Aftermarket Wideband O2 Sensors
 

I've been having that issue with my MTX-L Plus where it throws an E8 code after a couple of months. 1st sensor was the one that came with the gauge, 2nd was an aftermarket "AA-Ignition" sensor, and this time around I went with an "Automotive Leader" o2 sensor and threw on a copper plate heatsink/shield. All plug and play and all, just this time I noticed the gauge itself was reading a little richer than the reading in Tuner Studio. So I just used the "Custom Linear WB" calibration setting in Tuner studio and got them both pretty close. Was this the right thing to do? Any dangerous things that could possibly happen from the recalibration or use of non-Bosch sensors?

Is this the first time using a heat shield on the o2 sensor? Where is the o2 bung located? Yes using custom calibration in TS to get AFR's to match is correct.

When my MTXL started having E8 issues my solution was switching to an AEM uego

Yes it's my first time. I just got a copper plate and drilled a hole through it and placed it in between the o2 sensor and the downpipe. Bung is 12 o clock right before the cat would go. It's the Kraken 3" T25 downpipe.
https://cimg2.ibsrv.net/gimg/www.mia...62b413f389.jpg


Yeah I'm hoping this won't be my fix, I don't feel like dropping $200 on something that I already have basically. :bang:

That heat shield looks good. I also only use official Bosch LSU sensors.

I smoked three genuine 4.9 sensors on a LC2 in the stainless kraken downpipe in less than a season. My modding to make it fit MKturbo exhaust doesn't leave much room for a heat shield. Besides that, the LC2 is slower than snot canbus or not. I finally gave up, switched to a 14point7 S3 ADV, and I went back and added a relay triggered by MS so it's never running if the engine isn't. I've made it further on this one 4.9 ADV sensor than I ever did on a LC2 with 4.9 sensors.

My old MTXL 4.2 in a mild steel downpipe around the same distance from turbo never had a single problem, it threw E8 once or twice but never came back after a restart. I'm also helping build a j35 with a 7163 in the back of a classic mini, it smokes LC2 sensors like they are going out of style. Hard fitting stuff in the space available, but it's a stainless downpipe too...

What's weird is I've also got a few other non innovate 4.9 controllers, some I made and some more 14point7 stuff. They all use the "burnt out" E8 coded sensors just fine, and they all converge within 0.3-0.5 points.

I don't much care for innovate stuff any more. Not any faster, much more sensitive to sensors.

I have had much better luck with 4.9 sensors over the old 4.2. I have also had better luck with the AEM over Innovative. Now I just use a Haltech Canbus Wideband.

Saying that I have had a few customers get E8 codes or have 4.9 sensors die early with AEM products.

Glad it looks good, we'll see if it does anything lol. I was tempted to grab Bosch sensors but I couldn't get myself to spend the extra $30 for something that's probably gonna get burnt out again.

I think I had a sensor crap out on me even when the car was stock and naturally aspirated right before I put in my turbo stuff. Does running the wideband without the engine running help burn these things out? Is it the temperature shock issue?


Is it the sensors that are the issue? Or is it the wideband units? I've heard mixed ideas as to whats the reason for all the E8 codes and whatnot.

From lots of reading I guess innovate runs a proprietary non standard protocol on a bosch sensor. It's supposed to be more sensitive, faster responding, somesuch like that. It does not use a bosch controller IC. Most other widebands do use a bosch IC, like most AEM stuff from what I've read. Since my dead to innovate sensors work on bosch/other controllers, they're just extra sensitive.

Whenever the wideband is on it's powering the heater because it needs heat to work. Most dump a lot of current in really fast so they can respond as fast as possible. If any condensate, maybe even lots of e85, hits the sensor it could crack and destroy it, and there's a lot on an engine cranking over for the first time. I guess even the cold air moving is enough to hurt them, they get smaller and smaller with each generation. Bosch specifies a condensate period/delay in the ramping of heater current, up to the controller to implement it properly. Powering it yourself is a surefire way to make sure it happens properly and at the right time.

http://www.bosch-motorsport.de/conte...or_LSU_4.9.pdf

The heater period is on the bottom left of the second page if you are curious.

Ah ok. So I guess I'll have to integrate some delay timer or switch then. Are there any common practices people here do? Or is just kinda free reign and do whatever that works.

I have talked to some workers at Bosch by pure luck. A lot of the LSU sensors are built here in SC. One of the engineers I talked to went into tons of details about all sorts of info on the LSU 4.9 sensors. What I ultimately took from it all is China is copying them, and to only buy Bosch. Same problem as with the RX8 injectors.

I use a GPIO with a seconds and RPM trigger. It's almost perfect, but megasquirt does not count time since engine running, just total on time. IOW, you won't get a delay if you key on and wait too long to start, but it will still never run if the engine is off. Maybe in a future firmware version we will get an engine running counter...

Yeah I might try out a Bosch sensor if this sensor I got craps out again.

Interesting. I might give a switch a shot. How long should the wideband wait until it gets turned out though? A couple of seconds?

I have mine set to 30 seconds, ends up being ~20 seconds after waiting for the pump to finish priming.

Alright cool. I'll look around for delay circuits or figure out something with the MS outputs or something. Thanks!