Toyota Celica Supra Forum banner

1 - 16 of 16 Posts

·
Registered
Joined
·
1,266 Posts
Discussion Starter #1
I just purchased an Innovate MTX-L wideband kit to see what's happening with my AFR. I have a Pacesetter header with the O2 bung welded at a 3-into-1 collector. If I'm reading correctly I can use this bung for the Innovate sensor and use one of the analog outputs to signal the current OE ECU (rather than keeping the current 02 sensor and adding another bung).

What I'd like to find out is the OE signal range that the ECU is expecting, and there are two analog output options available on the MTX-L:

1. 0 - 5v range 0=AFR of 7.35 5=AFR of 22.4

2. 1.1 - 0.1v range 1.1=AFR of 14 0.1=AFR of 15

Does anyone know this information?
 

·
Registered
Joined
·
3,295 Posts
You want option 2. If it is like the older models you should be able to hook it up to a laptop and change the settings or data log if you are into that.
 

·
Registered
Joined
·
491 Posts
Revvhappy, I am searching out that information. Something interesting I found when researching that problem; the Toyota service manual indicates in one of the factory tests that if the O2 Sensor is indicating an excessive rich condition, that the problem is typically a symptom of a failed PCV system. I NEVER would have thought of that, but I suppose its possible; but that's assuming a totally stock engine and surrounding components.

Still looking for the info if I can find it.
 

·
Registered
Joined
·
3,295 Posts
We use the unheated narrowband type of sensor, a fairly common one which is the bosch 12051. It also comes with an adapter flange which is good for using wideband in the stock location.
http://www.summitracing.com/parts/bch-12051
http://www.oreillyauto.com/site/c/detail/BOS0/12051.oap?pt=03860&ppt=C0018

They all generally have the same output range
http://www.gonefcon.com/whatifdyno/theory.htm

I've got the older version with no gauge sitting around i've been meaning to hook up forever:
http://www.celicasupra.com/forums/showthread.php?73534-84-p-type-up-keep-and-basic-mods&p=907009&viewfull=1#post907009

Here are a couple relevant older posts i link to there, 0-1 is what you want.
http://www.celicasupra.com/forums/showthread.php?50656-Innovate-wideband-to-output-OE-O2-sensor-signal-instead-of-stock-sensor
http://www.celicasupra.com/forums/showthread.php?56462-lm-1-lc-1-gauge-options
 

·
Registered
Joined
·
1,266 Posts
Discussion Starter #5
Thanks, Gamble! After my search last night I was fairly confident that it was Option 2 but wanted confirmation. For the moment I will only be datalogging to get an idea of the current AFR conditions. I wonder if the Innovate software will let me "fool" the ECU to tune for leaner running at WOT. I'm quite interested to see what happens when the ECU goes "open-loop" on the track. Next Friday will be a very interesting day at Road Atlanta; I'll have datalogging of AFRs, video, and GPS data. If I was tech savvy I'd find a way to combine them with a track overlay on the video including AFR readings.

It may not matter for too long as I'm thinking about adding a standalone system with a MAP sensor. I really only want it for fuel mileage, but if I get a little more power that's OK too :)

Khlause: My engine has been simplified for endurance racing, so it has no PCV, EGR, vacuum lines, or coolant lines (for the intake). I could see how a malfunctioning PCV would affect the AFR readings - adding oil (more petroleum product) into the combustion process would result in a "rich" condition.
 

·
Registered
Joined
·
1,266 Posts
Discussion Starter #6
I installed the Innovate MTX-L today using the 1.0v - 0.1v output wire spliced into the OX wire of the ECU. Everything is working fine, the photo is at fast idle and not warmed up. I took the car for a test drive and all was good for about 6 miles. The engine was up to temp, the check engine light had come on earlier with no change in AFR at street speeds. I did a short full throttle burst on an empty road (2nd gear) and all of the sudden the AFR jumped to 18 and wouldn't come down.
Any ideas?

** the gauge is mounted temporarily using a piece of coroplast, it was easy to drill and slide into this space.

 

·
Registered
Joined
·
3,295 Posts
That's odd, but might be normal behavior for all we know. I know the later gens i've seen discussions of guys confused as to why the car goes full rich sometimes but it is expected under certain conditions apparently?

Do you have two o2 bungs? I vote you toss the stock one back in as before, and have this one set up completely separate and log things to see if it acts the exact same way. Nice mounting, and big ups for getting it done so fast. Looking forward to seeing your progress.
 

·
Registered
Joined
·
1,266 Posts
Discussion Starter #8
I only have one bung. AFR of 18 suggests very lean, which I don't understand since it still seemed run fine. I'll try again next week when I return from a short trip.
 

·
Registered
Joined
·
3,295 Posts
Oh duh brain fart, yeah full lean can be when you just let off the gas as the injectors shut off.

Did you calibrate it? Either pull the sensor out, or just let it sit a day if you want to calibrate it still attached to the car, don't want to mess up the readings.
http://www.innovatemotorsports.com/resources/whycalibrate.php

Where is the bung? If it is on one of those pacesetter headers it could be reading super lean as clean air is getting back up in the extension tube between exhaust pulses.

Looks like if they read way way off sometimes the sensor itself is bad. You can test it out with a torch of about any kind as well to make sure it is at least seeing the changes.
https://www.youtube.com/watch?v=2RfzzuTJUW8
https://www.youtube.com/watch?v=o6iXNBPPhwc
 

·
Registered
Joined
·
1,426 Posts
Yeah, the sensor location might be an issue depending on where it is and how it's mounted. And with this in mind, I wonder if the sensor was really getting hot enough to work properly in both conditions. Is that particular data logged? If not, it might be something you could check with an IR meter, which should get you close enough data to know. Maybe it's heater is AFU/NFG. And maybe it's not getting good samples in some way.
And like Gamble also stated, it should be calibrated when you're sure it's operating OK otherwise.
The 18 reading wouldn't surprise me at all when letting off the throttle, but picking it back up a bit should very quickly return it to normal to slightly rich. And by the time it returned to idle it should have settled back to normal too.
Is it possible that it's holding a min / max reading in the display or something similar?
What fuel are you running? WilliamB82 pointed out that 14.1 would be in balance with E10. With that in mind, can you input an offset to the sensors readings to compensate? That way it could fool the ECU into thinking that was 14.7 and it would adjust accordingly. That might free up more power and / or mileage?
I should RTFM again. But knowing the sensors actual location, etc. would be useful.
 

·
Registered
Joined
·
1,266 Posts
Discussion Starter #11
My Pacesetter header has the bung welded into one of the 3-into-1 collectors. I'm running 93 octane E10. The weird thing was the AFR looking awesome for 10 minutes at operating temp (13-15 at various throttle openings), then something told it to go lean. I'll get more data (including a log when I borrow a computer with a serial port) next week. I'll also mount a Gopro aimed at the gauge to see the exact moment the AFR freaks out.
 

·
Registered
Joined
·
3,295 Posts
That sounds like a great plan, i was hoping to do the same thing myself as far as pointing a cam at the cluster to log that info on the cheap. I was actually thinking of using a webcam hooked up to the laptop that is already logging over serial though, and screen capturing things side by side for maximum lazy.
 

·
Registered
Joined
·
1,266 Posts
Discussion Starter #13
After another drive today I'm a bit concerned. The AFR started at 11 when the engine was cold, then progressively went leaner and leaner until the gauge was maxed out at 22.4 (high limit). The coolant temp never rose and the car seemed to run fine and idled perfectly. The reading would change slightly depending on throttle application, but even full throttle would barely bring it under 22 at the end of 10 miles of driving.

At one stoplight I switched the ignition off and back on again which made the CEL shut off but the AFR remain constant. Any ideas what would tell an ECU to "go lean".

GoPro video available...
 

·
Registered
Joined
·
3,295 Posts
Toss the stock narrowband back in with a gauge and see if it acts the same? Anyone local have a wideband you can borrow to see if the reading matches yours? Apparently a failure sign of wb sensors is that once warmed up they just about max out lean, if your car was running that lean you would hear/feel it. Have you thought about wearing headphones to listen for knocking? Just google diy knock detector and you'll get a ton of hits. http://www.nicoclub.com/archives/diy-knock-detection.html
 

·
Registered
Joined
·
1,266 Posts
Discussion Starter #15
The more I read about the Innovate products the more I'm inclined to believe it's a faulty sensor and not my car. I'll give them a call Friday and see if they can help.
 

·
Registered
Joined
·
1,266 Posts
Discussion Starter #16
Update with good news: I replaced the exhaust gasket even though it didn't seem to be leaking, then recalibrated the O2 sensor. 15 miles later the system seems to be functioning as intended. AFR at idle and cruise are 14-15, and drop to 12-13 at WOT.

Can I say for sure that it wasn't the exhaust gasket? NO, but the gauge reacts differently that it did previously on start-up. It used to go straight to an AFR reading, and now it takes a 5-6 seconds reading Htr (heater) before I get a number.

The Check engine light came on as usual after 10 minutes of driving, but had no effect on the AFR. Now it's fun to watch what's happening in real time. The off-throttle popping is definitely a lean condition.

I'll have more data after the track day on Friday. It will be a bit chilly (40-55 degrees) and I hope to have a computer the car to datalog.
 
1 - 16 of 16 Posts
Top