Symptoms of a dying AEM wideband 02 sensor?

Femi · Apr 3, 2009, 11:08 PM

I've had my wideband O2 sensor and my AEM EMS on the car for about a year now, and over the last 3 months or so, I have noticed:

- After hot startup, the wideband sometimes just stays on full lean for maybe 10-20secs after the start, coming down to the normal idle AFR after blipping the throttle a few times. On other days on warm start, it works normally, showing my usual 13.5:1 idle AFR as soon as i start the car. If I attempt to drive the car while it is reading lean, you can actually hear the car stuttering slightly.. Again, letting the car sit until the wideband comes down to the usual 13.5AFR, or blipping the throttle a few times sorts this out.

- This could be my imagination, but when going into boost, it seems a bit slow in operation, almost like a lazy or drunken sensation to the way it operates, or as if it isn't even monitoring MY engine.. I know that widebands read lean when you let off the gas, but should not do so almost haphazardly out of nowhere. I could jab the throttle, and the wideband is still registering what happened after maybe half a second after the deed is done.. Other times, I could jab the throttle, and the wideband reads lean for what seems like an eternity in response to my opening the throttle up initially.. Sometimes as I am going into boost, it would read rich like it should, and then all of a sudden snap to reading leaner and leaner, which of course signals me to get off the gas out of fear. This is not even at WOT either. Might be 60-70% gas.

I have sent logs to my tuner in the past, and he also passed a comment last time about there being a bit of a lag between what the engine is doing and what the wideband O2 sensor is reading according to the logs, and how it can be dangerous.

- When driving in open loop (I think it is open loop that it is called when you drive out of boost with the AEM reading the wideband??), I have noticed that when I shift gears at at anything approaching boost, the car seems to stutter or hesitate slightly as I reintroduce the throttle, and then accelerate.. almost as if the wideband is reporting info to the ECU out of sync or something.. And after shifts, the wideband starts creeping up to the lean end of the scale as I re-introduce the throttle.. And when I say "shifting", I'm not talking about your race driver stylie shifting.. I'm talking regular shifting...No speed shifting here.

The car was not doing any of this after it was tuned last July. The laziness in the wideband really became apparent at start over the winter.

My setup is such that it idles at around 13.3-13.5:1 due to cams and injectors. The sensor has been on the car for a year now, and about 8-10K miles.

The car has been tuned on race gas once, but that was when tuned last July, and has never seen C16 or any other race gas after. Just 94 octane pump gas.

Car's oil consumption is normal for a built motor also. Nothing unusual.

Sensor is mounted pretty close to the turbo's hotside.. Around 3-4 inches from it.

Are there any other characteristics or symptoms to look out for when the O2 sensor has had it? From what I read, a slow responding O2 sensor is a very dangerous thing to have, but there doesn't seem to be much info hanging around on how to catch it before driveability or engine safety really becomes an issue.

I have seen the AEM video on testing the widebands on youtube, but that seems to tell you how to catch a wideband that is already gone... Not one that is in the process of leaving the building..

Only piece of info on sensors I have found is a rule of thumb from some website, which states that you should change them out as often as you change spark plugs... Now I change plugs on my car every service interval, due to the rich idle mixture I run (basically richer than 14.0:1 99% of the time), but not sure if I should be changing sensors that often also

The Hydra I used before had an NTK sensor, which gave me no issues for the year that I used it, and it had to deal with even richer idle mixtures in the 12's on a daily basis. No leaded race gas on that one though...

Or are there any checks the AEM EMS can perform, or readings to look at on AEM Pro that would alert you to an O2 sensor in need of replacement?

Again, Not sure if this is indicative of a tired O2 sensor, or just tuning issues that need ironing out, or perhaps a pre-turbo leak or something?

I would like to rule out the wideband sensor itself though before I go looking elsewhere on the car though.

BenJ · Apr 3, 2009, 11:17 PM

Mine lasted all of about 6 months and it is pretty obvious when it starts to go. Mine started by showing lean occasionally and after about a month or so it just sat on full lean all of the time. Do you run meth by chance? The reason that I ask is that the rich micxture can foul or kill the o2 sensor quicker....as it did mine.

Femi · Apr 4, 2009, 03:15 AM

Quote:

Originally Posted by BenJ

Mine lasted all of about 6 months and it is pretty obvious when it starts to go. Mine started by showing lean occasionally and after about a month or so it just sat on full lean all of the time. Do you run meth by chance? The reason that I ask is that the rich micxture can foul or kill the o2 sensor quicker....as it did mine.

Interesting.. I do have methanol injection installed in the car, but have not used it. I run off straight 94 octane gasoline from Sunoco, and the C16 once as mentioned in my original post when the car was tuned last summer..

Perhaps this might explain partially why my sensor has not died off so quickly, although I suspect that the constantly rich mixture that I have to run has done a number on the sensor.

I honestly never knew that O2 sensors go out so quickly though... I'm almost thinking that I would prefer a wideband that has a much longer life than 6 months to a year.

It would also be nice to have one that is tolerant of methanol injection as well as the occasional shot of C16...

awdboosted · Apr 10, 2009, 01:14 PM

If you do a WOT pull and it reads much leaner then it has in the past (ex. car is set at 11.2 and reads 12.8-13.00). The sensor is no longer any good. They go bad fast from what I’ve observed. My last 4-6month if I’m lucky. Sometimes the nlts or back fire cracks the porcelain and kill it. It’s a good thing they are cheap and I always keep extras.

czilla · Apr 10, 2009, 01:22 PM

they go real fast i bought 5 of them from club rsx just for backups

czilla · Apr 10, 2009, 01:23 PM

14.8 reading is also a sign of a crapped out widebnd.

awdboosted · Apr 10, 2009, 03:50 PM

Quote:

Originally Posted by czilla

14.8 reading is also a sign of a crapped out widebnd.

Err.. I never saw that as an issue. Do you have the wb hook up to the battery??

czilla · Apr 10, 2009, 03:57 PM

no just a reg power source

ktk · Apr 11, 2009, 06:14 PM

Quote:

Originally Posted by Femi

Sensor is mounted pretty close to the turbo's hotside.. Around 3-4 inches from it.

I think this right here is your problem. The Bosch sensor that comes with the AEM wideband is quite sensitive to high temperatures. Mounting it so close the the hotside would absolutley be a detrimental effect on the life of the sensor itself.

In a street driven or roadcourse driven car there is really no need to put the wideband so close the hotside. It can be mounted near the bottom of the downpipe just before where your cat would normally go. In fact, there is even a little rubber stopper in the floorpan of the passenger footwell that I suspect Mitsubishi engineers placed there specifically for routing wideband wires into the cabin!

The only real benefit I can see of having the sensor mounted so close to the hotside is that if your car is a fully dedicated drag and rides around on a trailer. The sensor will quickly get up to temperature and give accurate readings within the few seconds from when you fire up the car to when you line up the car and begin your run.

Accuracy wise, assuming that you have no leaks between your hotside and your cat, mounting it in the downpipe will be fine and extend the life of the Bosch sensor significantly. From what I've seen the NTK sensor is a bit more resilient when it comes to high heat, but its also significantly more money...

zbomb · Apr 11, 2009, 06:18 PM

How to check your AEM WB

Edit...Seen you saw this...I Sorry.I'm going to leave it up for anyone else.

Femi · Apr 16, 2009, 05:26 PM

Quote:

Originally Posted by awdboosted

If you do a WOT pull and it reads much leaner then it has in the past (ex. car is set at 11.2 and reads 12.8-13.00). The sensor is no longer any good. They go bad fast from what I’ve observed. My last 4-6month if I’m lucky. Sometimes the nlts or back fire cracks the porcelain and kill it. It’s a good thing they are cheap and I always keep extras.

Gotcha!

I'm considering switching my wideband to the NGK AFx with NTK sensor, but am unsure as to whether it would be more trouble than it is worth to start rewiring everything again... I would really like something that I didn't have to constantly change sensors out really...

I have a new sensor that I will give a whirl to see if it does any better.. I can't do it right now though due to some damage I sustained to the car a few days ago...

Let me ask this though:

1) Is there a decrease in the driveability of the car out of boost with a cruddy wideband?

As I mentioned, I was experiencing a bit of a hesitating during shifts as soon as I got back onto the gas.. Not sure if that was due to the wide band or not. This is during normal shifting, so the car was not really staying in boost between shifts so to speak.

It was not like that after I got tuned last summer, but seemed to show up as the weather cooled last fall and into winter, and as I got more miles on the wideband.

2) Is there any other way (apart from getting the shock of your life as the wideband reads leaner than usual in boost) that the wideband needs to be replaced? Are there any driveability-related issues one should look out for?

Or is an NTK wideband probably the way for me to go? The NGK AFx wideband allows for the ability to perform a free-air calibration, which allows you to maintain the accuracy of the whole system as it ages. Something the AEM does not allow...

Thanks everyone!

awdboosted · Apr 18, 2009, 07:46 AM

-With 02 feed back on you may see a decrease in gas mileage. If you want to see if it's your 02 sensor is going bad. Run a tank with out 02 feed back. That shoud take care of 1 and 2. Assuming you're not useing the 02 feed back to correct a very poorly tuned low throttle.

As far as the ntk senors. They last a little longer and cost a whole bunch more. From my observation. I just recommend tuning a car in 75 degree temp with a new sensor. (The temp correction suck on AEM. I've been starting to mess with it.)

Femi · Apr 18, 2009, 08:20 AM

Quote:

Originally Posted by awdboosted

(The temp correction suck on AEM. I've been starting to mess with it.)

Is this more an issue with the AEM-provided basemaps, and the unwillingness of most tuners to actually adjust them? Or is it more a problem with the EMS itself? I have noticed that a lot of tuners seem content to just play with the fuel, boost and timing, and leave as much as possible as it came on the basemap, assuming that AEM did their homework with regard to that aspect of things.

awdboosted · Apr 21, 2009, 02:03 PM

AEM just provides us with safe start up maps.. It will make your car run but it I've had to adjust all kind of stuff to make the car run better. I’ve observed as it gets colder outside the car run much richer. To make a proper map AEM would have to simulate all kinds of weather. AEM is out of Cali.. They don't understand cold weather. To them 50 is brutal cold, when in IL it get down to 10. So they just take a math equation and run with it. It beats start up a car with no base maps. So I thank them for work they put into the maps.