problems everywhere lol
Thread Starter
Joined: Apr 2003
Posts: 1,580
Likes: 0
From: Houston, TX
problems everywhere lol
sorry guys. I just suck at computers. I was trying Jack's load thingy for evoscan. Not much luck w/ that either. It "digested" a log and output a small portion of the log with the load added (yay!), but it was only up to 85 samples... significantly less than were in the log. Also, it did output the file... but not cleanly. I got an error (which coincidentally I sometimes also get with evoscan) for which I have attached the screenshot and details thereof. If anyone can help, remember that I suck at all this xml whatever stuff lol.
click here to help me lol
click here to help me lol
You are dropping the excel file in the wrong spot. You need to drop it exactly where it has the letters and not in the white empty area. I got that error the first time I tried it also because I dropped it in wrong spot
Thread Starter
Joined: Apr 2003
Posts: 1,580
Likes: 0
From: Houston, TX
lol, yeah I did that initially, but no I got it in the right spot and I'm still getting that error. Like I said, it does make the file... but it's truncated and I get that error at the same time... just not exactly as I'da thought it would have happened. I thought there would be some sort of confirmation, but perhaps it's not there b/c of the error screen.
There's two possibilities.. A problem with the source file, or you exited the program before it was complete.. The file is not complete until you see the bottom box show the new filename.... If it doesnt and you exit the program, you'll get a partial file. And the program is drag and drop, there was a reason for that, basically you can take several files and drag and drop them to convert them, and the file is written to the same location where the source came from..
Again, if you exit the program before its done (it should finish before it lets you exit, but since I forgot to disable the quit button, and I never added a dialog for when the files are completed) I wrote this tool for myself, so I never took error handling or user error into account.. I can set it up to disable the exit button and show a status of some sort (or a completion message) if you guys want.. Like I said, I wrote this tool for myself so I did very little to make it very intuitive, just easy for my "workflow".. Thats why it hovers on top of everything and is drag and drop..
Again, if you exit the program before its done (it should finish before it lets you exit, but since I forgot to disable the quit button, and I never added a dialog for when the files are completed) I wrote this tool for myself, so I never took error handling or user error into account.. I can set it up to disable the exit button and show a status of some sort (or a completion message) if you guys want.. Like I said, I wrote this tool for myself so I did very little to make it very intuitive, just easy for my "workflow".. Thats why it hovers on top of everything and is drag and drop..
Last edited by MalibuJack; Sep 4, 2006 at 07:38 PM.
Thread Starter
Joined: Apr 2003
Posts: 1,580
Likes: 0
From: Houston, TX
I understand, but now I see that the error message is probably what cut the program short and caused a truncated file. So what do I do about the error message? That seems to be the problem. Again HERE's the link to the zip file containing a screenshot and a .txt of the error details.
thanks all.
thanks all.
looks to me like a value was blank and it wanted a decimal format number in there.
anywhere youve left blank that you could enter 0 (zero) or a value for?
just based on the error. its a runtime error generated because the code did not handle the exception. I know it has expected a decimal and gotten a blank value because of the type conversion failure.
Any IT questions feel free to throw my way am pretty experienced engineer
but ive barely touched EcuFlash until now so give me time to learn hehe
if you have no values left blank or everythings as expected let the writers of ecuflash know and theyll have to sort it out with a patch or in the next version man
Trav
anywhere youve left blank that you could enter 0 (zero) or a value for?
just based on the error. its a runtime error generated because the code did not handle the exception. I know it has expected a decimal and gotten a blank value because of the type conversion failure.
Any IT questions feel free to throw my way am pretty experienced engineer
but ive barely touched EcuFlash until now so give me time to learn hehe
if you have no values left blank or everythings as expected let the writers of ecuflash know and theyll have to sort it out with a patch or in the next version manTrav
Trav, I hereby accuse you of being an IT engineer and demand your help with disassembling this blasted Evo ROM
You do not have to say anything blah blah blah but it is driving me nuts and I could use some help.
You have a PM!
You do not have to say anything blah blah blah but it is driving me nuts and I could use some help.You have a PM!
Trending Topics
Originally Posted by jcsbanks
Trav, I hereby accuse you of being an IT engineer and demand your help with disassembling this blasted Evo ROM You do not have to say anything blah blah blah but it is driving me nuts and I could use some help.
You have a PM!
You have a PM!
I'm definately quite experienced in C and years in security and some protocol work, I'm not familiar with some of the evo ROM concepts yet so let me bring myself up to speed over a week or 2 and ill definately do anything i can to help
My tactrix cable or what-not is on its way, along with reflash jumper block etc
Originally Posted by Honki24
hmmm, there are blank values, I'll try to fix that. Thanks man! I'll letcha know.

If youve entered everything as expected, then ECUFlash could be doing something buggy internally and hence the same runtime type conversion error will occur. In which case there's nothing you can personally do other than submit the error to the ECUFlash authors and see if they can assist, along with the bin or hex you are trying to apply/save if this is possible.
Sorry I couldnt be more specific help,
Trav
Originally Posted by honki24
sorry guys. I just suck at computers. I was trying Jack's load thingy for evoscan. Not much luck w/ that either. It "digested" a log and output a small portion of the log with the load added (yay!), but it was only up to 85 samples... significantly less than were in the log. Also, it did output the file... but not cleanly. I got an error (which coincidentally I sometimes also get with evoscan) for which I have attached the screenshot and details thereof. If anyone can help, remember that I suck at all this xml whatever stuff lol.
click here to help me lol
click here to help me lol
Honestly, I never had a problem determining load based on ignition timing tracing, but this will be a good tool for the newbies to help them get a feel for what is going on if it remains accurate.
I posted this in the Log Analyzer thread on my forum.. But understand I wrote the program for my own use, and made several assumptions because of that.. I didn't make any extra effort to handle errors that I knew didn't happen in my workflow.
Although the program runs fine as long as you select the right columns, it might have a problem if you ONLY select 3 or 4 columns and the last column shown is either AFRMap or InjPulsewidth.. This was a coding error (shortcut) taken on my part to speed up the digestion process.. My workflow does turn on ALL of the log data since I do get a very high sample rate with my laptop.. But I have turned several off out of curiosity and had no trouble with it.
What I did see happen FREQUENTLY was if someone opened the CSV file in Excel, do something with the data, then re-save it.. it WILL NOT WORK, partly because Excel reformats the date and time columns slightly which is different than the original Evoscan log format. If you want to work with the data in Excel, and need to cut/trim/save, you should do it after converting it. (Again, that was not part of my workflow so I never initially noticed that Excel reformatted it)
You don't have to have all the columns enabled, and it won't croak on blank data, however it will choke if some of the columns that are needed are missing..
The load calculation is most useful in determining load if you have several adjacent cells with similar values, are experiencing knock, or your octane number isn't 255 (for whatever reason..) Conditions like that would make "hand tracing" quite a bit more difficult and time consuming. But I do actually do that on occasion if its necessary.
I have gotten some word that the functions that the analyzer does, will be part of the next release of Evoscan, I think he also mentioned he fixed the logging so the sample format (time stamp format) is easier to bring into data log lab, so you won't need this tool any longer at that point..
Although the program runs fine as long as you select the right columns, it might have a problem if you ONLY select 3 or 4 columns and the last column shown is either AFRMap or InjPulsewidth.. This was a coding error (shortcut) taken on my part to speed up the digestion process.. My workflow does turn on ALL of the log data since I do get a very high sample rate with my laptop.. But I have turned several off out of curiosity and had no trouble with it.
What I did see happen FREQUENTLY was if someone opened the CSV file in Excel, do something with the data, then re-save it.. it WILL NOT WORK, partly because Excel reformats the date and time columns slightly which is different than the original Evoscan log format. If you want to work with the data in Excel, and need to cut/trim/save, you should do it after converting it. (Again, that was not part of my workflow so I never initially noticed that Excel reformatted it)
You don't have to have all the columns enabled, and it won't croak on blank data, however it will choke if some of the columns that are needed are missing..
The load calculation is most useful in determining load if you have several adjacent cells with similar values, are experiencing knock, or your octane number isn't 255 (for whatever reason..) Conditions like that would make "hand tracing" quite a bit more difficult and time consuming. But I do actually do that on occasion if its necessary.
I have gotten some word that the functions that the analyzer does, will be part of the next release of Evoscan, I think he also mentioned he fixed the logging so the sample format (time stamp format) is easier to bring into data log lab, so you won't need this tool any longer at that point..
FWIW the only time I ever saw any "Decimal" errors was when the data was opened in excel then resaved as a CSV..
For it to crap out after 85 or so samples is just strange since my logs are typically 5mb in size, are you including the header line (first line of the log) ? That line is required for the program to determine the right format.. I also realized I made an assumption about what column the time stamp data is in.. and I saw some of you guys had different output at times for those columns..
Don't forget I also have a Data.XML file that I included with the program.. It was only tested with that file. If your using your own I can't be sure of anything its doing since you could have duplicated column names, different order, etc..
For it to crap out after 85 or so samples is just strange since my logs are typically 5mb in size, are you including the header line (first line of the log) ? That line is required for the program to determine the right format.. I also realized I made an assumption about what column the time stamp data is in.. and I saw some of you guys had different output at times for those columns..
Don't forget I also have a Data.XML file that I included with the program.. It was only tested with that file. If your using your own I can't be sure of anything its doing since you could have duplicated column names, different order, etc..
Last edited by MalibuJack; Sep 7, 2006 at 08:52 AM.
Thread
Thread Starter
Forum
Replies
Last Post
fostytou
ECU Flash
2
Mar 16, 2009 08:33 PM




