Wife's '08 Mazda5 has had an appreciable drop in mileage over the past few months. It used to average 24-25mpg and now it's more like 19-20.
I changed the plugs (laser iridiums), checked the purge solenoid, changed the oil (M1 5W20), cleaned the MAFS, and checked tire pressures. My OBD2 software won't read the wideband o2 sensor for some reason. Monitoring bank 1 sensor 2 (I assume this is the downstream sensor) it sits at 0.88 to 0.85 volts during normal city driving. This sensor should show 0.45 volts right?
The long term fuel trim is plus 10%.
Is there an OBD 2 reader for Windows that will work with my Bluetooth setup so I can check the mA of the wideband?
141,000 miles on the van btw.
No, the rear sensor may not run at 0.45 volts in normal conditions. While that is rich (IMHO), I also know that other calibrators are comfortable with that downstream voltage, and it's still running correctly.
Most rear O2 sensor info will average 0.6 to 0.8 volts- since lower than that, and it's running to lean (barely), and higher than that mostly is too rich.
A question- when you monitor it- when you do a tip out of the throttle, and the engine turns the fuel off, does the rear sensor go to close to 0 V?
BTW, checking the current of the front sensor isn't going to help much- outside of the fact that there is current. What's the short term fuel trim? That's pretty much the output (corrected) of the front sensor.
Yes the voltage goes to 0V when coasting. Short term fuel trim is -5%. It's the 5 speed automatic if that makes a difference.
The system goes into closed loop fairly quickly which I assume is an attribute of the heated oxygen sensor.
Yup.
So it sounds very much like the closed loop air fuel control system is working properly.
Possible that the trans isn't locking up properly. Not sure why, but a suggestion.
It turns about 2500rpm at 65mph which seems like it's locking up but it's possible it's not doing it during normal driving.
The rear sensor takes a bit to start showing normal numbers, it's around .045 to .1V at idle sometimes.
Is a calculated load of 30% normal at idle with the trans in D while holding the brakes? Maybe I have an accessory on it's last leg.
The rear sensor takes a while to get going, for a variety of reasons.
30% load seems pretty high- is that fully warmed? What about in neutral?
and nothing is on, right?
The rear (sensor 2) O2 sensors aren't used for mixture adjustment. They are use to monitor cat health. Speaking of, is it plugged up at all? If it's constantly running rich (theory below), it won't last too long.
You may have a small exhaust leak pre-sensor 1. This would introduce free air into the exhaust and make the sensor read leaner. It would then try to correct the perceived leanness by adding fuel, hence crappy mileage.
bluesideup wrote:
Is there an OBD 2 reader for Windows that will work with my Bluetooth setup so I can check the mA of the wideband?
Not an exact answer to your question, but I just bought Dashcommand for iOS (I believe there is a Windows version) and a WiFi connector. I can read the wideband on my Volvo with this. You do have to monitor something different than the standard O2S12 in Dashcommand and it may take an aggressive PID search to get it.
I'm not so sure the downstream doesn't have any impact on fueling. I was chasing a lean misfire problem on my Miata that seemed to have been caused by a downstream sensor stuck at 1.245V. I'm not 100% sure this was the case, but it seems the most likely explanation. The upstream would read very low voltage and the short term fuel trim was still spiking negative when it would miss. It was kind of like the short term trim was spiking negative to see if it could get any different signal out of the downstream sensor. I didn't have the miss issue before putting in the faulty O2 sensor and it went away after. I did clean grounds at the same time I put in a new sensor, so it is possible that fixed the problem as well, but I kind of doubt it.
bradyzq wrote:
The rear (sensor 2) O2 sensors aren't used for mixture adjustment. They are use to monitor cat health. Speaking of, is it plugged up at all? If it's constantly running rich (theory below), it won't last too long.
You may have a small exhaust leak pre-sensor 1. This would introduce free air into the exhaust and make the sensor read leaner. It would then try to correct the perceived leanness by adding fuel, hence crappy mileage.
Yes, they are- especially in this car- both by it's age, and the fact that it's using a WB sensor in the front.
As for the leaks- the long and short term trim would have to be very far out of where he's saying they are for a leak to have that kind of impact on the fuel economy. The sum of the trims is just 5%- and it's a 25% change in fuel economy.
But, one can check for a pre-upstream leak pretty easy- open hood, listen- you'd hear that pretty easy.
30% load is with the HVAC system off and the car in Drive. I will try different configurations and post when the computer charges up.
I wil look up the dash command program and see if I can load that. There aren't any audible leaks in the exhaust and I'll double check for intake leaks.
Thanks you all for the help so far.
alfadriver wrote:
Yes, they are- especially in this car- both by it's age, and the fact that it's using a WB sensor in the front.
As for the leaks- the long and short term trim would have to be very far out of where he's saying they are for a leak to have that kind of impact on the fuel economy. The sum of the trims is just 5%- and it's a 25% change in fuel economy.
But, one can check for a pre-upstream leak pretty easy- open hood, listen- you'd hear that pretty easy.
It would be the first I I've ever heard of that uses a post-cat O2 sensor for primary fuel trimming.
You can ignore the short term trim. It's not stored. The plus 10% long term trim is the one to focus on.
Does your fuel have ethanol in it? Did it used to?
Does the long term trim stay at about +10 all the way through a drive cycle, or does it vary with RPM and load?
bradyzq wrote:
alfadriver wrote:
Yes, they are- especially in this car- both by it's age, and the fact that it's using a WB sensor in the front.
As for the leaks- the long and short term trim would have to be very far out of where he's saying they are for a leak to have that kind of impact on the fuel economy. The sum of the trims is just 5%- and it's a 25% change in fuel economy.
But, one can check for a pre-upstream leak pretty easy- open hood, listen- you'd hear that pretty easy.
It would be the first I I've ever heard of that uses a post-cat O2 sensor for primary fuel trimming.
You can ignore the short term trim. It's not stored. The plus 10% long term trim is the one to focus on.
Does your fuel have ethanol in it? Did it used to?
Does the long term trim stay at about +10 all the way through a drive cycle, or does it vary with RPM and load?
You would not be the first I've heard that from. But fuel trimming based of of the rear sensor has been around since the sensor went in exhaust in 1996. I've personally been using since 1999, when I started to do emissions specific work. For WB set ups, since those sensors are not as accurate around stoichiomety as NB sensors are, it's used both to trim the fuel and to trim the front sensor. So I'm 100% sure that this car uses the rear sensor to trim the fuel.
Still, a 10% shift in the long trim fuel (which we don't know the long term history of- could have been there for many years), isn't nearly enough to change the fuel economy 25%. And I would not ignore the front trim just because it's not stored- it's error is what feeds the long term trim. That's how the system works.
But it is a good idea to track the fuel trims over a whole drive. There should be a handful of ranges where it's kept.
After battling the flu over the holidays I finally am able to check on the van again. Last gas tank was 245 miles and filled with 12.6 gallons, slightly less than 20mpg.
At idle in Park with no accessories or lights on fully warmed up the load is 22%, 27% in Drive. With the A/C and lights on it's 31% in Park.
Long term fuel trim is still at plus 11% and the rear o2 is giving weird numbers. Fully warmed up it's registering .085V then settles at .115V. I'll post back after a short drive.
The secondary o2 comes alive within 100 yards of driving but randomly drops to those low readings and stays there for a bit (randomly meaning not during decel). Idle seems to read 0.875 plus or minus and cruising stays near the same number.
Is there an OBD 2 program for my PC that will let me read something useful from the primary o2? This program just shows "Closed Loop-oxygen sensor normal".
For anyone else's reference I ended up throwing parts at the problem which I hate to do but the noise coming from SWMBO was getting unbearable. A new primary o2 didn't fix the mileage but the new secondary o2 has fixed the problem. I bought it from Amazon, the old one was registering strange voltage (0.045V). Now I get 22mpg city and 30mpg highway (subtract 10% if SWMBO is driving).
