Mitsubishi Evo4-EvoX and Subaru EvoScan Datalogging Project

User topics relating to software that logs data from ECUs

Moderator: Freon

Postby Jon [in CT] » Fri Sep 29, 2006 6:12 pm

Evo4Mad wrote:the LC-1 support doesn't work with the LMA3 device attached,
Good job.
Jon [in CT]
 
Posts: 352
Joined: Sat Jan 01, 2005 10:23 am

Postby dryad » Sat Sep 30, 2006 4:37 am

I know this might be a shot in the dark here, but is it a possibility for the Turbo XS Tuner Pro wideband support?
dryad
 
Posts: 8
Joined: Tue Jun 20, 2006 6:43 am

Postby hexman » Sat Sep 30, 2006 11:10 am

Evo4Mad

I tried to send you an LC-1 log but it failed. Your disk quota has been used up for email.

I will try again. Essentially if I only logged AFR the app would not error. It was only capturing data every so often. The values were off too. Ranged from 19 to 0. Should have been 14.7 (idle)
hexman
 
Posts: 23
Joined: Fri Apr 21, 2006 5:08 pm

Postby Evo4Mad » Sat Sep 30, 2006 3:11 pm

thanks hexman, yes the LC1 is quite slow in comparision to how fast EvoScan is.

yes you will see values of 19.x which is what the Innovate LC-1 is outputting, either the Innovate Spec for calc A/F is incorrect, or the LC-1 settings are incorrect.

I have sent an email asking Innovate to correct it.
Evo4Mad
 
Posts: 332
Joined: Mon Jun 13, 2005 11:58 pm
Location: New Zealand

Postby hexman » Sat Sep 30, 2006 3:53 pm

This Innovate forum is pretty good. You need to ask Klaus about the issues. He seems to be the man for Innovate tech issues. He frequents the forum often and is quite knowledgeable and helpful.

http://www.innovatemotorsports.com/forums/index.php
hexman
 
Posts: 23
Joined: Fri Apr 21, 2006 5:08 pm

Postby hexman » Sat Sep 30, 2006 5:29 pm

It is wierd from my log i got a 19.644 value and it took about 3 seconds to get that value again. In between there were 0's and 3.xx. Something is odd here. The rate should be much higher.

Found this on the Innovate Forum:

Me? I don't think it would be cool for me to do that in detail. If you have the protocol document, you should be able to do it without much trouble.


Big Hint: as each byte comes in, look at the bits that must be 1 and those that must be 0 in certain words, and you'll be able to tell what part of the packet they are. Synch on the header (throw away data until you see the header... that's in case you start reading in midstream). Save header and other recognized words until you have a complete packet, then handle the other bits as per the document. I believe there is one error in the PDF... a certain bit is 0 that I expected to be 1. It could also be an error in my comments (in my code) that were written based on the PDF.

And this:

The digital output runs at a constant 12.28 speed.
The packet rate is actually 800kHz divided by 65536.
The LC-1/LM-1 read the AFR data from the sensor at the fastest rate the individual sensor can support at the current condition (speed of the sensor is dependent on Lambda, fastest at stoich). This can be up to 300 - 400 times/second on the LC-1. If the analog outs are set to "instant", that's also the update speed of the analog out. For the serial output the data is collected and averaged over the packet rate.

Regards,
Klaus

Here is that thread:

http://www.innovatemotorsports.com/foru ... ample+rate
hexman
 
Posts: 23
Joined: Fri Apr 21, 2006 5:08 pm

Postby NoQuarter » Sat Sep 30, 2006 5:43 pm

Evo4Mad: I don't know if your reading the LC1 the same way as me, but this may help.

If the LC1 is sending a value that represents 14.7 AFR, You should be seeing a value of 500 being sent from the LC1.

Take the 500, divide by 1000 = .5
take .5 + .5 = 1
take 1 * 14. 7 = 14.7 AFR

So, AFR = ((LC1Reading/1000) + .5) * 14.7

This comes from using 1024 steps to represents the AFR range of 7.35 to 22.39.
NoQuarter
 
Posts: 12
Joined: Mon Aug 14, 2006 10:12 am

Postby evoralph » Sat Sep 30, 2006 7:15 pm

NoQuarter wrote:Evo4Mad: I don't know if your reading the LC1 the same way as me, but this may help.

If the LC1 is sending a value that represents 14.7 AFR, You should be seeing a value of 500 being sent from the LC1.

Take the 500, divide by 1000 = .5
take .5 + .5 = 1
take 1 * 14. 7 = 14.7 AFR

So, AFR = ((LC1Reading/1000) + .5) * 14.7

This comes from using 1024 steps to represents the AFR range of 7.35 to 22.39.


Nice dude. :D
evoralph
 
Posts: 10
Joined: Wed Sep 06, 2006 2:10 pm

Postby hexman » Sun Oct 01, 2006 5:34 am

Here is the formula from Innovate's protocol sheet

AFR=((L12..L0)+500)*(AF7..0)/10000

The "L" and "AF" values are subpackets.

L is the lambda value and AF is 14.7

Somehow I think the problem is in the reading of the packet. Capturing of subpackets is in the wrong sequence or the wrong bits are being captured. Reading the packet midstream?

I am getting a value every .10 of a second or so which is about what the rate should be. However it goes from 19.xxx to 0 to 3.xxx which seems to indicate that the wrong bits are coming into play. The value should stay in a much narrower range.
hexman
 
Posts: 23
Joined: Fri Apr 21, 2006 5:08 pm

Postby dudical26 » Sun Oct 01, 2006 11:23 am

dryad wrote:I know this might be a shot in the dark here, but is it a possibility for the Turbo XS Tuner Pro wideband support?


I have a turboxs tuner and I am going to try and follow the steps that Evo4Mad said to use to help and integrate a wideband into the software.
dudical26
 
Posts: 13
Joined: Fri May 12, 2006 9:33 am

Postby Evo4Mad » Sun Oct 01, 2006 12:57 pm

Dont worry about posting LC-1 stuff, I already know all that stuff. your Lc-1 loggers are outputting incorrect lambda values to that spec.. Innovate has pointed out a missing packet in the logging, I'll trace it and get out a version to you guys asap.

Cheers,
Hamish.
Last edited by Evo4Mad on Sun Oct 01, 2006 1:15 pm, edited 1 time in total.
Evo4Mad
 
Posts: 332
Joined: Mon Jun 13, 2005 11:58 pm
Location: New Zealand

Postby Evo4Mad » Sun Oct 01, 2006 12:59 pm

email me your TurboXS logs and I'll add its support

email me your PLX logs and I'll add its support.
Evo4Mad
 
Posts: 332
Joined: Mon Jun 13, 2005 11:58 pm
Location: New Zealand

Postby hexman » Sun Oct 01, 2006 1:45 pm

Evo4Mad wrote:Dont worry about posting LC-1 stuff, I already know all that stuff. your Lc-1 loggers are outputting incorrect lambda values to that spec.. Innovate has pointed out a missing packet in the logging, I'll trace it and get out a version to you guys asap.

Cheers,
Hamish.


You have to admit the LC-1 stuff is more productive than the whining :wink:

Besides when we think out loud (contribute on the board) we all learn. Even though I do my share of programming, by reading and exploring this stuff (the ECU code is somewhat new to me) it widens our knowledge.
hexman
 
Posts: 23
Joined: Fri Apr 21, 2006 5:08 pm

Postby Evo4Mad » Sun Oct 01, 2006 3:50 pm

Innovate LC1 support now working and tested, party time :) , thanks to Mike Laak for testing all my versions on his LC-1 ... If you are online a lot over the next few days and wanna be a tester for a whole bunch of new features I'm adding. add me to you msn... hamish.ahern at hotmail.com thanks.

A new version will be released (emailed) this coming weekend with all the new features.
Evo4Mad
 
Posts: 332
Joined: Mon Jun 13, 2005 11:58 pm
Location: New Zealand

Postby laakness » Sun Oct 01, 2006 4:42 pm

No prob Hamish. I'm glad I can contribute to the project in some way.

A note to everyone else...it works great so far :)
laakness
 
Posts: 68
Joined: Thu Jun 15, 2006 11:00 am

PreviousNext

Return to Data Logging Software

Who is online

Users browsing this forum: No registered users and 7 guests

cron