Air/ Fuel Ratio gauge

[liquidgroove]

1st, I am a noob, so don't shoot me.. ;)

2nd, I have a narrow band A/F ratio gauge from Westach. The input singal is damped, so the needle doesn't dithering around like the LED ones.

But overall, I think narrow A/F gauge is not that usefull for fine tune mixture. Even Westach's gauge give you mixture ratio, but still difficult to fine tune with it.

[SamE30e]

Nevarrrrr. My project took another route. I'm finding a clean 325is and making a monster out of that. The e will be the beater, just chip, exhaust, wheels brakes and stereo.

[DaveSmed]

Ok cool were putting the narrowband in the geo then. :pimp:

You say it like there is a working O2 sensor on that car....

Put the LC-1 in it! Or even better yet, give it to me!

[crxb16t]

If you like it, it's fine.

Ok cool were putting the narrowband in the geo then. :pimp:

[SamE30e]

I did put it back to where bosch had it. Like I said, I got it from the wreckers and it was messed with, so I just fixed it.

[DaveSmed]

I like it.

There ya go, thats what matters. It is useful to get a rough idea what mode Motronic is in, like acceleration enrichment when it pops in open loop.

However, knowing the limitations I mentioned above, you can also see why it is not helping SamE30e out with his situation, and he needs to actually buckle down and diagnose what is wrong with his car, like a stuck or clogged injector.

Sometimes it can be the opposite of what you think too, my car had a leaking injector, and I was running too lean. It wasn't leaking bad enough to where it was dumping crazy amounts of fuel, but it was enough that Motronic saw there was too much and pulled back the trim tables. This made 5 cyls run too lean, and one run too rich as it found an average that kept it happy. I could get the car to run great for about 10-15mins by unplugging the DME and resetting the learned values. A new set fixed it.

[delatlanta1281]

Mine does just that, when I let my foot off the accel, it goes red for two seconds the bounces up. When I push on the accel, it goes blue for two seconds and then bounces down. On the highway over 2000 rpms it bounces right in the middle.
I like it.

[DaveSmed]

If you like it, it's fine. It is unfortunately useless as a diagnostic tool to figure mixture, yet provides some neat insight into what DME is up to, like when it decides to pop into open loop when you floor it for example. Then it is showing a rich condition (hopefully!) and when it isn't bouncing around the DME can't use it. Kinda cool.

If your immensely bored and have an old chevy (pre 96) usually connecting pins A and B of the diagnostic connector with the vehicle running will give you a similar picture via MIL flashes. (engine off is when it flashes the codes)

[delatlanta1281]

Whatever.
I like my guage. I think it works well for what I use it for. I don't plan on tuning my car of of it. So it helps me to know how rich or lean my ecu is adjusting to cope with potential problems. Nothing wrong with a narrow band guage if you are using it for that is there?

Oh, and "tuning" your AFM is stupid. Don't make me explain that one too. Put it back to where Bosch had it, they guy that put it there knew a bit more than you and me. Work on finding your REAL problem.

+1

[DaveSmed]

...sorta kinda.

It actually works BY swinging wildly from rich to lean at the DMEs command.

There are a few different parts to your catalyst, that take care of different pollutants and work in different ways. Simply put, one part works better with a higher concentration of oxygen in the exhaust, and the other works better with less oxygen in the exhaust.

Given that stoitch is NOT the optimum mixture for a few different scenarios, the DME also needs to know an approximation of what the mixture is when it is adding fuel for say a cold start. It also needs to tend to the catalyst.

How it does this is by varying the mixture a small amount multiple times a second. If you had a graphing scope, the oxygen sensors output would resemble a sine wave. That way both parts of the cat are happy and able to work efficiently by alternating between rich and lean, and it allows the computer to plot where it's corrections have placed it by allowing it to time the peak to peak intervals between the sine waves. Longer interval above stoich, richer mixture.

Enter the gauge in question, and this goes for a digital multi meter without a graphing feature too. A momentary display is crap for reading any damn thing unless you sit there with a stop watch, have incredible reaction time,
and plot a graph to notice the peak to peak differences. The peaks themselves do not change, its the timing between them, and before passing the stoitch line that changes. A digital multimeter with an averaging function and decent response is useful for this however.

Now, say you for some reason think your running rich (using lots of gas, we'll go with that one) When the DME is adjusting the mixture based on the o2 sensor values, it stores correction factors called fuel trim in its RAM. These values are immensely helpful in diagnosis, as it takes all of the computers interpret of the numbers, and shows what the computer decided to do based on that information. That value is based on a complete picture of the DMEs operation parameters, including the times when it decides to resort to hard coded tables (open loop) for engine operation, for say a cold start (when the o2 data is inaccurate) or for WOT (where the o2 data is inaccurate, too slow response time, and the DME already knows what it wants to do for maximum output) Interestingly, in a way, the o2 does still affect open loop, as the fuel trim values learned from the sensor are still applied to the final calculation.



Oh, and "tuning" your AFM is stupid. Don't make me explain that one too. Put it back to where Bosch had it, they guy that put it there knew a bit more than you and me. Work on finding your REAL problem.

[SamE30e]

Which is what I just did..

[nando]

sure, it's rather simple. it's an oscillating 0-1v signal with .5v being stoich (a wideband is typically a linear 0-5v). the ECU tries to keep the engine running as close to stoich as possible - and a narrowband only outputs rich, stoich, or lean. so if it reads lean or rich, it will adjust the PW until it reaches stoich - but the ECU has no idea how rich or how lean, only that it isn't 14.7:1. and of course above a certain load (about 70%) it switches to open loop mode and goes off the VE table only. it's pretty reliable and works fine for a static tune/stock engine.

[DaveSmed]

Does anyone actually understand how the narrowband input into the DME is supposed to work, what kind of signal it is, and how the DME uses it?

[delatlanta1281]

yep, I just read it on the internet. I have no idea what I'm talking about. you can totally "tune" your car with a narrowband sensor and turning the wheel on the AFM - that's how all the pros do it anyway. :roll:

I understand that Nando, but in my experience, the ss meter helped bring to light a potential problem.

[SamE30e]

yep, I just read it on the internet. I have no idea what I'm talking about. you can totally "tune" your car with a narrowband sensor and turning the wheel on the AFM - that's how all the pros do it anyway. :roll:

I'm well aware that you can't tune your car with a narrow band, otherwise I wouldn't have the LC-1 with my MS that i'm putting together. I used it for a ROUGH idea of the problem. And the problem was I have a AFM off a car from the yard and the cap wasn't sealed on it, so obviously someone had tampered with it. So I wanted a rough idea as to what the problem was.

Did it work for me? Yes
Does it make you not sleep at night? I sure as hell hope not, but sure seems like it does.