Long Term Fuel Trim at 20% 04 Mazda 3 2.3L

Joined
Jun 15, 2022
Messages
21
I've been trying to track down a CEL on an 04 Mazda 3 2.3L (CA Emissions Edition). It will throw a system too lean code since the LTFT reaches 20.3% at light to medium loads. It reads normal fuel trims at idle. The O2 Sensor 1 will fluctuate in voltage from ~0.05 up to ~.9. Sensor 2 is a bit more stable, but I've got to go back and check my recordings.

I've checked the PCV system and replaced the valve, AOS gasket, intake manifold gasket, and TB gasket.
Any idea what a normal O2 sensor voltage should be? How else can I check for a bad sensor before replacing it?
Similarly what should the MAF reading be at idle and such? I want to make sure the sensors are giving me reasonable values to see if I can rule them out.
Beyond that could it be fuel system related or something in that complex and troublesome intake manifold?

Any help would be appreciated.
 
Yeah, I'm not really sure how to troubleshoot the sensors via readings. I don't just want to parts cannon it right now.

I might go smoke test it to see if it's ingesting air somewhere past the MAF, but it idles fine so it doesn't seem like there's a vac leak.
 
Its adding fuel at +20%, So in no particular order.

MAF is reading low - which isn't likely because if it were dirty it would read low at idle but maybe not? Could also be the temperature sensor inside the MAF. I had that once, or I think so - took a long time to find. Finally replaced MAF out of frustration and voilla.

It could be vacuum leak, but you get the most vacuum and the lowest overall flow at idle - so probably not.

It could be a sticking injector - not putting in enough fuel. You could pull the plug and see if one of the plugs appears to be lean and the others rich. Could also be low fuel pressure. In my experience these are all possible but less likely.

Could be the primary A/F sensor - aka the first 02 sensor. Reading OK at low flow but not at higher flow.

If it were mine, I would clean the MAF sensor only because its a good idea and almost free. Use ONLY MAF sensor cleaner. If that did nothing I would next replace the primary A/F sensor with OEM or OEM branded - ie if a Bosch comes out, put a Bosch back in.

Good luck with your troubleshooting.
 
Last edited:
Clean MAF sensor isnt a bad idea so I agree with that. Dirty injectors are possible so running a tank with a heavy dose of detergent additive might help.

To troubleshoot the O2 sensor thats straight forward swap the front and the rear as they are typically identical sensors. If the fluctuations follow the sensor its a bad sensor if not the sensor is fine but the wiring still could be bad.
 
When was the last time the O2 sensor was changed? They usually start getting sluggish or quit after 120k miles for heated sensors.
 
Well under load you need more fuel. Have you at least checked fuel pressure?

Graph B1S1 to see if it's likely working correctly
 
  • Like
Reactions: BC1
Thanks for the help.
I've cleaned the MAF with MAF cleaner and checked the intake boot when I had it off for the PCV change. No difference/issues that were visible to me.
The car is a CA emissions version so it has 2 pre cat O2 sensors. I don't know which one is S1B1, but I don't think they are interchangeable since the connectors are different. I need to look at the post cat one still though. I do however suspect the O2 sensors just on a hunch.
I'll see what I can do about uploading my OBD data later.

Vehicle has ~170k miles.
 
Lean condition under load only - I would be looking at a dirty or defective MAF sensor.

Sometimes cleaning the MAF sensor works and sometime it does not. So it is best to replace with an OE part. Aftermarket is a hit and miss. Good luck.
 
I do however suspect the O2 sensors just on a hunch.
Wouldn't surprise me based on their symptoms. Any basic logging app you should be able to see them cycle at regular intervals from either 0.1 to 0.9V for a narrow band, and about 0.5 to 4.5V for a wide band. The actual exact voltages aren't that important unless there way out of wack, but the sensor at steady speed should cycle in a regular pattern. There are lots of youtube vids showing this. Given there are two sensors, they should cycle at about the same general rate, although probably not in sync. No idea why there are two in one stream - never seen that before?
 
O2 sensor values look OK from my live data on my scan tool. Getting an approx alternating voltage reading.
These were all taken with the car in Park and at either idle or up to 2500RPM. I wasn't able to get the LTFT to read high since I could get it under load and take pics.

At idle:
IMG_0350.webp

IMG_0353.webp


I am seeing a slow delay when on throttle then letting off.

IMG_0354.webp
IMG_0351.webp
 
MAF values with increasing RPM. I think this is OK, but I need to double check it at a few constant speeds:
IMG_0352.webp
 
I found my pics of the freeze frame data from when the P2177 code was thrown:

Closed Loop
ECT: 186 deg F
Load: 67%
STFT: 8.5%
LTFT: 20.3%
Manifold Pressure: 27.7 inHg
Engine Speed: 2195 RPM
veh speed: 38mph
ign timing: 22deg
IAT 105 deg F
MAF flow rate: 3.5 lb/min
Throttle Pos: 34.1%
 
It will throw a system too lean code since the LTFT reaches 20.3% at light to medium loads. It reads normal fuel trims at idle.
I would test fuel pressure. This seems to me like the fuel pump is too weak to maintain pressure at higher loads.
 
  • Like
Reactions: BC1
The 20% LTFT means there's a pretty bad vacuum leak That is, air is leaking into the intake system after the MAF. It could be a cracked hose or pipe, a bad intake gasket, etc. Have the intake professionally smoked tested.
 
Back
Top Bottom