Wideband as Narrowband

[3400-95-Modified]

Right after a run too you are also going into deceleration tables, so its somewhat typical to see the WB drop wicked lean right after you get out of the throttle.

Mine does that on any deceleration, and I'm sure a WOT drag run will be more apparent too.

[Superdave]

lol.. i love when my wideband reads 99.99 when slowing down in gear at the track

[3400-95-Modified]

Ok I have my wideband installed... has been for a while, I've been using UVScanning to log the a/f ratio since I don't have a gauge..

Now here is my question. I log a a/d input, its 0-5v now I need to do a calculation to get that to a/f ratio... default in the program there was a wideband fake pid... the calculation was, %AD1% * 3.008 + 7.35, so I was using that before.... BUT I just found on PLX's website it says, Air Fuel Ratio = 2 * voltage + 10, or for me it would be %AD1% * 2 + 10

I just ran this through some of my logs... my last very chitty drag run, and it comes up corrected as 14.2-13.8 a/f ratio... when the logs were reading a 13.1 or so.... I know it was lean to begin with but holy hell this explains why the damn car only did a 14.8 and felt like a lead sled...

[sprucegagt]

Don't forget, you still have to correct for the voltage offset if your using the AVT hardware to read the wideband output.

[3400-95-Modified]

How do you figure out the voltage offset? Since I am using the AVT unit.

And then which part of my formula does that affect?


Found a little more info on it but I need to see if I can set my PLX to go full lean or whatever to get the offset... I can also use just a 5v input from a DC power generator right?

[sprucegagt]

I have the notes for this at home. I'll post it up after work.

[3400-95-Modified]

Found this on the PT website, but this applies to the LC-1... I just need to figure out how to make the PLX do a 14.7 aka 2.35v readout OR a max 5v readout to get my offset.

Search is your friend. Foghorn had a good post about using the LC-1 with the powertuner.
Here are my notes on the matter that I copied from his post.

The formula 'out of the box' for an LC-1 with stock parameters is;

(N0/51) * 3.008 + 7.35

Where does 3.008 come from? Simple, it's an expression of the range. Stock the LC-1 provides for an AFR range of 7.35 to 22.39 over a voltage range of 5 v, so.. (22.39-7.35)/5=3.008.

If, as you must, check for voltage offsets then you may have to change the values. The LC-1 can be programmed to output a specific Voltage which you can then measure with a voltmeter. In my case I had only 4.9 volts as opposed to 5 so this leaves me with (22.39-7.35)/4.9=3.069

Where does the 7.35 come from? Simple, at 0 volts the AFR would be 7.35. In the case where the AFR is richer (lower numerically), the LC-1 in stock form will not read lower, but in all cases the minimum AFR value must be added to formula to show the correct value.

If you can adjust the sampling rate of the O2 sensor then I would suggest you make it as slow as possible. The LC-1 is capable of 'infinite' sampling rates but typically can be 300 samples per second or more. At that rate it's capable of reading variances from cylinder to cylinder and in fact the guage would be constantly changing. If you can slow it to 1/3 rd of a second or thereabouts that's what I would suggest.

I have simultaneously run the PT with the LC-1 connected while I also ran the Innovate Logworks program just to compare the results. Why would I do that? Because the serial output from the LC-1 does not suffer from any ground offsets, meaning what you see is what you get in the Logworks so you can compare to the 'adjusted' value displayed in the PT.

I can tell you, and Jerry has seen it too, that there can be a substantial difference between the PT and Logworks, up to 0.6 AFR or more in some cases. There are a couple of things you can do to correct that, both include manipulating the formula used in the PT.

The first method is to compensate differently in the formula for the ground offset. Some suggest measuring the maximum voltage output of the LC-1 (free air or AFR22.39), in my case it was 4.9V, then adjusting the formula like so;

((N0/51)+0.1) * 3.008 + 7.35

I don't like this method because the ground offset is only compensated for perfectly at very lean conditions. Alternatively you can just program the LC-1 to output 2.5V, which is the middle of the scale and equal to 14.7 AFR, and then read the voltage on the analog output and then use that value in the example above.

The second method, and the one I prefer, is to manually manipulate the formula range value in order to a result that closely matches the readout in Logworks. This can best be done with a co-pilot and a lot of trail and error.

In my case, from memory, I think I ended up with;

(N0/51) * 3.039 + 7.35

I found this method to be most acurate over the range that interested me most, 9.5 to 15 AFR.

[sprucegagt]

Yeah, that's it. The only thing different is the offset does not change due to AFR. It's constant. So set your Wideband output to a set number (i.e. 2.5 volts) and then see what UVscanning is telling you. Take that difference and add it to the forumla for the PLX.

[3400-95-Modified]

I add it or subtract it from the %AD1% number correct?

So it would look like ((%AD1% +-X)*2 +10)

Also on just open air my O2 should read 14.7 if I'm not mistaken, and I turned the key on started the scanner and the voltage in UVscanning was bouncing from 2.35-2.37 or so, which is pretty darn close.

[3400-95-Modified]

Well I tested and found it Formula is now ((%AD1%-0.02)*2+10) so I only had .02 positive adjustment to remove... Not too shabby if you ask me... that only equates to 0.04 a/f difference.