O2's reading lean. Wiring problem?

Hi everyone. My name is Anthony and I have a 2005 Mazda 3 with the 2.3l engine. Recently I have been having trouble with my up oxygen sensors (upstream and downstream) reading lean (~.25 v). Its only a 4 cylinder so there's only one bank. This problem really came out of the blue as one day I checked and they were working perfectly fine and then testing them another day, this happened.

Now the weird thing is that the sensors don't respond while idling or by just putting my foot down on the throttle, but I can get them to jump when I let off the gas.

While driving down the highway with cruise control on and my scan tool plugged in, I was monitoring the O2 signals and they by themselves start working for a very short amount of time, and then go back to reading lean. This will keep happening throughout the drive.

I have gone and done the test where I back probe the signal wire with the sensor unplugged, touching both the pin and the positive terminal of the battery. This made the signal go up from the bias (.2v) to 1.275v if I remember correctly. Whatever the number was, I got the same result for my downstream and upstreams which then made me believe my wiring was fine, but I also know my sensors are fine.

My thought now is that maybe there is some small break somewhere in the harness which occasionally touch? Or maybe a bad ground?

If anyone has any idea on what to do I'm all ears I just don't know where to go from here. Thanks in advance.

Can you drive the Sensors RICH by inducing propane into the Air Intake?

I suspect the sensor heating circuit isn't working fully warm the engine up and hold the engine at 2500 to 3000 rpm for about 2 minutes and return to idle to see if the up stream sensor gives the characteristic rich lean rich lean roller coaster. Also check the loop status is going closed loop.

@ pole71

I did try to do this, all I had was a little propane torch so I don't think it really released enough gas to cause any change. I didn't even hear a change in rpm. I think I'm definitely going to have to retry that test.

@Andy.MacFadyen

I have done this before, but its only when I let off the gas that I get the characteristic lean-rich signal for like 2 seconds.

What are telling you the other parameters?

LTFT, STFT, pulse width, MAF

That also sounds as dirty MAF sensor, but those parameters can tell you if ecu is reading the same that you are reading in O2 sensor.

@guafa

At idle STFT is maxed at 25 and LTFT is at 0. At about 2000 rpm they're both maxed. As for pulse width my tool isnt advanced enough to measure that so I dont know. MAf is new because I broke my other one (long story), but it reads 2-3 grams/sec which I think is normal.

Definitely sounds like you're looking for a fuel delivery issue. It seems to me that the O2's are probably telling you the truth.

A fuel pressure check would be a good idea, but there's no test port on your 2.3L engine. Makes that test tough to do, especially if you don't have a test kit with the connectors to tee in the gauge.

IMO, the next easiest test would be to watch Calculated Load during a WOT run through 1st gear. Floor the gas from a stop and let the engine force an upshift into second gear. Record the highest Load reading. If it's over 85%, there's no MAF or volumetric efficiency issue. Find a way to test fuel pressure. If it's lower than 85%, there's a MAF or VE issue.

@ Tyler

Okay so I have done a fuel pressure test, but I plan on redoing it with a new gauge. When I did it the gauge read ~55 psi but it was the needle was fluctuating very quickly like + or - 2 psi. When gassing it the needle became more steady at like 55-56 psi, but I just thought the fluctuation was just due to the cheap harbor freight gauge. Would this be indicative of a bad fuel pump?

Yes with those fuel trims fuel delivery ---- looking for a flow restriction such as a blocked fuel filter .

ajherrera58 wrote: @ Tyler

Okay so I have done a fuel pressure test, but I plan on redoing it with a new gauge. When I did it the gauge read ~55 psi but it was the needle was fluctuating very quickly like + or - 2 psi. When gassing it the needle became more steady at like 55-56 psi, but I just thought the fluctuation was just due to the cheap harbor freight gauge. Would this be indicative of a bad fuel pump?

Not necessarily? I'm not sure what the spec is, but that pressure sounds reasonable. It'd be interesting to know if the gauge maintains 55 PSI during a test drive (with the O2's showing lean).

Have you noticed any other symptoms? Let's say misfire at wot, hesitation, rpm lower than normal at idle (any other which can give you more clues).

I think you should also check for a vacuum leak.

Fuel pressure also sounds good for me and 2-3 g/s at idle is ok (it would be nice to have peak number at wot)

@ guafa

Yes so my car does have a very noticeable hesitation at around 2000- 3000 rpm. Its annoying to pass on the highway because I'll be sitting there for a few seconds foot all the way down, and really no response. At idle my engine does bounce the revs, with it being the worst at startup or when just stopped at a light, the car will pretty much stall and bring itself back to life. When it is warm the idle will probably be around 750 ish rpm but fluctuating, like it almost sounds like a rotary with the pulsations that come out the exhaust. I have checked for vacuum leaks, but I found nothing after smoking it.

@Andy.MacFadyen

Yeah as weird as it is I don't believe my car has a fuel filter? I have checked the injectors and to me they didn't seem to be dirty and the car held fuel pressure at 50 psi when shut off.

@Tyler

Yeah that would be a good test to do, but unfortunately my gauge is at home so I wouldn't be able to test that. I'm away at college right now.

ajherrera58 wrote: @Tyler

Yeah that would be a good test to do, but unfortunately my gauge is at home so I wouldn't be able to test that. I'm away at college right now.

Understood. That Calculated Load test is your next best test, assuming you have a scanner with you.

Also have a good look at the intake tube and any associated ducting. Unmetered air can behave in a similar way, from something like a cracked tube or a loose PCV hose.

@Tyler

Oh okay I see, but I did forget to mention I have a manual transmission. Is there any other method I could use? Thanks.

ajherrera58 wrote: @Tyler

Oh okay I see, but I did forget to mention I have a manual transmission. Is there any other method I could use? Thanks.

No problem! Same method works. Just wind out 1st gear at wide open until redline or close to it. Be responsible, of course. Don't break anything and don't attract police attention. Highway on-ramps are ideal.

The overall idea is to get the engine breathing the most air it possibly can, which will be reflected in the MAF g/s and Calculated Load reading. Wide open is key because you won't see the peak values any other way. This method allows us to rule out stuff like airflow mismeasurement, cam timing and other engine breathing issues.

@Tyler

Okay thanks so much I'll try to go out later today and try it. I'll post the results

@Tyler

So I did perform the test, and my load value didn't go over 85%.