Darkelvis

First off, I know I need a wideband 02, I know narrowband is ----, I don't want to start this arguement again, I don't have the money for a wideband right now.
89 crx si, 14b, DSM 450s, TurboEDIT, EcuControl datalogger (cable installed by xenocron), autometer a/f gauge.

Went for a datalogging drive last night and something really messed me up. I wasn't expecting the a/f gauge and the datalogger to tell me the same a/f but at many times they were doing the complete opposite thing. When the a/f gauge would go stoich/lean the datalogger would read rich, when the a/f gauge would spike rich the datalogger would be sitting in the middle of stoich. Sometimes they would both be stoich or rich or stoich/lean (but never just lean), but other times they would just keep doing the complete opposite of each other. For a pretty fair chunk of the drive the gauge was reading stoich/lean-stoich/rich but the datalogger wasn't even close to the gauge's reading.

Anyone have any ideas what this could be? Shitty gauge? An error in the program? Something I'm not thinking of? Any and all responses are greatly appreciated (except for "narrowband is useless get a wideband")

Chris Harris

The problem is, with a narrowband...the ECU is constantly trying to balance. Going back and forth from Stocih/Lean <--> Lean/Stoich

And its such a small window and the O2 sensor is sometimes so far down the pipe, there is a delay and you really dont know what you are running.

TurboEF9

I've seen the code for the O2 sensor in ECUControl... I wouldn't rely on it for narrowband.

It's a linear formula applied to a non-linear voltage input line. Not reliable at all. In this instance, if you insist on using that ricer light show gauge, then I would say use it over the datalogger. At least that is calibrated for voltage... even though it too won't measure what you're looking to measure. Which would be a target AFR of ~12.1:1 in boost.. but hell, what do I know.

[rant]
On a side note, why the hell are people so ------- cheap when it comes to Engine Management? We give them free software to do what they wish to thier car, they get endless free support for thier problems and questions...

Would you run stock injectors on with a tuned ECU on 9psi? No, everyone knows that's stupid... and you buy larger injectors.
Would you run a turbo with 1.75" exhaust? No, everyone knows that stupid. So you buy upgraded exhaust.
Why would someone who spends all of this money (most turbo setups costing around a $1000 depending on how cheap you go), and time.. to waste it all and skimp on tuning tools?

The only things we ever ask anyone to buy is a chip burner, and a wideband. Everything else is either free, or you don't "need". ::sigh:: I'll never understand.
[/rant]

Darkelvis

[rant]
Not to be a dick, but it isn't my car therefore it isn't my "ricer light show gauge" and who said I was being cheap? Right now I'm saving up for a chip burner, ostrich, and wideband, don't believe me? Ask Xenocron about it (the burner/ostrich). Sorry that some of us don't want to spend the money on a car that isn't ours and we have to wait for other people to come up with the money first. I appreciate the free software, I appreciate everyone that has helped me here, and I appreciated you taking the time to read my question and offer up an answer, but I don't appreciate being treated like ---- when I'm asking a question that to me is legit. All I am trying to do is learn, just like everyone else had to do here at some point in time.
[/rant]

I'm not trying to tune the car with the datalogger or the gauge, I just wanted to know what the problem was. I just figured that while they wouldn't be the exact same when the gauge moved towards rich the datalogger would also move towards rich, when the gauge moved towards lean the datalogger would also move towards lean. Guess I was wrong.

TurboEF9

Yes, you are.

For the nonlinear plotting you'd have to create a 2-d array that contains preloaded voltage points and thier equivelant AFR. This has not been done, nor do I think Nick a ---- to do it since Narrow band is a waste of time. And why I "use your ricer light show" is because this is what your A/F guage does, only with hardware not software. For each voltage reading (assumed A/F ratio) an LED will light up. You normally get more than one because the voltages are so close together..

A) You didn't say it wasn't your car
B) You said you were "datalogging air fuel ratio" when you saw all of this happening
C) And if you know narrowband is useless, why the hell does it even matter what the readings are? It's *USELESS*, as in "having no use".

raprazant

jeez relax man. I think he's asking a legit question. I would like an answer to this as well. But xenocron's reply really makes sense on why it could be wrong and not showing the same numbers (datalog vs. af gauge).

TurboEF9

I have looked at ECUControls code and I stated _exactly_ why it's off.

As far as "relaxing", no one is uptight but you and Darkelvis. It's OK dad, he can handle himself.