Notices
Tuning, Diagnostics, Electronics, and Wiring HP Tuners | EFILive | Hand Held Programmers | Stand Alone PCM's | Electronics | Wiring Diagrams

Smoothing VE tables ??

Thread Tools
 
Search this Thread
 
Old 06-20-2006, 11:25 AM
  #1  
TECH Addict
Thread Starter
 
Yelo's Avatar
 
Join Date: Feb 2003
Location: Geneseo, NY
Posts: 2,864
Likes: 0
Received 1 Like on 1 Post
Default Smoothing VE tables ??

I've been looking at all of these VE tables people have been posting lately and I'm wondering if my policy of getting a VE table as smooth as possible within the min/max or fixed area limits is a good or a bad thing ??

My thinking is that an engine doesn't have "peaky" airflow or rapid changes in airflow patterns so why should your VE table....my table is naturally pretty smooth from all of the tuning I've done on it but I always use polynomial smoothing on an area after I've made changes to it to ensure it's close to the other cells around it. I just can't ever see an engine changing by 5-10% efficiency within the space of 1-2 cells....

Opinions ??
Old 06-20-2006, 12:29 PM
  #2  
Moderately Differentiated
iTrader: (4)
 
dewmanshu's Avatar
 
Join Date: Nov 2004
Location: Maryland
Posts: 27,563
Received 3 Likes on 3 Posts
Default

I agree, but after logging for a week it's hard to ignore what the vehicle wants. I think with ve tuning with no filtering you'll find alot more "radical" spikes and valleys. Just by filtering out any TPS below 5% and increasing the cell count to 5 required (HPT advises 10) I get a more consistent and smoother result. Imagine a 10 count and some type of TPS filter.
Old 06-20-2006, 01:59 PM
  #3  
LS1TECH Sponsor
iTrader: (4)
 
Magnus's Avatar
 
Join Date: Nov 2001
Location: Chicago, IL
Posts: 2,402
Likes: 0
Received 0 Likes on 0 Posts
Default

When possible, try to hand smooth the table rather than using the poly smooth on the entire table..

Hand smoothing is very powerful. Just click the point you want to smooth on the 3D graph and drag it to where you want it to be.

If you see a odd spike that looks out of place, lower it a pinch, and raise the surrounding cells to compensate.

Also, when using the paste special, make sure you have a decent min frame hits value in your histogram config. A cell that has only a couple of hits could end up having a high value like 20% or -25% depending on what the vehicle was doing at that point, if you paste that into your VE table you will create a spike.

Load all your data in the histogram and look at the cell hit counter (the C up at the top). It will give you a good idea of what cells are actually GOOD data and what cells are just oddballs that really should be filtered out.

- Keith, HP Tuners
Old 06-20-2006, 02:06 PM
  #4  
TECH Junkie
iTrader: (30)
 
12secSS's Avatar
 
Join Date: Nov 2001
Posts: 3,728
Received 1 Like on 1 Post
Default

Yup, I agree with Keith entirely! I always use the 2d Graph to smooth out my transistions and from the various vehicles I have tuned, people comment on how smooth the power curve is. You need to take into consideration abrupt fuel transitions when looking at a log file, those fuel transitions are not good results to use. They are basically the ECM commanding more/less fuel based on the TPS% (or other variable) and the air flow not being able to instantly change. Just keep smoothing your VE table and all will be good. I apply this theory to my timing tables as well.

__________________
George P. Lara: 2001 Camaro SS #0391
HPTuners for '98-'02 F-Bodies & '99-'02 Trucks



Jenibella Racing Crew Chief
1998 Camaro Z28: 9.94 @ 137, 1.45
Old 06-20-2006, 02:16 PM
  #5  
TECH Addict
Thread Starter
 
Yelo's Avatar
 
Join Date: Feb 2003
Location: Geneseo, NY
Posts: 2,864
Likes: 0
Received 1 Like on 1 Post
Default

Thanks Keith !!

I've only ever used poly smoothing on a area that I've been working on...smoothing the entire table seems like it would produce wacky results. Hand smoothing is DEFINITELY the way to go, I use hand smoothing all the time to correct minor variances in places on the table

I use the cell count all the time to determine whether the date I've gotten is a fluke - or the real deal. I'd be scared to tune without looking at the cell count because ANYTHING can happen once, if it happens again and again you can start to believe it's really happening
Old 06-20-2006, 04:32 PM
  #6  
TECH Apprentice
iTrader: (1)
 
1slowsilverado's Avatar
 
Join Date: Nov 2005
Location: Daytona Beach
Posts: 310
Likes: 0
Received 0 Likes on 0 Posts
Default



Hand smooth with 3d graph or elmanual with the table, I dont use poly smooth entire, But I will poly smooth between horizantol bounds when I start to get close. Wash, Rinse, Repeat x by % from this point. I try to hit as many cells as I can, With smooth throottle tranistions an ignoring any thing low map on decel esp in the higher rpms. I like to get a real good average, I personly wont trust anything with less than 30+ counts witch can still be off. Most are in the 100's. Dont log right away with a hot engine drive around for 5 mins or so avoid any type of heat soak..
Old 06-20-2006, 05:00 PM
  #7  
Moderately Differentiated
iTrader: (4)
 
dewmanshu's Avatar
 
Join Date: Nov 2004
Location: Maryland
Posts: 27,563
Received 3 Likes on 3 Posts
Default

Originally Posted by 1slowsilverado
Dont log right away with a hot engine drive around for 5 mins or so avoid any type of heat soak..
I bet this is the biggest culprit with people who are trying to learn.
Old 06-21-2006, 02:50 PM
  #8  
TECH Junkie
iTrader: (30)
 
12secSS's Avatar
 
Join Date: Nov 2001
Posts: 3,728
Received 1 Like on 1 Post
Default

Or vice versa, allow the engine to come up to operating temp, and not be in cold start.
Old 06-21-2006, 03:17 PM
  #9  
TECH Junkie
iTrader: (2)
 
02sierraz71_5.3's Avatar
 
Join Date: Oct 2004
Location: Cornelius, NC
Posts: 3,473
Likes: 0
Received 1 Like on 1 Post
Default

Throttle transitions are smoother and it helps with keeping trims from jumping all over. I hand smooth then bump up the decimal to .00001 and hit the polynomial button. Repeat process. Adjusting the decimal to max resolution makes a big big difference in getting a baby butt smooth VE that works well otherwise if youve just got a resolution of .1 and hit polynomial it will make adjustments that are too big.

we are adjusting by percentages so if you see a +7 trim that actually translates to .070, this is why adjusting the decimal place is important for smoothing.

my .02

BTW: Its nice to know the HP tuners guys look at the truck website too. Thanks for the great product.

Last edited by 02sierraz71_5.3; 06-21-2006 at 03:24 PM.
Old 06-21-2006, 04:46 PM
  #10  
blownerator
iTrader: (20)
 
BlownChevy's Avatar
 
Join Date: Mar 1986
Location: Chatsworth, CA
Posts: 18,745
Likes: 0
Received 7 Likes on 7 Posts
Default

Soooooooooooo I guess this is not good?:



Quick Reply: Smoothing VE tables ??



All times are GMT -5. The time now is 11:04 AM.