B4349 Scaler for LS2 TB upgrade (Real FACTS)

[kwhiteside]

There are lots of guys upgrading thier LS1/LS6 Engines with the FAST Intakes and LS2 TB (or similar). If you've searched for what adjustments to make you will be scratching your head because about half the people have convincing technical arguments with spreadsheets that the scaler number needs to go down.

theory here is when you go from 78mm TB to a 90mm TB, original setting on a C5 Corvette is .0255 for B4349 and it should be reduced to .0191.

There are equally as many guys doing real live tunes saying an increase to .0320 is the only way to go.

My story is that I started with a Z06 with headers, FAST92, LS2, and a jacked up tune I inherited, then began my EFILive learning experience. After a while, I ripped the intake and TB off and went back to stock. I worked by tune in shape and proved it at Road Atlanta doing over 50 laps hitting high speed red lines over and over. I did notice my car was slower without the extra air from the FAST and LS2 TB, so I'm going for it all over again. This time I know for sure my base tune was great. Idle was great too.

So what to do with the B4349 TB area scaler?

Since one of the above theories is going the wrong way, wouldn't it be pretty noticable if you applied both settings and did some logging? That's exactly what I did today. I'm going to repeat tomorrow because my wideband must have come unplugged and didn't log. Regardless, I'll share a couple points from todays logs that stuck out.

The logs will have _0191_ or _0320 in the name and the only difference between the tunes they run on will be the setting for B4349. I've done everything in the AutoVE tutorial so no MAF, no Fueltrims, open loop speed density.

Notice the difference in the MAF Grams/s being logged with everything else being nearly the same. I believe the .0320 tune is reporting much lower then it should be. Also note the .0320 intruduces some severe Knock Retard. I also noticed the .0320 tune dropped below 700rpm many times. Nothing close to stall at this point, I'm guessing because my base tune was solid.

This is preliminary as I will repeat tomorrow, but at this point, .0191 seems to be the correct value.

If you want me to log some other pid and look for something specific, feel free to make suggestions.

Ken . . .

[joecar]

Ken, very interesting, thanks for the info...:cheers:

[redhardsupra]

all you need:
http://www.marcintology.com/tuning/TBresizing.xls

[kwhiteside]

Red, I pm'd you earlier about this ss.

Seems it is very misleading for corvette owners.

The default setting for LS1/LS6 78mm TB is .0255. The spreadsheet has a very different value. I found a multiplier to make it correct, and then used that on the 90mm cell.

This is how I came up with .0191. I used your spreasheet as shown below. Also some other tuners said they ended up using .019 which was a good confirmation.

That spreadsheet should probably either be updated or at least explained that the settings do not start with the corvette / camaro settings. Somebody said it may be correct for trucks using LS1/LS6??

[redhardsupra]

Could you explain the 'I found the multiplier' part? How did you do it?

The reason for the discrepancy is that I calculated the area purely from geometry. In reality you got some metal in the way, which makes the effective area slightly smaller. The general approach however is I'm fairly convinced correct, as I've used it on non-circular TB's by altering the formula for the area according to the desired shape.

I've never encountered any problems with the values generated by this spreadsheet, neither have anyone raised any questions/issues with. If you have any solid evidence pointing toward something new, please let me know, I'd be more than happy to alter the spreadsheet and give you credit for it.

[kwhiteside]

Could you explain the 'I found the multiplier' part? How did you do it?



Well, start with the stock corvette LS1/LS6 tune for 78mm TB. Right off there is the problem. Value is .0255 not .020928

In one of the original threads somebody explained the formula problem, but I wasn't getting that technical.

I took .0255 / .020928 and got 1.218482 multiplier. I then moved down to the 90mm cell and multiplied .015719 * 1.218482 to get the new value of .019153.

I think your approach seems to be correct. But if you are looking specifically at values to change B4349, shouldn't you at least start with a formula that matches. 78mm is .0255 according to B4349, not .020928.

Highlander was the one who also suggested .019 which matches your ss if starting from the correct point.

This is the original thread where my same point is brought out
http://forum.efilive.com/showthread.php?t=4816&highlight=b4349+90mm

That said, out of respect for you, knowing that you have lead a lot of the current understanding, I will add a third tune with .0157 to run my logging with tomorrow and see how it reacts. I would like you to try to make me understand why GM puts .0255 into b4349 when you calculate .0209 in you computations. You must admit, something is missing there since it is the values of b4349 we are adjusting. How did gm come up with .0255?

[kwhiteside]

Bruce Melton quote
"From the repository it looks like most stock Camaros and Corvettes are .0255 and trucks are .0208."

[kwhiteside]

and this says it better than I did.

http://forum.efilive.com/showpost.php?p=43960&postcount=16

[kwhiteside]

I've done two runs with my wideband working again. After analyzing these side by side looking for the "Right One", I've come to realize a few things. I start looking for my strong pulls at full throttle. I'm thinking which one give me better power. Neither :doh2: With either the .0191 or .0320 setting, at full throttle they are almost always identical. Hmmm, why would that be? Because the blade is full open. What we are messing with here is how much air the TB is letting in at each step or increment of the blade opening. This is why people start fussing with this setting when they have idle problems. My car seems to be idling fine with either setting and the only difference I can make sense of is that the .0320 tune runs .5 richer at idle. But then there is the 20 degree difference in temperature that may account for some of that.

My head is hurting and I'm not sure what to make of it at this point. I'll upload a couple screenshots, my tunes, and my logs in case any of you think it is worthy to figure out.

[kwhiteside]

Here's a full throttle 4th gear pull showing that things are almost identical.

[kwhiteside]

Obviously, I've bored most of you. I'm thinking now that if you start with a rock solid tune to begin with, you most likely wont have to make any changes to the scaler. A couple months ago when I slapped the FAST onto my engine with a jacked up tune, I could not stop my car from stalling out every now and then. Now I can change the scaler either way and my car idles really well regardless.

That said, and without much input from others, I picked the .0191 setting and have finished AutoVE and started on AutoMAF. While my tune is shaping up very quick, in the back of my mind I have noticed two things now that suggest I may have chose the wrong path.

First symptom - the .0320 tune ran .5 richer at idle. This indicates that scaling to a greater value, adusts the throttle blade to let less air in.

Second symptom - My first AutoMAF run found my MAF reading too much air for the initial opening steps to the throttle.

I'm thinking I can go back to my 0320 tune and setup for AutoMAF, do a log run, and see if those early steps are less lean. If that turns out to be the case, it seems to me to prove that the .0320 scaler setting actually adjusts the LS2 Throttle body to let less air in.

Keep in mind the whole point of the exercise is to take your new bigger 90mm throttle body and prevent it from letting massive air into the mixture so your car will idle properly. Then those step setting should also carry through wide open throttle to be more correct.

http://downloads.dataxperts.net/Tuning/Rich%20at%20idle%20for%200320.png

http://downloads.dataxperts.net/Tuning/AutoMAF%20too%20much%20air.png

[kwhiteside]

Tested the 0320 setting in AutoMAF mode with two log runs, and it ran even leaner. I also had a noticable surge or hunt at stops and could see my AFR gauge move back and forth in rythm to the hunting. Sure didn't like that. Wish I could have shown something indusputively conclusive.

I'm going to move on with my .0191 setting.

[joecar]

I'm not bored... I just don't know enough... I read your posts, and I would agree that the 0.0191 seems to be the more correct one (for want of better terminology).

[redhardsupra]

Bruce Melton quote
"From the repository it looks like most stock Camaros and Corvettes are .0255 and trucks are .0208."

The fact that the same size TB's generate a completely different effective area estimates to me signifies that GM is either fudging this one, or they're accounting for something else with it. Obviously in fbodies the area somehow has to be correlated to the airflow, and that must get translated to IAC steps. In Vettes you gotta convert the effective area to the angle of the TB to keep the engine going with enough airflow.

To my knowledge, noone ever fully described the process of airflow estimation at idle in general, effective area <-> IAC steps/ETC angle transformations in particular.

What Jeff described in the post you linked to makes no sense whatsoever. Division of the fudge number by the ratio (not difference, difference is a minus) of areas would suggest that he was attempting to scale the fudge value, assuming that whatever fudge that number carried is heteroskedastic. We dont know that, as we dont know what are they're fudging for. Also if he wanted to scale it, he should've been multiplying it not dividing it. Mathematically failing to do bad science is pretty much why I haven't agreed to anything that Jeff proposed in the last 5yrs on the boards ;)

If the fudge value is to account for the area that's obscured by the TB at different angles, then it's definitely not linearly scalable as Jeff would have you believe.

I dont think we should be looking at the scalar alone, I think we need to figure out how that scalar gets converted to the desired airflow. The full idle airflow model would be hugely beneficial as well, because we know all RAF, MAF, and Dynamic Airflow all play a role, but how, when, and to what extent they kick in is a mystery.

Also, idle is usually short pulse territory. This means that WB readings, and injector data can be screwing with you in a big way, so you attributing it to some TB scalars might be completely wrong.

[mr.prick]

Besides the spreadsheet and (X/Y)*X = {B4349} someone has posted a method for this:
What you have to do is simple (http://forum.efilive.com/showpost.php?p=99281&postcount=8)

The real question is why is this a sticky?

[joecar]

Just so I can track it for a few days... he posted some real data for us to look at.

[kwhiteside]

read ... read... heteroskedastic
read ... read...
big mystery

Got a little chuckle from me as I read that.

You know I've only been doing this for 2 months, and I know you've been doing it for many years, so I'm in no way trying to discount anything you've said. Keep in mind that us new guys get frustruted when we do use the search button and find experts pulling us in opposite directions.

Two months ago when I started with my FAST and then began tuning for the first time, I wanted to pull my hair out. The car's rpm's would divebomb and die and I tried everything, inlcluding idle tuning and throttle cracker stuff. I finally bailed on the FASt and went back to stock, started the tune over and got it right on. Then I took the car to the track and missed the extra ponies. Gotta try again. Now it is going very smooth. Smooth with the exception of what I reported today with the .0320 tune and the AutoMAF process. The car started surging / hunting at idle. I could literally see my AFR gauge sweep back between 13 and 16 AFR. I started to have flashbacks of the days were I couldn't get the car to idle right. I was hoping to prove it technically, but am not sure I know enough about deciphering my logs to do that. Real World Seat of Pants test, I'll stay away from that hunting idle for sure. Technically, the lower settings made most sense to me, that is why I went that way on my own. Perhaps you .0157 value is even better than the .0191. I showed how I got my .0191 value with your spreadsheet. I suppose it is quite heteroskedastic and yes it just repeats the GM fudge. I don't know if your value is even better, I didn't test it as I ran out of energy doing my log runs. Can't do that on public highways for ever without incurring a big cost eventually.

Ken . . .

[redhardsupra]

read ... read... heteroskedastic
read ... read...
big mystery

Got a little chuckle from me as I read that.

You know I've only been doing this for 2 months, and I know you've been doing it for many years, so I'm in no way trying to discount anything you've said. Keep in mind that us new guys get frustruted when we do use the search button and find experts pulling us in opposite directions.
Try 5 years, see how that fuels your frustration ;) we seriously knew about the same amount about idle mechanisms then as we know now.


Two months ago when I started with my FAST and then began tuning for the first time, I wanted to pull my hair out. The car's rpm's would divebomb and die and I tried everything, inlcluding idle tuning and throttle cracker stuff.
Heh, I tune that with RAF, idle spark, and the TB scalar you didn't like. Rarely do I need anything more than that. Remember that Fuel, Air, Spark are the big villains, the rest are merely minions and henchmen.


I finally bailed on the FASt and went back to stock, started the tune over and got it right on. Then I took the car to the track and missed the extra ponies. Gotta try again. Now it is going very smooth. Smooth with the exception of what I reported today with the .0320 tune and the AutoMAF process.The car started surging / hunting at idle.
Too much fuel, seriously that's the correct diagnosis like 99% of the time.

I could literally see my AFR gauge sweep back between 13 and 16 AFR. I started to have flashbacks of the days were I couldn't get the car to idle right. I was hoping to prove it technically, but am not sure I know enough about deciphering my logs to do that.
Practice makes perfect, and you're absolutely right, reading logs is the way to tuning, not arbitrarily making changes to the tune <HINT HINT>

Real World Seat of Pants test, I'll stay away from that hunting idle for sure. Technically, the lower settings made most sense to me, that is why I went that way on my own. Perhaps you .0157 value is even better than the .0191. I showed how I got my .0191 value with your spreadsheet. I suppose it is quite heteroskedastic and yes it just repeats the GM fudge. I don't know if your value is even better, I didn't test it as I ran out of energy doing my log runs. Can't do that on public highways for ever without incurring a big cost eventually.

Ken . . .

Huh? You're doing idle at highway speeds? :confused:

[kwhiteside]

Huh? You're doing idle at highway speeds? :confused:

Ok, you taunt me with that one. I tested the 0320 idle stuff on the way to the gym, and on the way back today. Mabye 10 stoplights each way.

Most of my log runs take 7 minutes to get to clear highway with many lights (idle stuff). Yes we are talking about idle, but I'm still tuning. To me, tuning is for when I go to the track and hit redline in 4'th gear of 140mph. Now I don't go that fast of course during my log runs, but I do get the 90-105 cells kicking in. I use both 3rd gear redline and some quick 4th gear floor board up to speed, then break to get down from go to jail speed. Actually similar to what happens at the track. Floor it, then get ready for a turn. Oh, and each run also includes a drag strip run as there are some nicely isolated on ramps that allow you to go full out 1st thru 4'th without drawing attention. Of course no other cars around me during testing.

The only time idle even matters at the track is when cruising thru the paddock. Having your car die there amongst all the 250k Ferraris and Vipers would be like doing bench press at the gym and getting stuck on your chest, then having the throw the weights off to one side so everybody in the gym gets a good laugh. (Don't try too many reps at 285 without a spotter :help2:)

http://downloads.dataxperts.net/Road%20Atlanta%20downhill%20139.png

[monzaaddict]

{B4349} ETC THROTTLE AREA CONVERSION (% per square mm)

Newbie here. After looking at this for most of the night it dawned on me that the reason redhardsupra's chart perfectly calculates the correct value for a truck throttle body and not the F/Y body throttle body is because the F/Y body throttle body in addition to the throttle blade has that bypass passage (this passage is not present on the truck throttle body). The fudge factor in the F/Y body tune is to account for the bypassed air.