Help with VE Tuning......

[Behold]

I never seem to be able to get a clean VE Table. and by clean I mean all between 0.98 - 1.02 And also my VE Table has some really high bits. do i just need more filtering to get more acurate data???

Second as there is a cross over on the between the normal VE and the Boost VE what should you do on the two 105KPA Columns???

Thanks

[Gregs]

I never seem to be able to get a clean VE Table. and by clean I mean all between 0.98 - 1.02 And also my VE Table has some really high bits. do i just need more filtering to get more acurate data???

Second as there is a cross over on the between the normal VE and the Boost VE what should you do on the two 105KPA Columns???

Thanks
i'm assuming this is on a cos, the 105kpa columns should be the same. you need to smooth out the cells with the peaks and then go make a log. you could try applying the ben, smoothing out everything and then do a ben @ 50% times the ve table. you should be filtering the data good and hide cells with less than 30-50 cell counts.

how are you acquiring your data through the wideband, are you using the serial port?

[Behold]

Yes im using COS5. I started with 105 on both the same. when you smooth out the table you endup with an adjusted 105 that then don't match. which table is used???

I generally do 2-3 runs then smooth out the peaks. but i have a few areas that always re adjust.

What do you mean by "then do a ben @ 50% times the ve table"

And yes im using an LC-1 in the serial port of my V2.

[Gregs]

Yes im using COS5. I started with 105 on both the same. when you smooth out the table you endup with an adjusted 105 that then don't match. which table is used???

I generally do 2-3 runs then smooth out the peaks. but i have a few areas that always re adjust.

What do you mean by "then do a ben @ 50% times the ve table"

And yes im using an LC-1 in the serial port of my V2.



-you should always have a smooth ve table in the car, having that spike in there and then doing a log is just going to mess stuff up

-under paste options go paste>paste and multiply with labels > paste and multiply (50%) with labels. its not necessary but you could try it and see if doing it that way helps.

105 kpa need to be the same, the computer can use the 105kpa on either table so they have to be the same. just chose one column and paste it to the other table after the smoothing

i don't know if you want your tune to be out on the web but if you want to post it up with a log i could take a look at it

[Behold]

I'll smooth out a tune and do a run and post them later this week.

[joecar]

What are you using for the BEN pid...?

Are you using EQIVRATIO x Lambda...?

[Behold]

What are you using for the BEN pid...?

Are you using EQIVRATIO x Lambda...?

the ben pid is the built in ls1 serial. {EXT.WO2AFR1}/{GM.AFR}

EQIVRATIO x Lambda??????? should i use {EXT.WO2EQR1} by something?

[mr.prick]

Serial BEN pids should be:
"({GM.EQIVRATIO}/{EXT.WO2EQR1})" or "({GM.EQIVRATIO}*{EXT.WO2LAM1})"

Please ignore the {GM.AFR} pid. ;)

Filtering helps a bit too.
Omit MAP transitions greater than 5-10kPa per frame.

IMO this works better than filtering TP%.

[Behold]

Hmmm... Ill change that and see what my result is. What difference will I see on the BEN result?

[Behold]

Also Im taking it thats a custom PID I need to create as I cant find it???

Right I have now been relooking at stuff.

What is the prefered method now of VE Tuning? Im in SD Mode as my MAF is gone. At the moment I use the normal AutoVE tuning. Is there a better way that im missing? CALC.VET?

Thanks Antony

[joecar]

Using EQR and/or Lambda avoids errors due to differences in AFR when alcohol content is not precisely known.

Look at the WO2BEN pid in the calc_pids.txt file attached to post #1 of this thread:A-New-Twist-on-CALC.-VE-Table..Computing-the-Entire-VE-Table (http://forum.efilive.com/showthread.php?15236-A-New-Twist-on-CALC.-VE-Table..Computing-the-Entire-VE-Table).

This implements what mr.prick said in post #8 above.

[joecar]

Serial BEN pids should be:
"({GM.EQIVRATIO}/{EXT.WO2EQR1})" or "({GM.EQIVRATIO}*{EXT.WO2LAM1})"

Please ignore the {GM.AFR} pid. ;)

Filtering helps a bit too.
Omit MAP transitions greater than 5-10kPa per frame.

IMO this works better than filtering TP%.I like this idea :cheers:

[Behold]

1211912120

Attached is my VE Tune and a quick blast (So not much over 50). it actually looks quite good on this run with the EQR and infact im not that far off in general in my amatuer opinion. One weird issue is I have 3 maps on my PC. 15-105kpa 15-285kpa and 105-285kpa. All three show different results and counts for 105kpa.... why???

I would be open to your feedback. Also I think i need to sort my new injectors.... 120 Duty cycle seems like im topped out...

Although its probably not as bad as it looks as the AFR is 12.8 at the time of the injectors topping out. Ill be working on my tune once out of VE

[Gregs]

1211912120

Attached is my VE Tune and a quick blast (So not much over 50). it actually looks quite good on this run with the EQR and infact im not that far off in general in my amatuer opinion. One weird issue is I have 3 maps on my PC. 15-105kpa 15-285kpa and 105-285kpa. All three show different results and counts for 105kpa.... why???

I would be open to your feedback. Also I think i need to sort my new injectors.... 120 Duty cycle seems like im topped out...

Although its probably not as bad as it looks as the AFR is 12.8 at the time of the injectors topping out. Ill be working on my tune once out of VE

your ve table looks like its really close, looking good keep at it!

one thing you you might want to know is that if you're smoother with the throttle you will get better data. try to slowly accelerate and decelerate and not so jerky with the petal. :)