March 20, 2010, 04:10:23 PM *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
News: AEM ECU Jumper/Extension Harnesses! Available NOW!!
 
Pages: 1 ... 3 4 [5] 6 7 ... 13
Print
Author Topic: Dyno Maths Library  (Read 86665 times)
FWombat
300+ Post Club
***
Posts: 438


« Reply #60 on: April 05, 2004, 11:05:08 AM »

Quote from: 93accordnos
Humm,  these seems to work for me, but for some reason after around 4k rpms theres a big drop off and then around 5500 rpms it comes back to where it was at 4k rpms.  smiley  smiley


Often times a big, sudden rise or fall in the power graph is due going up or down a hill while making the log...

A power spike can also be caused by a slipping clutch...

- Brian
Logged

Brian Geddes
1994 Mitsubishi 3000GT VR-4
TD05 14B turbos, custom headers/exhaust.
Currently under the knife...
93accordNOS
25+ Posts Club
*
Posts: 39


« Reply #61 on: April 05, 2004, 04:19:27 PM »

It's not a sudden rise or fall,its like just a gradual dip from around 150 to 110 back to 150.  It's definately not from unlevel ground (I did this on the GSP) and I see it on 3 different pulls.  My clutch definately isn't slipping either.  Any other ideas what it might be from?  When I get home tonight I'll try and post it up for you too see if you would like.
Logged

93accordNOS
25+ Posts Club
*
Posts: 39


« Reply #62 on: April 05, 2004, 09:35:37 PM »

Ok, I played with it some more.  Well actually I just put in the numbers same as last time and it worked better this time.  The dip is gone.  Heres my results off of a couple different pulls.  My question is, why is there such a sudden drop off after 6000rpm?  Motor spins 7k easily and pulls the whole way.  BTW the top two are with the engine speed reading at normal, the bottom two are with it at fastest.  BTW don't ask what that "dead spot" in the top two is, I have no idea, probably just my crappy laptop not wanting to log. Smiley

Logged

FWombat
300+ Post Club
***
Posts: 438


« Reply #63 on: April 05, 2004, 11:54:34 PM »

Quote from: 93accordnos
Ok, I played with it some more.  Well actually I just put in the numbers same as last time and it worked better this time.  The dip is gone.  Heres my results off of a couple different pulls.  My question is, why is there such a sudden drop off after 6000rpm?  Motor spins 7k easily and pulls the whole way.  BTW the top two are with the engine speed reading at normal, the bottom two are with it at fastest.  BTW don't ask what that "dead spot" in the top two is, I have no idea, probably just my crappy laptop not wanting to log. Smiley


If you're going to use the dyno maths library, you really need to use the internal logger.  The PC logger just isn't fast enough to get a really good sample rate.  Make sure that Engine speed is setup to log on one of the fast channels, and setup the internal logger at the fastest possible sample rate.  That will make sure you have good, clean data.

Then, once you've got logs from the internal logger, send email one to me, along with a copy of the math library you're using, and I'll tailor the filtering to your sample rate.  On the pictures you posted above, the reason the power graph comes up so late and drops off so soon is that it's overfiltered.  The filters in the library setup for the internal logger (250 samples/sec), but you're running on laptop logs (50 samples/sec, max).  Start using that internal logger!  Smiley

- Brian
Logged

Brian Geddes
1994 Mitsubishi 3000GT VR-4
TD05 14B turbos, custom headers/exhaust.
Currently under the knife...
93accordNOS
25+ Posts Club
*
Posts: 39


« Reply #64 on: April 06, 2004, 09:20:43 PM »

Ok did it with internal logging.  It looks to give more appropriate numbers, but has a lot of fluctuation.  Here's a link to download it.
http://home.comcast.net/~craibit/Abstract_from_internallog.stf
Logged

FWombat
300+ Post Club
***
Posts: 438


« Reply #65 on: April 07, 2004, 02:07:15 AM »

Quote from: 93accordnos
Ok did it with internal logging.  It looks to give more appropriate numbers, but has a lot of fluctuation.  Here's a link to download it.
http://home.comcast.net/~craibit/Abstract_from_internallog.stf


The fluctuation you see in the power curves is due to your Engine Speed (RPM) signal being noisy.  If you look very closely at Engine Speed in the areas where the power curves fluctuate, you'll see that it kind of "bounces" around...it'll go from 3000 to 3010, then back down to 3006, and up to 3012...etc.

In any case, try this library.  I turned the filtering up to smooth out the curves.  It's not perfectly smooth, but if it was perfectly smooth than any REAL peaks or dips would be masked as well, and the whole point of the library would go down the drain.  Smiley  

You'll still need to fill in appropriate weight, gear ratio, wheel sizes, etc...

- Brian


Edit:  I guess we can't attach files to posts any more.  Sad  PM me your email address, and I'll email you the new library with increased filtering.
Logged

Brian Geddes
1994 Mitsubishi 3000GT VR-4
TD05 14B turbos, custom headers/exhaust.
Currently under the knife...
v8huntr
25+ Posts Club
*
Posts: 25


« Reply #66 on: April 16, 2004, 07:30:02 AM »

i'm getting -130rwhp.  is there preset program for supra?  also how can i change car specs.
Logged
FWombat
300+ Post Club
***
Posts: 438


« Reply #67 on: April 16, 2004, 01:20:40 PM »

Quote from: v8huntr
i'm getting -130rwhp.  is there preset program for supra?  also how can i change car specs.


Go back and read.  There's no preset library for anyone (except 2nd gen VR-4s and Stealths, who can use the one I posted.)

- Brian
Logged

Brian Geddes
1994 Mitsubishi 3000GT VR-4
TD05 14B turbos, custom headers/exhaust.
Currently under the knife...
CHEETAH
25+ Posts Club
*
Posts: 68



WWW
« Reply #68 on: July 01, 2004, 09:26:52 AM »

brian, need some help, I dont have too many internal logs, i use a lot of laptop logs and i saw your recommendation to use internal for more sample rates. I get weird numbers if I try to apply your fwombat amtch calc to my log. I do not have VSS in my logs, can this still work ? what all is needed ? where can I email you my log sample where you can cater to give me hp/tw output if i give you slick size ? It doesnt not need ot be accurant, just consistent in terms of comparing each pass i make to see if the car accelerated better from more HP/TQ, etc.

feel free to PM me or email me gbrtuning@aol.com

thanks
Greg
Logged

Team Cheetah Racing
NHRA All Motor; IDRC All Motor
calvin_baank
25+ Posts Club
*
Posts: 56


« Reply #69 on: July 12, 2004, 04:46:11 AM »

can't wait to buy my aem for my fully  built b16
Logged

Calvin
visit for my project, www.CJperformance.com
Rotterdam, Netherlands
---
Builing my b16 with a 60-1 Turbo
---
silver2k1cobra
25+ Posts Club
*
Posts: 31


« Reply #70 on: July 23, 2004, 10:16:19 AM »

ok.  so there is filtering in RPS and Filtered _HP.  My question is I see different numbers in each.  on a graph what filter changes what?

I need more internal logs to work with this function and so far love what I see.  Just trying to get into the details to adjust the filtereing as needed and see if I can get usefull data from the laptop log even though it is slower.  since the only way to download the internal log is to have the car off.. nice to be able to see on the fly if changes are helping or hurting.

THanks again for all the time you have put into this.
Logged
FWombat
300+ Post Club
***
Posts: 438


« Reply #71 on: July 28, 2004, 02:35:41 PM »

Quote from: silver2k1cobra
ok.  so there is filtering in RPS and Filtered _HP.  My question is I see different numbers in each.  on a graph what filter changes what?

I need more internal logs to work with this function and so far love what I see.  Just trying to get into the details to adjust the filtereing as needed and see if I can get usefull data from the laptop log even though it is slower.  since the only way to download the internal log is to have the car off.. nice to be able to see on the fly if changes are helping or hurting.

THanks again for all the time you have put into this.


There are three places filtering is employed:
     RPS - (FILTER X Engine Speed) /60
     DeltaRPS -  (DIFFERENTIATE Y RPS)
     Filtered_HP - (FILTER Z Raw_HP)

In the default library,
     X = 50
     Y = 25
     Z = 125

I've found that these values give me good results with my internal logging of Engine Speed at the highest rate.  If you want to play around with them, feel free.  I'd recommend reducing Z first, then X , and Y as a last resort.  All three have the same effect - smoothing out the graph - they're just applied at different points in the calculation cycle.  

In general, you should be getting a graph that LOOKS like a dyno sheet.  If your graph looks smoother than a dyno sheet, you need to turn down the filtering.  In extreme cases of over-filtering (like when you apply the library to a PC-log), you'll get a graph that looks like an extremely smooth mountain.   If you're getting lots of sudden peaks and valleys in the middle of the graph, then you need more filtering.

- Brian
Logged

Brian Geddes
1994 Mitsubishi 3000GT VR-4
TD05 14B turbos, custom headers/exhaust.
Currently under the knife...
silver2k1cobra
25+ Posts Club
*
Posts: 31


« Reply #72 on: August 10, 2004, 07:14:43 PM »

Thanks, Such a easy tool.. I’m hoping that with logging to the PC with the fastest rate I can at least get estimated info back.  that way I can get real-time feed back to changes to the tune.  Hate having to turn the car off to get the internal log file.  But understand how that is the only way to get accurate data back.
Logged
93turbosi
25+ Posts Club
*
Posts: 64


« Reply #73 on: August 25, 2004, 12:53:08 PM »

lol i have -45 hp and - 35.51 tq at 6700 rpms and 15 psi of boost on a built sohc vtec.  awesome no wonder it feels so fast
Logged

David Eriquez
D-Series Turbo Ninja!
93 Civic Si

Before:
Stock D16Z6 w/ greddy turbo kit and blue box = 188.1 WHP / 167.4 WTQ @ 8psi Smiley

Now:
Built D16Z6 Eagle rods, Arias pistonsand greddy running AEM EMS and UEGO = NEED TO DYNO!
DSMotorsport
25+ Posts Club
*
Posts: 70


« Reply #74 on: October 14, 2004, 11:36:09 AM »

Just downloaded and used the math against my street pulls for 3rd gear and my datalogs from my last dragstrip pass.

seems to work perfectly for me with an Evo8.  got the gear ratio info off the web, I knew the weight from scale at the track.

Showing 297awhp and about 285-290wtq (can't see peak torque on the drag passes, as rpm doesn't get that low).

that seems pretty close to what I would be expecting on a dynojet for a 3400total weight car going 108 in the 1/4.

Im impressed...
Thanks to the maker.
Logged

03 Evo8 11.2 @ 126 Everyday Streetcar
90 Talon AWD  10.08 @ 147  
www.diamondstarmotorsport.com
AEM Performance Electronics Forum
   

 Logged
Pages: 1 ... 3 4 [5] 6 7 ... 13
Print
Jump to: