Correct call?

Today my boss has me checking out a CEL on a 2011 Mazda 3. Code was p0138 - rear o2 high voltage.

Data pid for the O2 showed it to be sluggish and looking like a dead sensor. However, the heater pid was not toggling back and forth from on to off. The front AF sensor heater was toggling, but not the rear.

I decided to scope the heaters on both and I found good 12v into both, but the ground on the rear o2 was not being pulsed.

This capture is orange representing the ground on the AF heater and green represents ground on rear o2 heater.



AF is probed sensor side, O2 is probed at ecu pin.

My only question here is why didn't I get a heater circuit code, instead of/in addition to the high signal code? Am I making the right call? That voltage spike resonating into the AF side of things tells me this car needs an ecu and an O2 to be safe.

I wouldn't call an ECU, just yet. If you are doing your back probing at the sensor connector, you still need to rule out the control wire. Move your back probe to the ECU connector. Do you still read battery voltage?

The voltage spike you are seeing is, most likely, Secondary Ignition feedback. The closer your test leads are to a coil/plug wire, the more you will see this.

Yeah but the 12 is dropping by about a volt and then spiking, each time. That looks like a failed driver as far as I understand ecm design.

And yes, that green trace is backprobing the ecm underhood.. The square wave is the healthy AF heater ground, backprobed sensor side just for comparison. The green O2 heater ground is the same spiky 12v on any point on both heater wires.

Deltron wrote: AF is probed sensor side, O2 is probed at ecu pin.

I see, now, you DID test at the ECU. I missed that. That makes me agree with you on the heater driver.

Ok good, I'm feeling better then. My boss was looking at me sideways making me doubt myself, but I'm pretty sure he's an idiot when it comes to efi so...

One thing I would double check is the possibility of the voltage you are seeing at the ECU being a bias voltage. An easy way to tell is to connect an incandescent test light to battery negative and touch the back probe. If it voltage disappears, you have more testing to do. If the test light glows, it's safe to say that it is NOT a bias voltage.

Is there some way the heater itself could cause that sort of trace? Or would I need to use an amp clamp to see heater health? I thought this voltage was the 12v into the heater that comes out the other side of the coil. If I use a test light to ground, I'll just turn the heater on. I've seen Paul do that in videos.

I, still, think that is Secondary Ignition feed back. It looks like a coil ramp and voltage spike. Does the heater share a power with a coil? Which engine is this. I'll look at the power distribution.

The 4 cyl

I'm using custom Aes wires on a maxiscope for that capture, also the same pattern I got on my uscope while in the air

Deltron wrote: If I use a test light to ground, I'll just turn the heater on. I've seen Paul do that in videos.

That is correct....IF the wire is good and that is NOT a bias voltage you are seeing. But, IF the wire is broken, and that is NOT voltage coming through the heater, but a bias voltage coming from the computer, the voltage will disappear and the heater will NOT turn on.

Ah, yes. Indeed. SO I could unplug the sensor and look for the voltage with a multimeter or use a test light and check the data pid. Those are my 2 options to rule out bias.

The O2 heater shares a power with the coils. I think the voltage drops you are seeing are the coils charging.

Deltron wrote: Ah, yes. Indeed. SO I could unplug the sensor and look for the voltage with a multimeter or use a test light and check the data pid. Those are my 2 options to rule out bias.

Yes.

I was curious if we could plug some math into this to add some info to the investigation into the cause of the spike. Those spikes are 42ms apart, roughly, with this capture being taken at warm idle speed. What's that equation to determine RPM?

Find the time for 1 revolution (in ms) . Divide 60000 by that number = RPM


There are 1000 ms in 1 second. 60 seconds in one minute. 60000ms in one minute.

Little late to the party. :blush: I agree with Andy that you're seeing ignition feedback into the heater circuit. NBD. As far as the heater PID not toggling, and the heater not getting grounded, that may be down to the PCM's strategy. Nissan and Mazda are selective about when and how they run the downstream heater.

Has anyone searched for TSB's on this code yet? I recall going down this path on a similar Mazda, and found a TSB that recommended cleaning the throttle body. :silly: Apparently, the carbon in the throttle body would 'hold on' to fuel, which would then get sucked into the engine during fuel cut decel. The PCM's looking for the downstream O2 to peg lean during fuel cut, so it'd get upset when the downstream O2 would suddenly go rich.

Dug up a couple examples from other Mazda's of similar years to yours. Both show the 'when I feel like it' downstream heater strategy. :silly: The first one wouldn't run the downstream heater at all at idle.

Tyler wrote: Little late to the party. :blush:

Better late, than never.

Well, file this one under things we'll never know. Wholesale.

Sometimes I hate being at a dealer. I often have interesting things ripped away and auctioned.

Thanks for the help guys.