Getting Knock Retard with new setup

**1slowbusa** · 07-27-2011, 01:32 PM

Hi,

If any of the senior tuners can give me a hand i would really appriciate it. I have posted on HP Tuners forum with my tune and scan.

http://www.hptuners.com/forum/showthread.php?t=35371

Thanks,

**1slowbusa** · 07-29-2011, 05:56 AM

Update,

everything is looking good, i have removed timing on the upper end to 25 and the car is running great. I am still getting KR at cruising in a higher gears but ive removed some of the timing and it is going away. My Lambda% error is around +/-3 which is looking really good. Hopefully today i can start on the MAF calibration.

**0rion** · 07-29-2011, 07:04 AM

Do you have a wideband? Your VE table is really choppy looking. I wouldn't worry about your timing until you get your fuelings squared away.

**0rion** · 07-29-2011, 07:11 AM

nevermind....I see it in your log. Why are you logging in lambda?

**1slowbusa** · 07-29-2011, 07:16 AM

According to the Greg Bannish DVD Lambda is the way to go. I have also smoothed the fuel since the scan i posted. Once i have got the SD mode to were i want it i will post it and i will move on to the MAF calibration.

How do you tune, in AFR? Why not in Lambda?

**0rion** · 07-29-2011, 07:17 AM

you need to log that wideband in AFR so you can take advantage of the histograms. I looked at just 1 point where you had knock on a throttle transition and you were at 16.5:1 afr

Get rid of all that stuff you don't need and only log the things you need to log. List your mods and where you're at in the tuning.

**1slowbusa** · 07-29-2011, 07:21 AM

If Frost can chime in, since i am about 5 hours away from him i was wondering if i could bring the car down one weekend or weekday and have him tune it on the dyno. How long would it take to tune? But it looks like i am going to need new injectors and i want to install a electronic cutout as well.
($450 seen on the website)

**1slowbusa** · 07-29-2011, 07:34 AM

Modifications are
G5x3 112LSA cam, AFR205 heads with 921 springs, LS6 ported oil pump, Double roller timing chain, EGR & AIR deleted, LS6 intake, SLP Lid, 1/7/8 TSP headers, TSP catted Ypipe, Stock muffler, Monster stage 5 clutch with flywheel.

**0rion** · 07-29-2011, 01:06 PM

have you done your idle tuning yet? I've never used lambda to tune......I use afr so I can use the afr error to tune the ve. You should put the car in speed density and dial in your ve using the afr error histogram. A lot of the things you're logging aren't necessary for what you need to do and you're not logging some things you should be. Look in the help files of the software and it gives you a step by step on setting up and tuning the ve in speed density. Get your ve on before tuning your MAF.

**1slowbusa** · 07-29-2011, 01:38 PM

I am using the LS1tuningguide.com for tuning the VE right now. The reason i am using the Lambda instead of AFR for what i understand, when using Lambda it doesnt matter what gas you use, like in my area 10%ethanal is added which can bring your AFR from 14.68 pure gasoline to as low as 14.1. Greg Bannish mentions this in his DVD and i have read multiple Forum thread to use Lambda. If im misunderstanding please correct me, i dont mind like i said im learning.

**0rion** · 07-29-2011, 04:27 PM

well......if you log your commanded eq ratio you can then create a lambda error histogram to do the same thing as the afr histogram. I've always just used afr error.

**1slowbusa** · 07-29-2011, 06:59 PM

Originally Posted by 0rion

well......if you log your commanded eq ratio you can then create a lambda error histogram to do the same thing as the afr histogram. I've always just used afr error.

Can you send me or post one of your scans so i can take a look.

**0rion** · 07-29-2011, 07:43 PM

I use afr and not lambda....I can send you a cfg that uses that. I can also help you figure out how to do it using lambda. Here's a good thread at hpt showing how to set up to log for lambda error. This will speed your tuning up quite a bit if you can use this percentage error to correct your ve.

http://www.hptuners.com/forum/showth...mbda+histogram

I'm tired and heading to bed right now but I'll set up a cfg for you tomorrow and get you started using the histograms. You want to get that fueling squared away before you worry about your timing.

**1slowbusa** · 07-30-2011, 05:15 PM

Thanks Orion,

I have EIO Pro and just figured out the other day how to put it in graph.

I have a Lambda error PID setup as my NGK wideband(output Lambda)/commanded Lambda

I then made a histogram (called Lambda_err_VE) using this Lambda error as my PID
(X header table i used Primary VE vs RPM vs MAP Units RPM)
(Y header is Primary VE vs RPM vs MAP units in KPA)

I then can drive the car normally and i get an Lambda error in the same setup as my Primary VE table

Then i can take the numbers and go the actual cell and multiply. After doing this i can then download and redrive the car and keep doing this until i get +/- 2 of my lambda error.

How does this sound? This is how the DVD is teaching but he uses a load bearing dyno and keeps it at a steady speed and increases once that cell smooths out.

**0rion** · 07-31-2011, 02:39 PM

yeah...that sounds about right. Email me that histogram you made and your cfg file to ibzrg6@aol.com along with a log and let me look it over. You should try to hit as many cells in the ve table as possible and set up a filter to have a minimum number of hits per cell. I think most recommend like 50 hits but I found it hard to hit some of those cells 50 times on the street so I usually do 25-30 hits and then hand smooth that table.

**1slowbusa** · 07-31-2011, 03:12 PM

Originally Posted by 0rion

yeah...that sounds about right. Email me that histogram you made and your cfg file to ibzrg6@aol.com along with a log and let me look it over. You should try to hit as many cells in the ve table as possible and set up a filter to have a minimum number of hits per cell. I think most recommend like 50 hits but I found it hard to hit some of those cells 50 times on the street so I usually do 25-30 hits and then hand smooth that table.

Email sent
Thanks,

**Frost** · 08-11-2011, 10:22 PM

Originally Posted by 1slowbusa

I am using the LS1tuningguide.com for tuning the VE right now. The reason i am using the Lambda instead of AFR for what i understand, when using Lambda it doesnt matter what gas you use, like in my area 10%ethanal is added which can bring your AFR from 14.68 pure gasoline to as low as 14.1. Greg Bannish mentions this in his DVD and i have read multiple Forum thread to use Lambda. If im misunderstanding please correct me, i dont mind like i said im learning.

You can use lambda, but keep in mind when you use AFR, you really ARE using lambda. The wideband just assumes gasoline and converts Lambda to AFR. As long as you think of 14.7 (or whatever your unit ascribes to lambda 1.0) as stoich, you ARE using lambda actually.

You can get the car here for tuning if you'd like, but I'd honestly want to keep it overnight to be able to visit cold-starting the next AM.

**0rion** · 08-11-2011, 10:42 PM

Originally Posted by Frost

You can use lambda, but keep in mind when you use AFR, you really ARE using lambda. The wideband just assumes gasoline and converts Lambda to AFR. As long as you think of 14.7 (or whatever your unit ascribes to lambda 1.0) as stoich, you ARE using lambda actually.

You can get the car here for tuning if you'd like, but I'd honestly want to keep it overnight to be able to visit cold-starting the next AM.

I've been reading since this thread and it seems like this is the "new" way since Bannish has been steering people towards it. I just don't understand it myself.....seems like you're making it harder than it has to be and adding more steps that aren't necessary for 99.9% of the cars out there. Maybe I'm just missing it though.

**Frost** · 08-12-2011, 11:48 AM

If you have NEVER been exposed to AFR then Lambda can be easier since stoich is 1.0λ. 0.86λ is about 12.6:1 AFR (14.7 stoich * 0.86λ = 12.6-ish:1). %error becomes native when you are looking for 1.0λ and getting 1.12λ; you need to add 12%. HOWEVER; since our widebands literally ARE lambda meters and they are converting based on assumed stoich, to me, converting back down is an extra step. Also lambda needs more points right of the decimal place for the same resolution that we are used to with AFR.

The argument is that lambda reaches past AFR because stoich can move based on fuel composition. However; when you see 12.6:1 on the wideband, it's ALWAYS 0.86λ since that is the base measurement anyway, and all of the alcohol content is already rolled in and accounted for.

Use whatever you are more comfortable with.