narrow band o2 calibration in TS correct?

[evo828]

after I calibrated afr through TS (selected Narrow band sensor); project properties also narrow band; the ego o2 correction behaved in exactly opposite way - when it was rich it was adding fuel - when lean (below 0.5v) it was decreasing fuel. I looked at the std_ms2geno2.log and realised that the calibration logic seems to be reversed (below 0.5v = rich; over 0.5=lean). So went to megatune - did the calibration from there and it is now fine (EGO correction logic is doing what is expected).

Is there something special in TS that I missed or is it an error?
What does the 02 in the logs mean - it is showing negative values - although the voltages measured (and shown on gauges) are positive (between 0-1V)

attaching two short logs - before and after the MT calibration - for illustration.

[LT401Vette]

That does look backwards...

[LT401Vette]

I just checked the ms2ReferenceTables.ini file in the inc dir:

Code: Select all

    	solution	= "Narrowband", 		{ adcValue > 92 ? 16.9 : 12.5 }

Sure enough it is backward. You can edit it to be:

Code: Select all

    	solution	= "Narrowband", 		{ adcValue > 92 ? 12.5 : 16.9 }

I will get it in the next release.

[evo828]

Thank you Phil. And what does the O2 value in the datalog mean with its negative values?

[LT401Vette]

It looks like that means there is a line of the ini file TunerStudio is not interpreting as it was meant to...
#if NARROW_BAND_EGO
egoVoltage = { 1.0-afr1*0.04883 }, "V" ; For LED bars...

TunerStudio does not observe operational hierarchy, so for that line to work as I believe it was intended, it needs to be changed to:
#if NARROW_BAND_EGO
egoVoltage = { 1.0-(afr1*0.04883) }, "V" ; For LED bars...

Or I need to make my math parser respect operational hierarchy

Thanks for reporting this...

[racingmini_mtl]

Or I need to make my math parser respect operational hierarchy

That might be a good idea because that's not the first time it's happened
But I'm just nit-picking.

Jean

[LT401Vette]

That might be a good idea because that's not the first time it's happened
But I'm just nit-picking.

That is the best answer, but naturally there is a reason I don't jump on it . It would break many of the formulas in MLV and it would be a sizable undertaking. My guess would be about 3 weeks work. That is looking at it as really going into my math interpreter and changing a number of things that have been nagging at me for some time. So that is what I want to do, but because none of the existing ini files are depending on operational hierarchy until this 1, I have been leaving that until there weren't so many features still on the todo list and I'm not quite so far underwater.

But I'm gonna take a look and see if I can come up with something simpler and less impactful.

[racingmini_mtl]

Maybe you can have a warning if there are 2 types of operators (+ and *, for example) in the same expression but no parentheses. But it may not be easier to do this than do the "correct" solution.

Jean