History of MembersPage/KoabiBrooks/Tuning
Older Newer
2006-10-14 17:08:43 . . . . 71-33-187-19.hlrn.qwest.net [added general AFR numbers]
2006-10-14 14:01:14 . . . . 71-33-187-19.hlrn.qwest.net [formatted]
2006-10-11 05:18:07 . . . . 67-40-143-208.hlrn.qwest.net [notes from g60 dave - currently unformatted]


Changes by last author:

Changed:
[1] g60: koabi, why is EGO correction turned off.

[2] g60: dont you have a Lambda table?

[3] g60: or was this a "tuning" session to get real numbers

[4] g60: you'll find if you get your map mostly smooth then turn EGO correction on it will smooth things out a fair bit

[5] koabi: everything is a tuning session so far

[6] koabi: i dont really have a lambda table

[7] koabi: alright, so i will try to adjust the VE table for smoothness

[8] koabi: then i will investigate getting a lambda table together

[9] g60: create a lambda table

[10] g60: with your targets

[11] g60: to get AFR.. multiply lambda x 14.7 for gasoline

[12] g60: once its full.. turn it on.. go for a drive

[13] koabi: i was thinking .87 ballpark..

[14] g60: see how far it has to correct to get you to the numbers you can then make pretty close guesses on your VE table entries.. make the changes and go for another drive

[15] g60: makes it easier to tune by yourself.

[16] koabi: so you are advocating doing lambda correction before i mess with the VE?

[17] g60: yeah.. if you are tuning on your own

[18] g60: you'll be surprised how much better your car drives

[19] koabi: ok

[20] g60: and using the datalogs you can then make the changes

[21] g60: if you are tuning with a partner <someone driving> leave it off as you can correct in realtime

[22] koabi: i have a lot to learn

[23] *** kalsket has signed off IRC ().

[24] koabi: sara is willing to boost her wrx

[25] g60: let it do 20% rich or lean correction as long as your wideband is properly calibrated

[26] koabi: dunno about the scirocco ;p

[27] g60: lol.

[28] koabi: my wb is close...

[29] g60: then try 15% to start

[30] koabi: i think bob's guide was updated, so ill try again.

[31] g60: your map isnt taht far out anyways

[32] g60: if you datalog with EGO correction on i could probably tune your car using your datalogs and a map

[33] g60: i've done it before with MS for customers who live many thousands of miles away

[34] koabi: i guess i am still confused on the point where lambda correction comes in if i am tuning myself as opposed to tuning on the fly, though. i am beyond novice

[35] g60: i know your beyond novice

[36] g60: but you cant drive and make corrections at the same time

[37] g60: so if you leave EGO correction on, in the datalog where it says EGO in the colum and its all 100 right now

[38] koabi: yeah

[39] g60: if it goes down to 80 then you know it was trying to lean out your mix by 20% to reach your target

[40] g60: or if it goes up to 106 then it was adding 6% fuel for that VE area to reach your target AFR

[41] koabi: ok - and if i can get a partner to drive - how do i go about tuning the VE table

[42] g60: its not a perfect science and assumes things like your ignition system is working

[43] g60: if you can get a partner to drive, just sit with the tuning window open on VE.. watch the lambda gauge on the left, and your target gauge on the right and make corrections by hand as they drive

[44] g60: ego correction disabled.

[45] g60: still nice to have your lambda table filled in

[46] g60: its a handy reference

[47] koabi: so i have to be pretty quick

[48] g60: yeah.. but not really at the same time

[49] g60: you just need a decent memory of where the green dot went and what the values were as it traveressed its path

[50] koabi: yeah

[51] g60: its likely to take nearly the same path each time through your table if the driver repeats the behavior

[52] g60: theres large chunks of your map that really never get used

[53] koabi: so i hire my brother who is very erratic

[54] g60: lol.

[55] g60: it all comes down to practise.. we had the 1.8T car tuned perfectly in under 2 hours..

[56] koabi: nice

[57] g60: and the miata i actually lost torque on the dyno when we tried making changes to the timing

[58] koabi: so you got it damn close

[59] g60: the map he drove in on was the best one we could come up with even after an hour on he dyno

[60] g60: we changed his timing by 1deg and his fueling by maybe 5% in a few areas

[61] g60: with a wbo2 and EGT you can do a lot yourself.

[62] koabi: i am sans EGT, but the WB should be a bit more precise, i'd guess

[63] koabi: ok i have an egt but the secret is i really dont want to pull the turbo/manifold off to put it on right now

[64] g60: you can tune without it

[65] koabi: i will try to talk my corvette driving friend into driving my car.

[66] g60: you'll get the hang of it pretty quick.

[67] koabi: i dont know much of the theory so i may not. but i'll try

[68] koabi: is .87 a decent target?

[69] g60: go richer on a turbo car

[70] g60: 0.86-0.82

[71] g60: if it were my VW i'd be targetting 0.82-0.83

[72] koabi: noted

[73] koabi: and that's at WOT

[74] g60: specially since you dont have 94 octane

[75] g60: yeah anything over 100kpa

[76] koabi: yeah stupid colorado

[77] g60: 100kpa target 0.89

[78] g60: 85-90 target 0.91 0.92

[79] g60: below 85 you can leave it all at 1.0 as a start

[80] g60: you can probably get a leaner cruise if you want

[81] g60: and you might need a richer idle.. 0.96-0.99 on some cars depending on your setup for a smooth idle

[82] g60: some cars will idle nicely at 1.08 lamba too like my suburban

Some general tuning guidelines from g60dave

If you can't tune for load situations, turn EGO correction on and create a lambda table. You'll find if you get your map mostly smooth (prior to EGO correction) then turn EGO correction on it will smooth things out a fair bit. This technique is good if you have to tune by yourself or tune a car that's not local to you.

create a lambda table with your targets. To get AFR.. multiply lambda x 14.7 for gasoline. Once its full.. turn it on.. go for a drive

see how far it has to correct to get you to the numbers you can then make pretty close guesses on your VE table entries..and use the datalogs then make the changes and go for another drive

about making lambda tables:

let it do 20% rich or lean correction as long as your wideband is properly calibrated or try 15% to start

if you leave EGO correction on, in the datalog where it says EGO in the colum and its all 100 right now. if it goes down to 80 then you know it was trying to lean out your mix by 20% to reach your target or if it goes up to 106 then it was adding 6% fuel for that VE area to reach your target AFR

----

If you are tuning with a partner <someone driving> leave EGO correction off as you can correct in realtime

if you can get a partner to drive, just sit with the tuning window open on VE.. watch the lambda gauge on the left, and your target gauge on the right and make corrections by hand as they drive with ego correction disabled - it is still nice to have your lambda table filled in because it's a handy reference.

you just need a decent memory of where the green dot went and what the values were as it traveressed its path. its likely to take nearly the same path each time through your table if the driver repeats the behavior. Also there are large chunks of your map that really never get used

it all comes down to practice.. we had the 1.8T car tuned perfectly in under 2 hours..

----

Lambda targets for gasoline

turbo car: 0.86-0.82 lambda for anything over 100 kPa - 0.82 to 0.83 recommended on 91 octane

100 kPa target 0.89

85-90 kPa target 0.91 0.92

below 85 kPa you can leave it all at 1.0 as a start, though you can probably get a leaner cruise if you want

you might need a richer idle.. 0.96-0.99 on some cars depending on your setup for a smooth idle

some cars will idle nicely at 1.08 lamba thoough

general guidelines:

12.8:1 - 13.2:1 / 0.87 - 0.90 NA

12.0:1 - 12.5:1 / 0.81 - 0.85 turbo