Why did VL400 use x/10 for all the wideband equations in 12P

[Charlescrown]

Trying to get the TunerPro logging with 12P to align with my wideband gauge reading and it's way off. I have tried many formulas and changing the equation seems to be the only way to change the output. He mentions changing VE learn parrams for wideband which I have just done and will see what happens tomorrow. Any info would be appreciated.

[antus]

change the two voltage points in the 12P cal. how to get the right voltage offsets for those points is a challenge. I didnt find a good way, even though I am sure there is one. I did some kind of excel table with a graph and fudged numbers untill the slope on the graph was right and the two points I was changing crossed both of the two vendor supplied reference points. Cant remember exactly how I did it now though. You shouldnt need to touch the ADX.

Its "VE Learn Params - WideBand 0v AFR" and "VE Learn Params - WideBand 5v AFR" I understand. At least it did work for me, even though its not related to VE Learn.

[antus]

Actually I do still have the spread sheet. I wouldn't recommend doing it this way, but it's an idea.

I cant answer any more questions about it, it was a long time ago now. But I can see the slope and I can see the 0v and 5v points for the bin and the reference points from the wideband docs. I think the sequence number equasion was just tweaked until it lined up, rather than properly calculating it. These days AI can probably come up with a better idea.

Then once its in the bin, tweak the high or low points to adjust / offset the slope to account for analog voltage variation so that TP matches the WB display closely enough that your happy.

[pman92]

I asked AI, and summarised it here myself.

You need linear extrapolation.
Given your wideband's calibration points (V_low, AFR_low) and (V_high, AFR_high), the slope of the line is:

Code: Select all

m = (AFR_high - AFR_low) / (V_high - V_low)

Then the AFR at any arbitrary voltage V_target is:

Code: Select all

AFR_target = AFR_low + m × (V_target - V_low)

So to calculate the AFR values at 0v and 5v for entry into 12P:

Code: Select all

AFR_at_0V = AFR_low - m × V_low
AFR_at_5V = AFR_low + m × (5 - V_low)

Example calculation
0.5V = 7.35 AFR, 4.5V = 22.39 AFR.

Code: Select all

m = (22.39 - 7.35) / (4.5 - 0.5) = 15.04 / 4 = 3.76 AFR/V
AFR_at_0V = 7.35 - 3.76 × 0.5 = 5.47
AFR_at_5V = 7.35 + 3.76 × 4.5 = 24.27

So you would enter 5.47 and 24.27 into 12P

[vlad01]

I've personally never had an issue.

As far as my experience goes, all WD kits with analog output states the 0 to 5v values.

I just punch those in, key on and watch the WB readings. All kits have used also have a built in sequencer with the specs in the data/info sheet.

I just look at how much out the min, then the two middle and then highest reading are, and move either the 0v or 5v or both accordingly, re-key on and repeat.

Takes me around maybe 5 to 10 min and I get all values anywhere to exactly spot on to maybe no more than 0.1 AFR out, and if I can't get every sequence step spot on, I just make it that the middle two are, and move the error to one of the two ends where it actually doesn't matter for the engine in real life.

So for example, I may enter in 10.2 and 21 for the 0-5v, and the sequencer gets me 10.1, 13.3, 16.6, 20. (10 and 20 being the min and max for my WB output) These are actual figures I recall from the top of my head on one of the ECUs I had a WB setup on.

Moving the WB to a different car and ECU would alter the 0-5v some, but I could always get the sequencer to read out virtually spot on with just some quick tuning.

I know some people would remove a pull up or down resistor on the ECU's input, I never bothered as I could always just adjust for that and get damn good accurate read out.

Kits that have a physical gauge, I never bothered reading into those or what their numbers give, only ever paying attention to the sequencer as that is the true calibrated points you can rely on. Those external gauges can and are sometimes out. Those innovate kits popular back in the day, I recall there was something like 0.4 or so, maybe more, error on the gauge to the calibration sequence on a car I was running up on the dyno.

[Charlescrown]

Thanks to all for the input guys. The issue I have it's a Chinese 30-300 and their is doubt in my mind with the accuracy. Waiting on my new XGecu T48 to arrive as I have given up on the GQ-4X it took me about 10 goes to write a bin file so it should arrive next week. I'll also swap the Innovate to the car and compare the accuracy. I have more faith in using that.

[Charlescrown]

Quick question for pman92.
I did what I thought was the best with your calcs and came up with 0.637 for the low and 24.24 for the high. My question is do I now drop those figures into the range low and high for wideband ADX?

[pman92]

Quick question for pman92.
I did what I thought was the best with your calcs and came up with 0.637 for the low and 24.24 for the high. My question is do I now drop those figures into the range low and high for wideband ADX?

Dont edit the ADX. They are parameters within the 12P binary itself. See Antus' comment above.

[Charlescrown]

Ok got it. Change the bin file in these 2 areas till it is close to the wideband output.

Screenshot 2026-05-15 081612.png

[antus]

Yes. My AEM came with specs for 0.5 and 4.5v hence the talk above. Maybe they have 0 and 5v now since that seems more normal, dont know. And then there can be some analog drift. So if your log is, say, 1 AFR different to the display you can tweak these a little in real time to fix that.