mine failed smog, with the 10% fuel trim. i ended up buying a cat, and it passed smog easily, and runs great, mileage is fine no check engine light, so maybe 10% fuel trim isn't a problem?

 

These problems could also be caused by an aging O2 sensor. We are lucky/unlucky in that (if I understand correctly), the longer-lived, more-reliable, rear narrow-band sensor will generate a phantom trim to correct for an aging front wide-band sensor, as well as delaying a trip into true limp mode if the front fails or goes too far out of spec. I think a way to test this might be to disconnect the rear sensor and drive gently on level ground at a steady speed. Normally, this should give you a LTFT ~ +10 (in your case) and a STFT ~ 0. If you suddenly get under the same conditions, say LTFT ~ 10 (because it takes a while for LTFT to change) and a STFT ~ +/- large, then the front O2 is likely defective.

If you decide to renew the sensor, get Denso, *not* Bosch, btw.

 

j9fd3s: So does your LTFT still stuck with ~10% after you replaced the cat? My cat was replaced along with the engine about 20k ago...hopefully it has not failed again.

HiFlite: Why ~10% when I unplug the rear O2? Is that just what people have found?

I am wondering though...and please don't hesitate to correct me if I say anything wrong here. I thought the cat is supposed to clean out the exhaust gas so the rear O2 should be low all the time? My rear O2 does stay high for the most of time IIRC, in cruise.

By the way the STFT in my car is only making small change in my car. I can kinda see it responds to the wide band sensor. It does not jump crazy like in my Protege5, which only has narrow band sensors.

 

10% because I'm going by what you said you have. As long as the LTFT memory is not reset by disconnecting the battery or loading an new tune, it will remember what you were running.

STFT will vary all over the place. LTFT should be rather stable - it averages the STFT readings and tries to zero them out over time and drive cycles. I still think what you or your app is reading, is mislabeled or still a point of confusion.

You can't really tell how the fuel trims relate to the front O2 readings without logging data and seeing how the measured reads match the AFR tables in the PCM.

As far as I know, there's no way to see the phantom trim generated by the rear O2 sensor.

 

Those pids you posted are the mode 6 data for emission monitor test results, if you give me the year of your rx8 ill let you know what they are. they changed every 2 years.

 

I searched around and it seems like if the rear O2 sensor is indeed failing, the PCM should throw a CEL. It seems like every once in awhile the PCM does a rich-lean/lean-rich transition test during open loop operation to make sure the sensors are functioning (and not removed). Also it compares the value between front and rear O2 sensors to make sure the cat is functioning (and not removed). There is no CEL on my car...

I do have one of those highly debated bluetooth scanner thing. :P Most of the data makes sense though. It graphs data in real time so I can see the trend between the STFT and the front O2 sensor. I can log too but it doesn't take long to see the control loop in action.

When I said STFT is only making small change I mean in cruise in a relatively flat ground. Compare to cars running narrow band front O2 sensor like my Protege5 the STFT is much more stable. Not sure why I brought it up in the first place.....

I attached some of the screenshots I got from the tool and app...You can kinda see the control loop in action between STFT and O2 sensor. The LTFT is just stuck high but this was before I cleaned the MAF. Now it only stuck around 10%.

I hope Team is not reading my thread or he is probably shaking his head...or well I am a poor home mechanic.

 

Yes! That was the end of my search...Mazda apparently does not public their mode 6 data...like the amount of things they use SST for...it's annoying.

I have 2004 GT manual. Thanks!

 

ya ive never had to look up mode 6 on a mazda, scanners I use all translate it... Im searching right now through mazda service info and not finding the ID you posted. what scanner are you using? and are you checking it on global data or through vehicle build?

 

OK this is going to get a lot of flame...
This is what I use:

along with Torque app. OK this is not what people traditionally use for scanner but whatever it's cheap and it's been working great, if not better. (kinda like fuel injection in the beginning...??)

No the app is not smart enough yet to know what the ID means. It knows most of the ID but not these few...strange. If your scanner can interpret it do you mind to look it up and tell me what it is? :P

 

Lol actually havent seen that one yet. well the $99 and $cc are not listed on any info I can find, most likely that scanner is displaying them incorrectly. Cheaper ones get confused by hexidecimal pretty easily (%, #, $, etc.) and will display the wrong digits. if you list every TID you got i may beable to single out what one its displaying wrong. as far as your driveability issue, if you have solid fuel trims at idle and lean trims cruising it is almost always the MAF. easy way to test it is do a WOT run and look at the calculated load pid, im not sure on an rx8 but most engines should be above 80% with a good maf on a good engine. or clean it like you did and if it changes then replace it, cleaning is a bandaid.
edit: and are you going into global info or are you telling the scanner its a mazda? usually global is more accurate for mode 6

 

I'm not guessing that the rear O2 sensor is bad, but that the front may be failing. The LTFT and STFT chart you show makes sense. However, the front O2 looks pretty noisy and doesn't correlate with what the STFT is doing.

Is your crankcase vent hosing stock? ie. did you install a catch can?

BTW, my stock tune with no mods on a 25k mile engine gave LTFT's around +6%.

 

his front o2 is flat lined at 1v, his rear o2 is the one bouncing around and not much but quickly.

 

Right It made quite some difference (from ~17% to ~10%) after I cleaned the MAF. That was pretty amazing actually...I was just surprised because on my Protege5 the MAF has never been cleaned and the trim is very close to 0% all the time.

Sorry those charts are confusing. Yes the app does not do a good job explaining which sensor is which one. The lower two boxes actually represents the same sensor (from what I believe.) One is the raw voltage and one is the interpreted O2 sensor equivalence ratio. I believed that then gets translated in the measured AFR. The rear O2 sensor has never been my concern so I left it off the graphs.

The O2 sensor equivalence ratio is always very close to one so it looks like a flat line. The graph should zoom in more...After awhile I took that graph off because I can't get anything out of it. It's better to just look at the measured AFR or raw voltage value anyway.

The whole car is stock. OK I have BHR coils but that has nothing to do with this. So maybe the rx8 gets high LTFT anyway? I just hate to see ~10% on the RX8 and how perfect the LTFT is on the Protege5....

I don't think the scanner knows the manufacturer. There is no option for that input. I will try to get all the ID's and post it here.

 

"TIP: If you have a scan tool, look at the Short Term Fuel Trim (STFT) and Long Term Fuel Trim (LTFT) values. Normal range is plus or minus 8. If the numbers are +10 or higher for STFT and LTFT, the engine is running LEAN. If you rev the engine to 1500 to 2000 rpm and hold it for a minute or so, and the STFT value drops back down to a more normal reading, it confirms the engine has a vacuum leak at idle. If the STFT value does not change much, the lean fuel condition is more likely a fuel delivery problem (weak fuel pump, restricted fuel filter, dirty fuel injectors or a leaky fuel pressure regulator) than a vacuum leak.

For more information about using fuel trim to diagnose a lean fuel condition, read this article on Fuel Trim by Wells Manufacturing (PDF file, requires Adobe Acrobat to read)."

from: Engine Vacuum Leaks

 

and to support the possible bad MAF sensor idea...(I know justifying a possible cause is probably not good...it prevents the mind from thinking objectively...) Anyway I feel I should put it out there. The LTFT is around 10%, but it's more like 9.4% @ 55mph, 10.x% @ 60 mph, 11.7% @ 65 mph, and 12.x% @ 70+mph. That is using cruising control on a relatively flat ground.

@idle it is around 1.x%. It jumps straight to ~10% after ~2000rpm

 

nah the one on the bottom left is your rear o2, im assuming the 1x2 means b1s2. factory airfuel/wideband sensors have never been useful on scanners, at least to me. rear o2 will usually float around .4v but can actually be used to control your fuel system if you go into open or closed loop fault.

 

It certainly is not a vacuum leak. I had a vacuum leak on the Protege5 not long ago and like you said the LTFT is high at idle but decreases as MAF gets higher. (the leak opening proportional to the throttle body opening decreases). The RX-8 is the opposite.

I just replaced the fuel pump. I have not check the fuel injector and pressure regulator. Of course I have not try another MAF and front O2 sensor...

 

what this means is.... you probably are not actually running lean but you maf is incorrect making your computer adjust fuel. think about it like this, @ 60 mph you are using 10.x% more fuel then needed (10.x% less mpg).
I dont actually own an rx8 yet, trying to get one this weekend so i dont know the specifics of what end of the maf fails more often but on every other car out there (especially fords) maf's fail all the time and this is how they fail. Then customers dont want to pay for the oem part so i put a napa maf in and ship the car, then they come back the next week and i put another napa maf in. sometimes i will order 2, and which ever brings the trims closer to zero stays in the car.

 

omg i need an rx8, im so used to posting on rx7club but these things are soooo much easier. anything with OBDII is simple diag

 

Narrow band sensors read voltage changes. However, I think that wide-bands read current, though there is a constant voltage supplied to them.

Perhaps borrow a proper OBDII tool to help interpret the Android is reading??

 

nah only current going through either o2 sensor is for the heater circuit, ecm is only using the voltage signal from them.
Maxchao I just realized you said you just replaced your fuel pump... what was the reason?

 

Wideband on the RX-8 isn't voltage based...the rear narrowband is

 

In the first page of the thread I was suggested to replace the fuel pump and I think that made sense...the RX8 fuel pump is weak and degrades overtime, and since I believe that I still have the original fuel pump at 120k miles, I decided to replace it.

Especially after I found an claimed OEM one on Ebay selling for $58.00...Cheap enough for me to just test my theory. In any case nothing changed after I replaced it.

Now my attention is going back to the MAF...this 'proportionally' increasing LTFT does sound like something is metering off the center line...but I guess it can also be the front O2 and fuel injector. Can fuel injector fail this way?

Actually I can clean the front O2 sensor as well...Maybe I will do that first.

 

Right well it's more like the engine is running lean based on the MAF...but running fine based on the O2 sensor. In the end of day it's the O2 sensor that controls the AFR in closed loop, MAF is just a starting point.

I get around 20 mpg on highway at 80mph. I don't know how some forum members get 22-24...Unless they cruise at lower speed...?

 

There are multiple LTFT data points; idle, cruise, tip-in, etc. Readers only display the LTFT for the condition it is reading at the moment. Technically there is not an emission issue until hitting +/- 25%. It's correct that a vacuum leak has the most impact on the idle LTFT. +10 does not mean the engine is running lean. It means that initially the PCM is receiving data that the engine was operating lean relative to its programmed target values and the PCM compensated the trim to bring it in line with the target AFR. There are so many possible issues on why trims will vary including normal variance, particularly with the OE map. You can't simply guess an O2 sensor etc. it needs proper diagnosis with the proper tools