Welcome to the North American Subaru Impreza Owners Club Friday September 4, 2015
Home Forums WikiNASIOC Products Store Modifications Upgrade Garage
NASIOC
Here you can view your subscribed threads, work with private messages and edit your profile and preferences Home Registration is free! Visit the NASIOC Store NASIOC Rules Search Find other members Frequently Asked Questions Calendar Archive NASIOC Upgrade Garage Logout
Go Back   NASIOC > NASIOC Technical > Engine Management & Tuning > Open Source Reflashes

Welcome to NASIOC - The world's largest online community for Subaru enthusiasts!
Welcome to the NASIOC.com Subaru forum.

You are currently viewing our forum as a guest, which gives you limited access to view most discussions and access our other features. By joining our community, free of charge, you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content and access many other special features. Registration is free, fast and simple, so please join our community today!

If you have any problems with the registration process or your account login, please contact us.
* Registered users of the site do not see these ads.
Reply
 
Thread Tools Display Modes
Old 12-21-2007, 06:10 PM   #1
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default Knock Summary Spreadsheet

A few months ago, I posted a feedback correction counter on enginuity.org. The spreadsheet attempted to tabulate knock using feedback correction as a signal for knock. However, it didn't work that well.

Rough correction (IAM), fine learning knock correction (FLKC), and feedback correction (FC) can all signal knock. They can also occur without one happening first. In other words, FLKC can occur even before FC has ever been triggered.

Anyway, I rewrote the spreadsheet so that it now tabulates knock based upon both FLKC and FC. It DOES NOT account for knock due to IAM dropping. This spreadsheet is meant to isolate light knocking. If IAM is dropping, another method needs to be used.

Don't forget to add the Fine Learning ranges in rows 112 and 115. That data isn't copied over from Airboy's spreadsheet because it's not used in the interpolation spreadsheet.

I'm still working on the filtering logic. Any comments are appreciated.

For those who haven't used this spreadsheet before, I'll write the instructions in a week or so provided the spreadsheet is working correctly.

Here's the link to the spreadsheet (thanks to ride5000 for hosting it):

http://www.ken-gilbert.com/wrx/engin...unter%2012.zip

EDIT: INSTRUCTIONS added to version 11. Look at the INFO tab.
EDIT: Version 11 accidentally deleted module1. I added it back in version 12.
* Registered users of the site do not see these ads.

Last edited by mickeyd2005; 01-10-2008 at 12:33 PM. Reason: Added instructions
mickeyd2005 is offline   Reply With Quote
Sponsored Links
* Registered users of the site do not see these ads.
Old 12-22-2007, 01:26 PM   #2
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

This is something I'm doing by hand right now, so I would love instructions as son as possible.
williaty is offline   Reply With Quote
Old 12-22-2007, 03:32 PM   #3
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

Actually, the NA tables are different sizes than the turbo tables. Can you make a version of this with NA sized tables please?
williaty is offline   Reply With Quote
Old 12-22-2007, 10:29 PM   #4
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

The only NA rom that I have is a 2006 Tribeca. It's timing table has 15 load columns and 14 rpm rows. My spreadsheet should be compatible with it.

Like Airboy's spreadsheet, my spreadsheet has an upper limit to table size because the tables start running into each other. There is no lower limit. You can have smaller tables.

I remember seeing some JDM timing tables that were quite large, so it is possible that some roms won't fit. If that's the case, let me know how large your timing tables are.
mickeyd2005 is offline   Reply With Quote
Old 12-22-2007, 10:44 PM   #5
sense of nature
Scooby Guru
 
Member#: 112759
Join Date: Apr 2006
Chapter/Region: MWSOC
Location: In the V hole!
Vehicle:
2018 STi Mikey Tuned
SexyRP

Default

Quote:
Originally Posted by williaty View Post
This is something I'm doing by hand right now, so I would love instructions as son as possible.
ditto
sense of nature is offline   Reply With Quote
Old 12-22-2007, 11:31 PM   #6
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

The 05 NA fuel table is 11 columns wide by 20 rows high. The 05 NA timing tables are 14 columns wide by 19 rows high.

Just for reference. I'm just back from logging. I'll be trying the spreadsheet as soon as I get the email flood beaten down.
williaty is offline   Reply With Quote
Old 12-22-2007, 11:56 PM   #7
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

OK, tried to use the table for the first time and it was a bit of a rough go. I'm REALLY glad you're writing this spreadsheet, so I would be willing to do any testing or provide any data that you would find helpful to bring this thing to perfection. Just wanted to make it obvious that the list of issues isn't a complaint, it's definitely in the name of making this baby work.

OK, the problems...

1) The "copy from airboy's" function doesn't work with NA sized tables

2) The "rom" page contains two tables, "AVCS" and "WGDC" that are not applicable. I zeroed them so they couldn't affect anything, which pissed off the "Color Cells" function. The displayed error was "Min same as Max cannot color".

3) In the "Input" worksheet, everything is fine until you click "Filter Data". The spreadsheet does a little song and dance, begins to fill in the tables in the later sheets, but then goes to the VE worksheet and halts with the error "Run-time error '1004': Application-defined or object-defined error". At the time it has the problem on the screen, it has a large bank of cells in VE sheet along the left side highlighted. Clicking debug brings up a bunch of code, including the following snippet, the line referring to "row, 4" is highlighted by the debugger.
Code:
        'Copy Data over to VE Sheet
        Sheets(VE_page).Cells(row, 1) = closest_value(rpm, Sheets(rom_page).Range("A39:A68"), 29)
        Sheets(VE_page).Cells(row, 2) = closest_value(load, Sheets(rom_page).Range("B38:AE38"), 29)
        Sheets(VE_page).Cells(row, 3) = closest_value(throttle, Sheets(rom_page).Range("B75:AE75"), 29)
        Sheets(VE_page).Cells(row, 4) = Sheets(input_page).Cells(row, ListBox_WGDC.ListIndex + 1)
        Sheets(VE_page).Cells(row, 5) = Sheets(input_page).Cells(row, ListBox_MP.ListIndex + 1)
4) After adding a couple of logs using the "Add Data" option, I switch to the "Knock" worksheet. The large bank of cells to the left is highlighted but empty. The tables to the right have some cells filled in.

5) In the "Knock" worksheet, the pivot tables to the right are partially filled in, but the column and row values are very wrong for an NA car. Is there any way to type in new values for the columns and rows without messing up the underlying math in your scripting?

6) In the "Knock" worksheet, the pivot table "Quantity of Data Available" has entires as high as 2.82g/rev. This is impressive, since this car won't exceed 1.2g/rev of load

7) In the "Fine Learning" worksheet, the tables have some values filled in. However, I can't tell what it's trying to tell me, because the top over every column is a "#REF!" error display.



EDIT: Regarding #2, I copied the AVCS and WGDC tables in from a virgin copy of the spreadsheet. The thing happily colorized them. However, it still complained about "Min same as Max cannot color" and left the Base Timing table all gray. Is this because I had to add rows and subtract columns?
williaty is offline   Reply With Quote
Old 12-23-2007, 12:13 AM   #8
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

1. Yes, Airboy's spreadsheet allows for a smaller table. Just copy and paste from enginuity. A little more work to move the cells around. My spreadsheet can handle 29x29.
2 and 3. The spreadsheet wasn't intended to be used on NA cars. I need to modify it to remove the references to the turbocharger.
4.5.6. Those are leftover values from the previous run.

merchgod only recently added NA cars to the enginuity defs.
mickeyd2005 is offline   Reply With Quote
Old 12-23-2007, 12:26 AM   #9
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

Is there anything I can do to help you get this working for NA?
williaty is offline   Reply With Quote
Old 12-23-2007, 12:57 AM   #10
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

Yes, winzip the spreadsheet that you have (with your data) and email it to me. My email address is my screen name at yahoo dot com.
mickeyd2005 is offline   Reply With Quote
Old 12-23-2007, 12:57 AM   #11
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

Oh and I need your rom too. Stock or modified.
mickeyd2005 is offline   Reply With Quote
Old 12-23-2007, 01:23 AM   #12
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

OK, everything is sent off to you.
williaty is offline   Reply With Quote
Old 12-24-2007, 09:09 PM   #13
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

OK, I continued playing with this today.

My problems with the Color Cells function were resolved by typing in the columns and rows for the base timing table rather than pasting them in. Not sure what excel thinks the difference is, but it obviously matters.

My problems with the Run-time error were resolved by logging an extra parameter (IAM in this case) and lying to the spreadsheet and telling it to use the IAM column as WGDC. After that, the program ran successfully.

EDIT: My issues with the FLKC column headers turned out to be due to the fact that the ROM for my car defines headers my car can't reach, so the table wasn't drawn out that far.

The Knock sheet is a little dodgy. It gets the columns right (i.e. pulls the values properly from the Base Timing table in the ROM sheet), but there aren't enough rows. It copies the row (RPM) values properly from the Base Timing table, but only up to 4800 RPM. The last 4 RPM rows aren't displayed at all. This means any FBKC events above 4800RPM effectively get ignored.
EDIT: For some reason, feeding it a second log caused it to extend the RPM rows all the way up to redline. God only knows what I did to piss it off the first time, I'll watch this over the next couple of days.



Other than that, this spreadsheet is AWESOME! It does in 30 seconds what it takes me the better part of an hour to do by hand. This is a total life-saver. Thank you SO much for writing this!

Last edited by williaty; 12-24-2007 at 09:22 PM.
williaty is offline   Reply With Quote
Old 12-25-2007, 02:24 AM   #14
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

I took a look at the spreadsheet that you sent me. The only thing that I did was:

1. Deleted all data from row 5806 onward because the log had changed parameters.
2. For Primary Waste Gate Duty Cycle, I selected Load. Basically, you just need to pick a column. Any column. It's not used in the logic. The only reason I included it was to tabulate it.

Based upon the above, it works just as expected.

Your interpretation of Fine Learning Knock Correction is not correct. The table is tiny compared to the base timing table. The active range 7 on your rom extends from 4000 to 4800 rpm. There is no active range above 4800 rpm.

That means that FLKC can change only up to 4800 rpm. You may see FLKC above 4800 rpm but that will be a result of FLKC that was learned between 4000 and 4800 rpm. It's not active above 4800 rpm.

Looking at your rom, it does appear to work correctly.
mickeyd2005 is offline   Reply With Quote
Old 12-25-2007, 02:27 AM   #15
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

One thing that you might notice is that FLKC is more sensitive right after a ecu reset. Right after a reset, you might get FLKC all over the place but as you drive, it starts to go away.

The datalogs must be ordered chronologically.

If you have a slow computer, you can actually watch the FLKC current change as time passes.
mickeyd2005 is offline   Reply With Quote
Old 12-25-2007, 02:34 AM   #16
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

Here is a screen capture of the results based upon the data you sent me. As you can see, it is tabulating FBKC and FLKC above 4800 rpm.



The following is a screen capture of the fine learning knock correction table. Note that the maximum goes up to 4000+ rpm. That's because that's how the rom is set up.

One thing that I did notice is that the stock rom ridiculously extends the load tables all the way out to 1.80 g/rev. You will never get that high a load on a normally aspirated car. I would recommend that you compress the ranges so that the max range is 1.3 g/rev. This will give you a finer resolution for FLKC. If SOA would do us the pleasure of making the FLKC as large as the base timing table then the ecu would pretty much tune itself.

mickeyd2005 is offline   Reply With Quote
Old 12-25-2007, 02:53 AM   #17
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

Quote:
Originally Posted by mickeyd2005 View Post
1. Deleted all data from row 5806 onward because the log had changed parameters.
Yup. The second email I sent you asked you to do this because my RIO clicked the wrong button and screwed up the log. I didn't notice that until after I had already mailed it to you.

Quote:
2. For Primary Waste Gate Duty Cycle, I selected Load. Basically, you just need to pick a column. Any column. It's not used in the logic. The only reason I included it was to tabulate it.
Yup. I worked this out earlier today on my own and got it to run successfully.

Quote:
Your interpretation of Fine Learning Knock Correction is not correct. The table is tiny compared to the base timing table. The active range 7 on your rom extends from 4000 to 4800 rpm. There is no active range above 4800 rpm.

That means that FLKC can change only up to 4800 rpm. You may see FLKC above 4800 rpm but that will be a result of FLKC that was learned between 4000 and 4800 rpm. It's not active above 4800 rpm.

Looking at your rom, it does appear to work correctly.
I wasn't complaining abotu FLKC not going above 4800, I was saying that FBKC didn't go above 4800. For whatever reason, the first time I ran the spreadsheet, the FBKC table only went to 4800. The second time, it went all the way up to 6400 like it should. I have NO idea what the difference was. Hopefully, it'll work fine from now on.

Again, as I said in my edit, this is now working flawlessly for me and is INCREDIBLY helpful. Thank you SO much for writing it!


I have some questions.

1) What logic does the FBKC table use to determine wither FBKC is normal weight red or bold red?

2) On the FLKC page, the top table is the most negative value ever achieved in that cell, and the bottom table is the most recent FLKC value in that cell, right? This means I can use the difference between the two to see if FLKC is trending down, right?

3) On the FBKC page, I'm trying to figure out what the "Number of Events/Total" table represents. I though it was the percentage of the total knocks, but in some cells I have 100% while in other places there's no way to group cells together to add up to 100%

4) is there any way to determine from this spreadsheet if knock happened with caused FBKC at the same time as FLKC was being applied


Again, this is an awesome tool and is saving me SO much time.
williaty is offline   Reply With Quote
Old 12-25-2007, 02:56 AM   #18
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

Quote:
Originally Posted by mickeyd2005 View Post
One thing that I did notice is that the stock rom ridiculously extends the load tables all the way out to 1.80 g/rev. You will never get that high a load on a normally aspirated car. I would recommend that you compress the ranges so that the max range is 1.3 g/rev. This will give you a finer resolution for FLKC. If SOA would do us the pleasure of making the FLKC as large as the base timing table then the ecu would pretty much tune itself.
Man, I really miss the obvious sometimes. For 3 days, I've been looking at the FLKC columns table going "Man, why the heck did they waste so much resolution on loads I can't reach". I never thought to re-scale the damn thing.
Man, I can have my head in the clouds sometimes
williaty is offline   Reply With Quote
Old 12-25-2007, 02:57 AM   #19
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

Looking at your screenshot of the FBKC table, I notice it topping out at 6kRPM. Does the FBKC table only get generated up to the max RPM contained in the log you feed it? Cause that would explain everything.
williaty is offline   Reply With Quote
Old 12-25-2007, 11:37 PM   #20
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

The FBKC table is generated using the data that is considered "valid". If you have no data that is valid in a set range then the table will be unpopulated in that range.

In the case where you weren't getting any data above 4800 rpm, it most likely is due to the high range FLKC was set. Once it is set, all the data above 4800 rpm will be considered invalid because the FLKC will be subtracting timing from that range even though the knock that set FLKC was set between 4000 and 4800 rpm. Therefore, you can not make any decision based upon the data above 4800 rpm.

On my LGT, the stock rom sets the top range from 5800 to 5800 rpm. That insures that the top range FLKC is always zero. I don't recommend this, but that's how my stock rom is set up. I would prefer the top range go from 5800 to 6000 rpm, so that one of the FLKC ranges isn't "wasted".

If you look on the "input" page, after the data has been filtered, there will be a column with either 0, 1, 2, 3, or 4 in it. There is a legend next to the filter button that explains what each number means. This will tell you why I decided some data is invalid or valid.

On the FBKC sheet, the logic which decides which data is bold red will probably be rewritten soon. I originally set that up when knock was based ONLY on FBKC. The new spreadsheet tabulates knock based upon both FBKC and FLKC and consequently, I need to rethink my logic. You should ignore what is bold or what is not bold and use the statistics in the other tables to guide you.
mickeyd2005 is offline   Reply With Quote
Old 01-03-2008, 01:13 PM   #21
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

Bump for instructions added.
mickeyd2005 is offline   Reply With Quote
Old 01-03-2008, 02:56 PM   #22
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

What are the changes between 10 and 11?
williaty is offline   Reply With Quote
Old 01-03-2008, 03:32 PM   #23
mickeyd2005
Scooby Newbie
 
Member#: 122595
Join Date: Aug 2006
Chapter/Region: NWIC
Vehicle:
2005 LGT Wagon
ABP

Default

There's a revision log.

Instructions are the only difference.
mickeyd2005 is offline   Reply With Quote
Old 01-10-2008, 12:21 AM   #24
wombat778
Scooby Newbie
 
Member#: 166941
Join Date: Dec 2007
Default

At the risk of sounding like a complete idiot, I am having real trouble getting this version of the spreadsheet to run. Specifically, everytime I use either the Filter Data button or the Color Cells button I get a "Compile Error: Sub or Function not defined" error. It seems that the two functions Excel is complaining about are Color_Cells and closest_value.

Any thoughts on what I am doing wrong? I have tried on several different PCs, with several different versions of Excel, and all with the same error. Version 8 of the spreadsheet worked great (but it did not seem to use the Color_Cells or closest_value functions). Thanks!
wombat778 is offline   Reply With Quote
Old 01-10-2008, 12:24 AM   #25
williaty
Scooby Guru
 
Member#: 71092
Join Date: Sep 2004
Chapter/Region: MWSOC
Location: Delaware County, Ohio
Vehicle:
2005 2.5RS Wagon
Regal Blue Pearl

Default

My guess is that the way you're getting the stock timing table, or other stock table, into the spreadsheet is pissing it off. When I tried to copy and past the tables in directly, I was getting errors like the ones you're getting. I ended up typing in the load (column) and RPM (rows) headers by hand, then copy/pasting the actual values of the tables in with Paste Special->Values

After that, it ran fine.
williaty is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Help build the Subaru spreadsheet to end all spreadsheets Unabomber Motorsports 7 06-04-2007 06:03 PM
Wheel & Tire Rotational Inertia Spreadsheet Jon Bogert Brakes, Steering & Suspension 13 02-18-2006 09:16 PM
Interesting Fuel Flow Observations, and Excel Spreadsheet sponaugle Factory 2.0L Turbo Powertrain (EJ Series Factory 2.0L Turbo) 2 10-10-2002 01:54 PM
17" and 18" Wheel Spreadsheet hooperman Brakes, Steering & Suspension 0 08-11-2002 01:46 AM
('93-'01) Excel Spreadsheet Template for Repair/Mod Journal Psi-Fire Impreza Forum 0 10-01-2001 10:21 PM


All times are GMT -4. The time now is 05:31 PM.


Powered by vBulletin® Version 3.7.0
Copyright ©2000 - 2015, Jelsoft Enterprises Ltd.
Powered by Searchlight © 2015 Axivo Inc.
Copyright ©1999 - 2015, North American Subaru Impreza Owners Club, Inc.