Iowa999

iTrader: (1)

My interpretation of that quote from another thread is that there is some misinformation on the internet. My deepest apologies for the shock if that comes as a surprise to you.

If you are no longer interested in the question of whether you need an aftermarket wide-band to tune your car, instead of yelling at people who try to explain to you why the answer is "yes," maybe you should start a new thread with a different title.

No matter what you decide to do, please - I beg you - stop saying stuff like this:
The "factory o2 sensor" on an Evo is a standard narrow-band. As such, it does not produce a current; it produces a voltage between about 0.2 and 0.8. Only a wide-band involves current.

todd6027

iTrader: (10)

no it cant

justastockevo

iTrader: (1)

a sensor not producing a current? how's that?

ok, another question, perhaps, more simple. when ECU throws voltage errors on 2 banks of oxygen sensor, where does it take it's data from? besides, which MUT requests are correct for logging o2 sensors?

http://evoecu.logic.net/wiki/MUT_Requests

here i see listed requests 13, 3c, 3d, 3e - all connected with some sort of o2 sensors. i remember that in older definitions there was request 1f, which was supposed to do with o2 sensor too. which is which?

Iowa999

iTrader: (1)

From your comment about sensors not producing a current, I infer that some basic information is missing.

A narrow-band has a single plate. The difference in O2 concentration on the two sides produces a potential - i.e., a voltage - that varies between 0.2 and 0.8 vDC, with a curve (as posted above) that is nearly a step function with an inflection point at a tad less than stoich and 0.45 vDC. The amount of current this sensor could produce if allowed to flow is tiny and irrelevant.

A wide-band has two plates with a "dead air zone" between them. The first plate is set up like a narrow-band. The other plate is powered by a regulated supply, the current to which is constantly adjusted to get the first plate to produce 0.45 vDC. The amount of current that must be applied to the second plate to get the first plate to 0.45 vDC tells you how much O2 was in the exhaust, since the amount of current determines the free O2 in the dead air zone. This is why wide-bands operate in terms of current. But not in terms of a current produced; it's a current required.

Do you see now why you can't just slap a wide-band on our ECU and be done with it? A narrow-band is a pretty standard sensor - (well, one of the two simple standards) - in that it produces a very simple signal: a voltage. But a wide-band is hugely more complicated, requiring a controller to produce a known and regulated current which can later be converted to a more-standard signal, such as a voltage up to 5. The key, however, is that there's nothing in an Evo's ECU that can do what a wide-band controller does. It isn't a question of MUT codes or software, etc. It's the lack of required hardware.

edit: I should have said that I'm describing how narrow- and wide-band Bosch-type O2 sensors work; there are other methods, but I highly doubt that they are used as original equipment on any "normal" cars like Mitsus and Scoobies.

justastockevo

iTrader: (1)

please... i know the difference... i know the sensor cannot produce any current, but it surely alters it, as it creates resistance, any sensor. the difference between the two is that wideband's sensor current needs to be measured. i know Ohm's law too, no need to post, thanks.

rather, please anyone tell me what can we log off our o2 sensors and what we can't? i've seen a post in the same thread, from someone under username Zeitronix, who told that ECU measures heater current, can't we log sensor's current as well?

Zeitronix

iTrader: (2)

There is much more into controlling a wideband O2 sensor than a source code. Quite a bit of dedicated circuit is required to DRIVE the wideband O2 sensor. Precise heater control is another issue all together. A simple "switching off heater off at high EGTs" is not going to work.

Iowa999

iTrader: (1)

I give up. Good luck.

Zeitronix

iTrader: (2)

justastockevo

iTrader: (1)

glad that you have joined, i guess you would have more first hand experience than most of the people

i read specs on denso 234-9xxx series sensor (the one that's installed on subarus), sounds easy to me - one wire is reference voltage, the other is signal, in terms of current. i thought our sensors heaters are same pulse-width modulated as denso's 234-9xxx, both turn off when they reach operating temperature and turned on when it's drops lower, am i wrong in this assumption?

and just for a sake of clarification, do you believe that our ECUs check those sensor circuits for current or not?

Iowa999

iTrader: (1)

Must resist....

The only difference between how Denso does it and how Bosch does it is this: while a Bosch wide-band applies the current to a second plate and the amount of current required to maintain .45 vDC in the first plate is measured, the Denso applies the current to the same plate from which the voltage reading is taken. But in both cases - as in all standard wide-bands - what provides the needed information is the amount of current that must be supplied to maintain a certain voltage. So, you can't use a Denso wide-band any more than can use a Bosch on our ECU without significant hardware as well as software changes.

Note, also, that the Denso approach only really works between about 11:1 and 19:1, while the Bosch works for any ratio. The Denso approach is also slightly non-linear. Fine for setting the fuel trims for closed-loop; not so good for open-loop tuning.

On the positive side, the Denso approach only requires four wires (including the heater) and doesn't require anywhere near as precise temp control inside the sensor.

On the negative side again, the fact that the Denso only has four wire can confuse some people as to how it works.

justastockevo

iTrader: (1)

have you checked bosch specs on their LSU sensors? they're as accurate as denso's in terms of AFR range and produce virtually the same non-linear curve.
LSU sensors are used by VW to set their closed-loops, while denso is preferred by japanese manufacturers mostly.

regarding hardware modifications, we know from fact that ECU checks for resistance on our heater wire, why is that we cannot log it on our sensor wire?

all this should make it trivial, in case we know our ECUs more.

Iowa999

iTrader: (1)

Maybe a Denso is as good as the Bosch. Not really important. Both require a controller.

OK, so now you are logging what is, effectively, the temp of the unit. And this is helping you to do what?

Speaking of knowing our ECUs more, have you figured a way around the core problem of our ECUs not having anything close to the required controller for a wide-band? Personally, I'd deal with this first and worry about temp control later.

Edit: in case anyone is having "issues" with my sliding back and forth into sarcasm, when I say "deal with this first" what I mean is "face the fact that this ain't going to work without major hardware and software changes"

justastockevo

iTrader: (1)

let's assume they're as accurate for measuring AFRs so far.

this only helps with unit temperature, right. but why are you all guys that sure the sensor's circuit can't be measured in the same way? then, having a reference voltage and a resistance one can easily have current from that circuit, effectively, making a wideband controller. what else i'm missing here?

andrewzaragoza

iTrader: (6)

the thing is that our narrow band hardware is simply not capable to give the resolution needed to properly tune our cars.

if our narrow band sensors were enough to give the correct afr wouldn't you think the companies like innovate would simply use a narrowband in their kits.

justastockevo

iTrader: (1)

let's say we replace our sensor (no matter which, front or rear) with denso 234-9xxx series. can we log it with our ECU? this way we would eliminate the controller. suppose we can modify software the way we want, while would like to keep hardware as simple as switching sensors.