E38 Lambda tuning approach with LM-2

[mowton]

Just want to double check that my current Lambda tuning strategy is correct. Sounds like it to me, but just don't want to be wrong with a 900 hp FI setup :shock:

My approach is to set the LM-2 to 14.2 AFR for all applications. Based on this, I establish a Lambda PID in calc.pids that takes the ext wb input and divides by 14.2. This establishes my lambda reading developed by the WB's O2 sensor. Then I setup a Commanded Lambda PID which simply takes the tunes Stoich value (i.e 13.6) and divides by the E38.AFRATIO_DMA value to arrive at a commanded lambda PID. Lastly I setup a Lambda error PID which takes the LM-2 developed Lambda value and divide by the commanded Lamda value. I then map that error value similarily as you would for AFR.

*CLC-00-111
factor 0.5 1.5 .4 "{E38.AFRATIO_DMA}/13.6"

*CLC-00-112
factor 0.5 1.5 .4 "{EXT.WO2AFR1}/14.2"

*CLC-00-113
factor 0.5 1.5 .4 "{CALC.E38Lambda}/{CALC.E38CMDLambda}"

and........

CALC.E38CMDLambda F111 CLC-00-111 factor Tuning "BEN lambda from serial WB"
CALC.E38Lambda F112 CLC-00-112 factor Tuning "BEN LM-2 Lambda"
CALC.E38LambdaErr F113 CLC-00-113 factor Tuning "BEN Lambda Err"


Here is a copy of the calc.pid text file.

14041

Thanks for the look

Ed M

[swingtan]

Why not just log Commanded Lambda from the E38 and measured Lambda from the LC-2 ? That's how I do it with the Techedge and then there's no need to worry about converting for the stoichiometric ratio. I'm running E85 and didn't even bother resetting the Techedge WB controller or the LA1 gauge, I just display Lambda on the display and log Lambda with the V2.

If you then want to run a BEN factor, just use the two Lambda data feeds.

Simon.

[mr.prick]

^^^^^^^^^^^^^^^^^^^^^^
({EXT.WO2LAM1}/{SAE.LAMBDA})

I suggest avoiding "nested" calc_pids. (a calc_pid based off another calc_pid)
The data rate tends to be slower.

[mowton]

I am more used to HP Tuners, but I couldn't find SAE.LAMBDA in the PIDS supported. Will the PIDS validate automatically when I open the E38 scan mode or do I have to manually select it to get all supported PIDS?

The only reason I didn't change the LM2 to Lambda was because mu buddy is still a bit uncomfortable about LAMBDA and feels a bit better if he can see the "normal values" :) I will eventually go to lambda display setting.

Thanks for the input

Ed M

[joecar]

Hi Ed,

In the scantool, enter your VIN and OS id (under the File menu item), on the PIDs tab uncheck Supported, and try doing Validate Pids.


When you log SAE.LAMBDA make sure it has the correct "sense" (i.e. it may actually be EQR on some OS's)...


so first log it and in PE see if it goes below 1.00 (indicating it is LAMBDA) or above 1.00 (indicating EQR);

and then modify CLC-00-110 as mr.prick and swingtan indicated above.


As for your friend, now is a good time to learn Lambda. :cheers:

[mowton]

Thank you mr.prick, Simon and Joe, I re-hooked up to the car, ran Validate PIDs and voila.......SAE.LAMBDA revealed itself. Much simpler with it and EXT.WO2LAM1 along with an Err Pid I set up.

Thanks again.

Ed M

[joecar]

Be sure to check the sense of SAE.LAMBDA...

( it may be Lambda or EQR )

( Lambda = 1/EQR )

[mowton]

Be sure to check the sense of SAE.LAMBDA...

( it may be Lambda or EQR )

( Lambda = 1/EQR )


was Lambda...1.0 at Stoich, .77 at PE :)

Ed M

[joecar]

Ok, so it is Lambda... :cheers:


using Notepad edit your calc_pids.txt as follows:

in CLC-00-110 change {GM.EQIVRATIO}*{EXT.WO2LAM1} to read {EXT.WO2LAM1}/{SAE.LAMBDA} as shown by mr.prick above.



Edit: I corrected my mistake (I had somehow become totally stupid today).

[mowton]

Ok, so it is Lambda... :cheers:


using Notepad edit your calc_pids.txt as follows:

Edit: I corrected my mistake in the post above (sorry I was totally stupid today).

This what I updated to and deleted all the other garbage......worked great, and much simpler

*CLC-00-110
factor 0.5 1.5 .4 "({EXT.WO2LAM1} / {SAE.LAMBDA})"

and

CALC.LAMBDAERR F110 CLC-00-110 factor Tuning "BEN Lambda err from Serial WB"

Ed M

[joecar]

You can name it CALC.WO2BEN.

[joecar]

I corrected my mistake in post #9 above.