Discussion and questions related to the course Understand AFR
Is there a way to tell if your wideband ls4.2 sensor is not accurate? Ive been told to change them regularly because the best way to know is to compare the readings to a new sensor. If the sensor reads the afr or lambda correct to the values entered to the software (7.35 @ 0v and 22.39 @ 5v) would this be enough to say the sensor was good? I display AFR and Lambda on the same screen with 14.7 as stoich. if the lambda and afr match up would this be good enough to say its good. I tuned my race bike and found it liked to run around the 13.5 - 13.8 afr. I raced it for abit and goes great. I then went to another race meeting where there was a well known tuner (im quite green) he tuned my bike - It showed 14 - 15 afr. he then ran it down to 12.5 afr - didnt gain anything. He said my sensor must be crook. however over that weekend the bike ran like crap so when I got home I put my tune back in. But now Im doubting my sensors and I dont know enough to ask the other guy about calibration or accuracy.
This is why VW, BMW, and a host of other OEMs took Bosch to task over the 4.2 and the 4.9 was designed and introduced. I've had several 4.2s fail in the same fashion as you're describing in a rotary environment, and even the controllers do something similar so it's not always traced back to the sensor itself.
Also remember is that the 4.2 was developed as not only a budget sensor, but a budget controller/heater circuit combination for OEMs to start integrating them into street vehicles for emissions control, not motorsport. Thankfully the 4.9 is here now but its still value engineered to be attractive to OEM pumping out 100s of thousands of vehicles a year.
Last but not least to answer your question: I always keep a spare 4.2 on hand that never gets installed permanently into a vehicle as a measuring stick. Seems to be the only way I've come up with the not only be accurate, but alleviate any fears of accuracy on a vehicle I'm working on.
Ok thanks. I do have both sensors 4.9 and 4.2. they have done not lots but a bit of work with rich, lean, 2 stroke and a bit of methanol so I was hoping there was a way to test them. I didnt know 4.2 was a lesser sensor. So a 4.9 is the recommended sensor for tuning?
In my own experience I generally find that the reading from a wideband sensor doesn't tend to slowly drift over time. Instead it will be accurate and then all of a sudden fail completely. If you've got a sensor installed alongside a stock ECU that still runs in closed loop mode at cruise then a good sanity check is to make sure your wideband is in fact reading stoic in closed loop operation.
Word, with regards to the 4.2 the reference cell gets contaminated with exhaust gasses at sustained 1600ish exhaust temps (from my experience, manufacturers may say less or more) pretty quickly. Everything is a OK and then no go. Other than that the 4.2 is completely fine and accurate for tuning in my opinion, it's just fragile (comparatively).
I've yet to see a 4.9 fail from anything other than getting smashed by a rock during an offroad excursion at Willow Springs or LONG multi season use.