Notices
Evo Dyno Tuning / Results Discuss vendor and member dyno tuning techniques, results and graphs.

Virtual Dyno is available!

Thread Tools
 
Search this Thread
 
Old Oct 7, 2010, 02:50 PM
  #181  
Account Disabled
Thread Starter
iTrader: (38)
 
Mellon Racing's Avatar
 
Join Date: Aug 2005
Location: Virginia Beach, Virginia
Posts: 9,319
Likes: 0
Received 1 Like on 1 Post
Originally Posted by bbarnhill
You have pretty much the latest copy. Im guessing the TPS trimming is working great? I was nervous about that part and the profile selection part.

After a little more testing on it and some minor changes ... it should be ready.
Originally Posted by tscompusa
that hump bug was pretty bad. there are no more of those are there?
Originally Posted by bbarnhill
there is if the data in the log shows a hump. Its with the log not VD. The RPM and time arent linear. We have looked extensively at this problem and I swear the logger developers leave a little to be desired here.
Brad it's working very good, nice work my friend.

tcomp that was the only oddity I saw in that particular log.

brad that issue is going to come up with 2byte and maybe 1byte rpm logging, would it make sense to filter out odd spikes that can't be real?
Old Oct 7, 2010, 05:25 PM
  #182  
Evolved Member
iTrader: (1)
 
bbarnhill's Avatar
 
Join Date: Oct 2008
Location: Kansas City, KS
Posts: 692
Likes: 0
Received 1 Like on 1 Post
I dont have a way to logically determine if a spike is abnormal or not. Any ideas? 1.0.7 knocks at some of the humps ...
Old Oct 7, 2010, 07:08 PM
  #183  
Former Sponsor
iTrader: (12)
 
Boosted Tuning's Avatar
 
Join Date: Jun 2005
Location: Chico, CA (Nor-Cal)
Posts: 2,383
Likes: 0
Received 7 Likes on 5 Posts
Originally Posted by Mellon Tuning
Brad it's working very good, nice work my friend.

tcomp that was the only oddity I saw in that particular log.

brad that issue is going to come up with 2byte and maybe 1byte rpm logging, would it make sense to filter out odd spikes that can't be real?
2byte rpm can have spikes just like 2byte load, but 1byte load and rpm rarely have spikes.

Last edited by Boosted Tuning; Oct 7, 2010 at 07:15 PM. Reason: grammer
Old Oct 7, 2010, 07:33 PM
  #184  
Evolved Member
iTrader: (1)
 
bbarnhill's Avatar
 
Join Date: Oct 2008
Location: Kansas City, KS
Posts: 692
Likes: 0
Received 1 Like on 1 Post
Im at a loss for how to trim spikes out. I dont know how to tell if they are real power gains or false. I know we can tell by looking at them but how can I convert that into logical statements? Hum ... the delimia
Old Oct 7, 2010, 07:36 PM
  #185  
Former Sponsor
iTrader: (12)
 
Boosted Tuning's Avatar
 
Join Date: Jun 2005
Location: Chico, CA (Nor-Cal)
Posts: 2,383
Likes: 0
Received 7 Likes on 5 Posts
Originally Posted by bbarnhill
Im at a loss for how to trim spikes out. I dont know how to tell if they are real power gains or false. I know we can tell by looking at them but how can I convert that into logical statements? Hum ... the delimia
I always use regular 1byte RPM (you only need 2byte if you rev over 796X) and I never every have any rpm spikes.

So I wouldnt be that worried. V7 beta 4 is working great IMHO.
Old Oct 8, 2010, 08:49 AM
  #186  
Account Disabled
Thread Starter
iTrader: (38)
 
Mellon Racing's Avatar
 
Join Date: Aug 2005
Location: Virginia Beach, Virginia
Posts: 9,319
Likes: 0
Received 1 Like on 1 Post
I rev to 9-9500 and a few of my customers do as well with larger turbo and built motors, but I agree this minor issue isn't a show stopper. Just something I noticed
Old Oct 8, 2010, 12:08 PM
  #187  
Evolved Member
iTrader: (1)
 
bbarnhill's Avatar
 
Join Date: Oct 2008
Location: Kansas City, KS
Posts: 692
Likes: 0
Received 1 Like on 1 Post
I dont know what to do about the two byte rpm problem. Any ideas? Anyone thinj we should get with hamish and see what we can do? Or
is this something that is an ecu issue? Where is the proper place to fix it as i dont mind working with whoever as long as we dont patch this up. Lets hit it at the source.
Old Oct 8, 2010, 12:17 PM
  #188  
Evolved Member
iTrader: (2)
 
l2r99gst's Avatar
 
Join Date: Mar 2004
Location: CA
Posts: 3,499
Likes: 0
Received 4 Likes on 4 Posts
It's a logging speed issue. There's not much that you can do about it while logging MUT. 2 byte RPM will have spikes anytime the first byte and second byte aren't in the same 'step' so to speak.

Logically, you may be able to write code to look at RPM dips or raises in the range of the step value (which will be 1000RPM) and it will always happen on multiples of 1000RPM.

Last edited by l2r99gst; Oct 8, 2010 at 12:23 PM.
Old Oct 8, 2010, 12:30 PM
  #189  
Former Sponsor
iTrader: (12)
 
Boosted Tuning's Avatar
 
Join Date: Jun 2005
Location: Chico, CA (Nor-Cal)
Posts: 2,383
Likes: 0
Received 7 Likes on 5 Posts
Originally Posted by l2r99gst
It's a logging speed issue. There's not much that you can do about it while logging MUT. 2 byte RPM will have spikes anytime the first byte and second byte aren't in the same 'step' so to speak.

Logically, you may be able to write code to look at RPM dips or raises in the range of the step value (which will be 1000RPM) and it will always happen on multiples of 1000RPM.
Eric, do you think theres any way to change 2byte RPM into a "1byte RPM" like we did with load?
Old Oct 8, 2010, 12:56 PM
  #190  
Evolved Member
iTrader: (2)
 
l2r99gst's Avatar
 
Join Date: Mar 2004
Location: CA
Posts: 3,499
Likes: 0
Received 4 Likes on 4 Posts
Originally Posted by Boosted Tuning
Eric, do you think theres any way to change 2byte RPM into a "1byte RPM" like we did with load?
That has to be coded into the ROM. Tephra did that for 1 byte load.
Old Oct 8, 2010, 12:58 PM
  #191  
Former Sponsor
iTrader: (12)
 
Boosted Tuning's Avatar
 
Join Date: Jun 2005
Location: Chico, CA (Nor-Cal)
Posts: 2,383
Likes: 0
Received 7 Likes on 5 Posts
Originally Posted by l2r99gst
That has to be coded into the ROM. Tephra did that for 1 byte load.
Yeah, I know it has to be modded into the ROM. But it could be done right?
Old Oct 8, 2010, 01:00 PM
  #192  
Evolved Member
iTrader: (2)
 
l2r99gst's Avatar
 
Join Date: Mar 2004
Location: CA
Posts: 3,499
Likes: 0
Received 4 Likes on 4 Posts
Sure.
Old Oct 27, 2010, 07:35 AM
  #193  
Account Disabled
Thread Starter
iTrader: (38)
 
Mellon Racing's Avatar
 
Join Date: Aug 2005
Location: Virginia Beach, Virginia
Posts: 9,319
Likes: 0
Received 1 Like on 1 Post
version 1.7 is up! as always, uninstall the old before installing the new
http://www.mellontuning.com/index.ph...dex&cPath=1_28

1.0.7 (October 26, 2010)
- Fixed a bug that would cause some resizing issues and save the wrong size and locations
- Fixed a bug that would not properly read CSV files with commas in the data or data was enclosed in quotes
- Fixed flickering on smoothing level selection
- Added RomRaider AEM UEGO support
- Added 2000-2009 Honda S2000 (6 speed) support
- Added Engine Load as a boost column for AEM logs
- Added print dialog to select printer to print to when the print button is clicked
- Added Hydra EMS log support
- Added GReddy EManage log support
- Added HHH log support
- Corrected Evo X incorrect weights corrected
- Corrected 07 STI Rear drive from 3.545 -> 3.900
- Corrected 02-05 WRX Wagon drag coefficient 0.35 -> 0.33
- Moved profiles to menu dropdown selection
- Increased graphable runs to 6
- Active profile is displayed on the bottom status bar
- If more runs are listed than can be displayed the side bar is now scrollable
- Profiles moved to the menu at the top of program
- TPS trimming is done automatically and no longer requires a setting
- Seperate HP/kW and lbft/Nm settings so any combination can be used
- Double buffering enabled to cut down on flicker when drawing graphs
- Show run level messages out beside the run control
- Enhanced error messages (E101-E102) to show file names causing errors
- Mustang Dyno correction factor changed to 0.95 to better match average MD
- Software version number printed on graph
- Using InstallShield now to install
- Code cleanup and unused code removed
- Performance analysis done to find bottlenecks
Old Oct 30, 2010, 11:45 AM
  #194  
Evolved Member
 
leecavturbo's Avatar
 
Join Date: Mar 2008
Location: uk
Posts: 1,392
Received 2 Likes on 2 Posts
do you have to trim out the non 100% tps data?
anyone got an example datalog i can look at?
Old Oct 30, 2010, 01:25 PM
  #195  
Evolved Member
iTrader: (1)
 
bbarnhill's Avatar
 
Join Date: Oct 2008
Location: Kansas City, KS
Posts: 692
Likes: 0
Received 1 Like on 1 Post
no you dont have to trim out the less than 100% TPS. TPS trimming is automatic now. Just please spare me the hassle of trying to graph a part throttle run. I assume you are smart enough to use a WOT run I cant handle that part in the code, atleast not with this algorithm.


Quick Reply: Virtual Dyno is available!



All times are GMT -7. The time now is 10:21 PM.