Calc.VET: correcting MAF and calculating VE (in single log)

[darcy]

Well, while logging my car, the status shows it is in closed loop, yet the LTFTs are not enabled. So what does that mean?

You're in closed loop, using O2 feedback to generate STFTs, but not turning those into LTFTs.

[COONDOG96]

OK somewere in this I'm missing a step or something. My VE map is in #like 82.324and 85.436 but when I go right by the new Tutorial they end up being #like 1.325 up to like 3.42 witch sound like % to me but I've gone back and redid all the maps and logs and still get them. Were am I going wrong?????

[Rich Z]

You're in closed loop, using O2 feedback to generate STFTs, but not turning those into LTFTs.

OK, that makes sense. Is there a good writeup anywhere here about speed density tuning in EFILive so I can stick my toes into that concept a little bit?

[joecar]

OK somewere in this I'm missing a step or something. My VE map is in #like 82.324and 85.436 but when I go right by the new Tutorial they end up being #like 1.325 up to like 3.42 witch sound like % to me but I've gone back and redid all the maps and logs and still get them. Were am I going wrong?????

Check the units of your scantool WO2BEN map and your tunetool VE display units...

they have to be the same (either % or g*K/kPa), and you can't mix units

[ I prefer g*K/kPa ]

Post screenshots (using shift-prtscn button).

[ferocity02]

I have a question for you experts. Instead of copying and pasting the calculated VE table into the tune as is, what would you think about doing it row by row. Let's say I want to do the 1600 RPM row... I use the normal ECT and TP filters, and then also filter out data under 1500 and over 1700 RPM. I think this would give a much better representation of the 1600 RPM row.

The issue I think I'm having is that when I log data on the highway, I'm spinning around 1750 RPM, which is almost halfway between 1600 and 2000 RPM, but all of this data is going to the 1600 RPM row. So my VE at 1600 RPM is too high because it's actually representing 1750 RPM.

Or maybe make the data table with increments of 2000 RPM and do some clever interpolation to make the final VE table.

Thoughts?

[joecar]

I have a question for you experts. Instead of copying and pasting the calculated VE table into the tune as is, what would you think about doing it row by row. Let's say I want to do the 1600 RPM row... I use the normal ECT and TP filters, and then also filter out data under 1500 and over 1700 RPM. I think this would give a much better representation of the 1600 RPM row.

The issue I think I'm having is that when I log data on the highway, I'm spinning around 1750 RPM, which is almost halfway between 1600 and 2000 RPM, but all of this data is going to the 1600 RPM row. So my VE at 1600 RPM is too high because it's actually representing 1750 RPM.

Or maybe make the data table with increments of 2000 RPM and do some clever interpolation to make the final VE table.

Thoughts?

I think that's a good idea ("bucketizing").

[tinindian]

Where is this Calc.vet tuning tutorial I keep reading about? There's got to be more explanation than just the few steps listed in the first post of this thread. Yes, I've read the summary notes. The problem is that I have an analog wideband and can't get the proper PIDs. I think I have all the crap taken care of in calc_pid.txt. The first post does a good job showing things for a serial WB. I've done AutoVE before and had no problems with that writeup. But now I have a 5-pin MAF and don't want to screw with the wiring so calc.vet seems like a good option. What am I missing?

[joecar]

Hi tinindian,

Post #1 has all the steps and explanation...

it also shows you how to handle analog wideband:

. . .

II. PROCEDURE for ANALOG WIDEBAND or V1

To obtain LAMBDA from an analog wideband, divided wideband AFR divided by the wideband's assumed stoich AFR.

Method 1:
Using NGK AFX as example (NGK assumes stoich to be 14.57):
edit CALC.WO2BEN1 (CLC-00-110) to "{GM.EQIVRATIO}*{CALC.AFR_NGK1}/14.57"

Method 2:
Using LC-1 as example (LC-1 assumes stoich to be 14.7):
define CALC.LAMBDA (a new CLC, say CLC-00-100) to be "{CALC.AFR_LC11}/14.7"
edit CALC.WO2BEN1 (CLC-00-110) to be "{GM.EQIVRATIO}*{CALC.LAMBDA}"

or simply combine those (i.e. same as method 1):
edit CALC.WO2BEN1 (CLC-00-110) to be "{GM.EQIVRATIO}*{CALC.AFR_LC11}/14.7"

. . .

the Calc-VET-summary-notes provide additional info to guide you along.

[joecar]

You can fail the MAF without un-wiring it (if you don't understand this, then you did not understand the AutoVE tutorial)...

but if you're going to go SD (MAF-less, failed MAF) then you would be doing the Calc.MAFT proceedure (can you see why...?)...

but you can only do it if your OS supports the pid GM.VE_TABLE_DMA (can you see why...?).

[tinindian]

Hi tinindian,

Post #1 has all the steps and explanation...

it also shows you how to handle analog wideband:


the Calc-VET-summary-notes provide additional info to guide you along.

I guess indirectly it tells you what to do but if it's truly a tutorial why not provide the whole thing for people that don't have serial wideband? Not even sure if this is correct:

*CLC-00-100
factor 0 20 .1 "{CALC.AFR_LC12}/14.7"

*CLC-00-110
factor 0.5 1.5 .4 "{GM.EQIVRATIO}*{CALC.LAMBDA}"

I know you're trying to get people to "figure it out" but why make it easy for people w/ serial & hard for people w/ analog?