up. Or, it shouldn't.

If you don't see the 450mv bias voltage with the sensors unplugged do not bother with the other steps. If it's less than 450/near zero with sensors unplugged you have a short to GND. We already know it's not a short to voltage, as your O2 readings are so low with everything plugged in and running.

Thanks again. Let me do some pocking around.

I really doubt it's the sensors themselves. both sets of sensors were brand new, Both sets were genuine GM, and bot sets yielded the same issue.

[kingtal0n]

I have also seen folks use a bare finger to act as the jumper - touch a pin in the +12v heater circuit or ground and then touch a pin in the signal line. Skin will still pass enough voltage to drive the signal high/low.

The narrowband sensors doesn't actually need voltage, just a ground. The Oxygen sensor makes its own signal voltage output.

There are 1-wire oxygen sensors, for example. They ground through the exhaust.
There are 2-wire oxygen sensors, the second wire is a redundant ground.
The 3-wire sensors have a 12v heater circuit for quick warmup. The heater is optional; the sensor will still ground through exhaust and produce a signal without the 12v connected.

There are also 4-wire narrowbands. The 2 wires for heater, 1 is signal, and 1 is redundant ground (it is a 2-wire and 3-wire sensor combined).

All of that said. I've tuned a hundred cars and never used narrowband sensors. The air fuel ratio 14.7:1 is neither ideal nor optimal. It is best for emissions because leaner A/F ratios tend to produce NOx emissions, however the engine will produce superior fuel economy and maintain cleaner spark plugs and fewer carbon deposits when air fuel ratio of 14.8 to 15.5:1 is used, instead of narrowband target.

[teckersley]

The narrowband sensors doesn't actually need voltage, just a ground. The Oxygen sensor makes its own signal voltage output.

There are 1-wire oxygen sensors, for example. They ground through the exhaust.
There are 2-wire oxygen sensors, the second wire is a redundant ground.
The 3-wire sensors have a 12v heater circuit for quick warmup. The heater is optional; the sensor will still ground through exhaust and produce a signal without the 12v connected.

There are also 4-wire narrowbands. The 2 wires for heater, 1 is signal, and 1 is redundant ground (it is a 2-wire and 3-wire sensor combined).

All of that said. I've tuned a hundred cars and never used narrowband sensors. The air fuel ratio 14.7:1 is neither ideal nor optimal. It is best for emissions because leaner A/F ratios tend to produce NOx emissions, however the engine will produce superior fuel economy and maintain cleaner spark plugs and fewer carbon deposits when air fuel ratio of 14.8 to 15.5:1 is used, instead of narrowband target.

Interesting?.

I do have a wideband. I just have not used it to date. My ocd has gotten the best of me trying to figure out WTH is going on with the O2 readings. I checked the harness side O2 connector for grounds and voltage at the pin connections. I have 1 with 12.5v, one is registering as a ground BUT one has 4.3v and the 4th has 4.6v.

I understand your point about the wideband but I would really
Like to figure this out.

[blindsquirrel]

This is me being nice and not responding to typically irrelevant garbage. See?

[teckersley]

This is me being nice and not responding to typically irrelevant garbage. See?

Key on

B1S1 .448v
B2S1 .443v

Using the power probe, I applied ground and power to each of pins A and B with no material change to the voltages on B1S1 or B2S1.



Ugh this is frustrating to no end.

[blindsquirrel]

Here is your diagram again. Use it.

screenshot.28-11-2021 12.26.33.png

Here's a simplified diagram with only the things that matter right now. No sensors. Voltmeter. Scanner.

screenshot.28-11-2021 12.26.33-.png

Key on. What does it show in the scanner for O2 voltages?

What does the voltmeter show at both 'B' pins?

Do those two numbers match?

If you apply ground to either pin B, does the scanner voltage for the corresponding sensor go down to near zero?

If you apply voltage to either pin B, does the scanner voltage go up to ~1.2 volts?

(none of the circuits here should have anything even close to 4.5 volts - if that's really what is present with the sensors unplugged, STOP, and verify each wire is in the right spot on both ends)

[blindsquirrel]

Key on

B1S1 .448v
B2S1 .443v

Ugh this is frustrating to no end.

Yeah, now try doing it on flat rate.

Those voltages are what the scanner shows as well?

(the unfriendly post was for that other guy, not you.)

[kingtal0n]

(the unfriendly post was for that other guy, not you.)

He didn't know how an oxygen sensor works so I explained it. But he hates learning and likes to think he already knows everything so it pisses him off to find out the truth I guess. This is me being nice.


Narrowbands are good for correcting your tune when the engine isn't perfectly tuned. For factory setups which may run 20 or 30 years who knows, everything is going to wander, injectors clog and wires get worse and sensors/signals wander. Combustion quality changes, fuel content (ethanol or whatever) can change. Things wander around. So the narrowbands keep the tune in check.

But you as the tuner, as a person who is intelligent and capable of reading a wideband and understanding the performance and danger of wandering signals beyond the basic functions of idle and cruise air fuel ratio.... like if the fuel quality changes or an injector clogs... The narrowband can adjust average air fuel ratio but unless you have a narrowband in every single cylinder runner it will just throw the entire tune off. And if fuel quality changes you would need to reflect it in more than just air fuel ratio. So the narrowband winds up being a false sense of security for novice tuners who think that they are important for some reason. Say, why DO You think the narrowband is important, anyways? who told you that it was? Where did you get the idea you needed to use them? I am just curious.

Tune the engine properly and pay attention to deliberate deviations to performance such as clogged injectors and wandering sensors. Because they affect WOT fueling and the narrowband has no domain for WOT function. Tune the engine properly for superior economy and avoid depending on narrowbands is ideal for performance. Modern stand-alone computer use wideband closed loop... there is no narrowband anymore. Its an archaic, unwanted feature in performance applications, the majority of. Modern Haltech, AEM, Megasquirt, Holley, etc... are performing closed loop functions with no narrowband sensors. Nobody would want or use them in performance application. Except of course people who are just 'used to using them' like our blind friend here. Hes spent so much time learning to deal with and tuning around them... it pains him to admit they are unnecessary and unwanted because it would make all the time and energy he spent learning how to tune with them pointless waste of time. That is the anger source I guess. Truth hurts sometimes


I want them to work for you, just for learning experience. Narrowband are very simple, just use 1 signal wire output to connect the output signal to the correct ECU input wire and enable them and thats it. That goes for every factory ECU in the world, they all work the same. All narrowbands from all vehicles do the same thing, whether 1 2 3 4 wires, there is only 1signal wire to worry about.

TO test if they work put a volt meter on the signal wire against the ground the ECU uses (or chassis ground) and use the scanner to force the a/f ratio to go richer than 14.7 and then leaner, watch the wideband if you need to. The signal will be near 1.00v (like 0.956v or something) when it is rich and closer to 0v (say 0.054v) when leaner. This is a 5 minute test to make sure the sensor is working. From there it's just the wire into the ECU. You dont need to connect the 12v and ground wires for the heater But the sensor may not work until it warms up in the exhaust for a few minutes.