2015 Mazda CX-5 2.0L GDI Naturally Aspirated Manual Trans running lean P0171

Sorry for being light on the details but was at a buddy's shop and didn't have much time to look at this. Car only has 80,000km/50,000miles on it. P0171 - lean bank 1. This is a recent issue. Owner said she noticed a bit of a fuel smell after a cold start. Gas mileage seems to have gone down quite a bit. Sometimes the vehicle does have a very slight stutter upon initial startup. Maybe just slightly extended crank. LTFT at +20% and STFT is +5 to +20%. Fuel trims stay the same as you rev the vehicle. A +30-40% fuel trim vacuum leak would make enough noise to hear and this thing isn't hissing at all. Fuel pressure actual mirrors the desired.

There's lots about this car that I don't like....

The MAF g/s is very low. Like 1.8g/s at idle and barely increases as you rev. 1500rpm is only about 3g/s. 3500rpm is only 15-25g/s. Upon a quick snap throttle I didn't even get higher than 2.5V.

I'm using a Snapon Verus Pro and questioning the authenticity of those g/s numbers...? The "Guided Component Tests" of the Verus lists a spec for both the MAF and the MAP at idle and at 2500 for voltage. This car matched both. The MAF only has a single exposed sensor wire. I'm pretty sure that's the IAT. The hot wire is way up inside the housing of the MAF so I couldn't inspect and you can't clean those. I sprayed up there anyways but of course no difference. The intake boot is free of cracks/tears and the PCV breather line is also intact.

The evap purge valve is clicking away at idle. When I pinch the vacuum line off I get a STFT improvement of around 10%.

So my questions are why are the MAF g/s numbers so low?? How is that not setting a code? Do I trust the numbers? Is it possible that the throttle is only barely being opened and the car is just going into ultra lean burn mode? It is a GDI... Would I have to go for a drive to get more familiar numbers? Maybe those numbers were accurate/normal for when stationary?

If this is a bad MAF or pirated air then why don't I have some sort of MAF/MAP correlation code..? This thing sets a P0171 almost immediately after going into closed loop after a code clear.

Also what's up with the improvement from pinching off the vacuum line to the purge valve? If that's being commanded on then shouldn't that have been accounted for? Also shouldn't that air contain plenty of HC's..?

Currently I like the idea of pirated/unmetered air the most for this. Maybe something with the PCV/EVAP purge..?

I guess bad gas could also be a decent possibility. It's GDI so I wouldn't think the injectors could get restricted but maybe they could get carbon buildup around the nozzles? Causing a more focused spray and less of a cone mist? Then maybe it just wouldn't atomize very well and get an incomplete burn..?

The guy doesn't have a smoke machine and I ran out of time on it. But ya, any thoughts/comments/experiences would be welcome in the meantime until I see it again.

Thanks.

I agree that the MAF seems very suspect. Can't recall ever seeing a MAF read lower than the engine displacement at idle. Next time around, I'd be interested in test driving with the upstream A/F sensor and Load PIDs up. A measurement error with show up in Calculated Load.

You'd think the purge volume would be full of gas fumes, but it depends on the content of the canister. If the vent valve is open and the charcoal is relatively dry, you'll end up getting more air instead. Not that big of a deal, usually.

Tyler wrote: I agree that the MAF seems very suspect. Can't recall ever seeing a MAF read lower than the engine displacement at idle. Next time around, I'd be interested in test driving with the upstream A/F sensor and Load PIDs up. A measurement error with show up in Calculated Load.

For sure. I immediately thought MAF all the way after only a minute of looking at this thing. But I just can't get over the fact that it has a MAP sensor to compare things to... And the voltage specs in the "Guided Component Tests" matched up for both the MAP and MAF. Not that I think that was a particularly extensive test but still...

Tyler wrote: You'd think the purge volume would be full of gas fumes, but it depends on the content of the canister. If the vent valve is open and the charcoal is relatively dry, you'll end up getting more air instead. Not that big of a deal, usually.

That's what I was thinking about as well with a dry canister. It had an improvement when I pinched the line but nothing overwhelmingly so. Totally made me think of a time wasting trap of a wild goose chase. If it's being commanded on it's being commanded on for a reason. Since it has a MAF and a MAP it should be able to compensate for that enough to not be the whole problem.

Also, as far as I can tell this is just a regular Otto cycle engine. No Atkinson or Miller cycle shenanigans... The MAF reading SHOULD be higher... But the sensor is $250 our cost so like $350 to the customer.... I just can't stop thinking about how it should be able to detect bad MAF readings... And what if it has intake valve carbon buildup... or bad gas... or carboned injectors causing a bad spray/atomization pattern..???

This car feels like it really, REALLY wants to be a $2000 money pit...

Oh ya, I totally forgot. I did unplug the MAF and the car idled around 1000rpm... about 300rpm higher. Can I take that as definitive proof of a bad MAF or unmetered air..? Or could that just be a part of the default strategy of running with a hard MAF fault..?? This was right when I was told to kick it out so I didn't have a chance to see if the fuel trims were improving or if it just kicked it into some kind of open loop fault...

I really wish I had a smoke machine and a gas analyzer I could throw on this thing...

I was thinking about suggesting a WOT snap test with a scope on the MAF signal wire, to see if you get that initial spike when the throttle is first opened. Problem is, lots of newer drive-by-wire systems won't allow WOT in Park, so you may have to get creative with that. :lol:

FWIW, I don't feel like carboned valves or clogged GDI injectors would account for this kind of lean condition. The cold start misfire and slightly extended crank, sure, that could definitely be a carbon buildup issue. If the lean were caused by injectors, I think you'd have a more consistent misfire.

The bad fuel, definitely. Or E85 in the tank. Mazda doesn't play the Flex Fuel game, as far as I know. The only problem there is that it won't account for the low MAF readings.

EDIT: Possibly a skewed upstream A/F sensor? Not likely, but a quick check against the downstream O2 could rule this out.

Tyler wrote: I was thinking about suggesting a WOT snap test with a scope on the MAF signal wire, to see if you get that initial spike when the throttle is first opened. Problem is, lots of newer drive-by-wire systems won't allow WOT in Park, so you may have to get creative with that. :lol:

Definitely the thing I want to do most when/if I see it again is a WOT run in 2nd. Unfortunately where this shop is located that could take 15+ minutes finding a stretch of road where you can do that...

Second I would want to do is block all PCV and EVAP lines trying to eliminate any other source of air into the engine and watch the MAF g/s and fuel trims for change. I did notice that when I pulled the PCV breather line out of the intake boot the STFT did go from +10% to close to +30%. Not sure if that's enough of a reaction to rule out additional unmetered air...? I didn't get a chance to see if the MAF numbers changed much.

Tyler wrote: FWIW, I don't feel like carboned valves or clogged GDI injectors would account for this kind of lean condition. The cold start misfire and slightly extended crank, sure, that could definitely be a carbon buildup issue. If the lean were caused by injectors, I think you'd have a more consistent misfire.

I'm of a similar line of thinking unless if maybe this is trying to do some major statified injection and the carbon is impeding swirl and causing incomplete combustion...? Not sure how much of a stretch that is. This thing is not misfiring. It runs smooth. Just maybe feels like slight lack of power.

Tyler wrote: The bad fuel, definitely. Or E85 in the tank. Mazda doesn't play the Flex Fuel game, as far as I know. The only problem there is that it won't account for the low MAF readings.

We don't have E85 in our area but we do have, supposedly, really bad gas as it's all trucked in from a distance. Likewise I wouldn't think bad gas would cause the low MAF numbers. Unless if, somehow, those numbers AREN'T actually low..?

Tyler wrote: EDIT: Possibly a skewed upstream A/F sensor? Not likely, but a quick check against the downstream O2 could rule this out.

Yeah, definitely want to do a WOT run looking at load, both O2 sensors, MAP/MAF, etc. The A/F sensor PID was in microamps. Would be anywhere from -150uA to +150uA. Most of the time it would be within +/- 50uA of 0. Not sure what the scaling is or if 150uA is a lot.

All in all I had less than 20 mins to look at this thing. It just makes me feel very uneasy with me trying hard not to spend lots of money on parts or lots of money on diagnosing time... :huh: