I did some logging......inputs?
Aug 3, 2006, 02:33 PM
#46
I would love to get into disassembly of the ROM (done some 68000 and 6502 assembly back in the day LOL - and some more recent Atmel AVR high level compiled stuff - and the SH2 looks quite similar), but whilst people have mentioned the subroutines, and I've disassembled some of the ROMs, it can be tricky to work out what addresses are being referenced with the various offset addressing modes. It would be great if more information were shared on this making it really an open source project, rather than just free software, which great as it is is a bit of a closed book to develop further.
A wiki of the 900 or so subroutines would be great...
Number one on my list would be to half the airflow frequency that is output to allow logging beyond 5000 RPM (and I'm hardly trying on a stock turbo).
Number two would be dumping that RAM and seeing what is going on.
I would love to read DSMlink as well.
A wiki of the 900 or so subroutines would be great...
Number one on my list would be to half the airflow frequency that is output to allow logging beyond 5000 RPM (and I'm hardly trying on a stock turbo).
Number two would be dumping that RAM and seeing what is going on.
I would love to read DSMlink as well.
Last edited by jcsbanks; Aug 3, 2006 at 02:37 PM.
Aug 3, 2006, 02:41 PM
#47
Originally Posted by jcsbanks
I would love to read DSMlink as well.
I'm not sure if you will find all the information that you are looking for about the ECU by looking at the app's source code, but it will be fun looking through it nonetheless, to see how they accomplished what they did.
Eric
Aug 3, 2006, 05:26 PM
#49
Originally Posted by l2r99gst
nj1266,
You don't have to believe me....I'm not forcing anything on you. I'm just telling your the truth. Take is as you may.
But, if you put your thinking cap on her for a second and think logically, you can come to the conclusion that I am correct, without even knowing me.
The guys that wrote the pocket logger basically had the ability to log timing and knock counts. That's all they had to determine how much timing was being retarded by the stock ECU. The pocketlogger is just a logger....it doesn't display timing maps or anything like that in the ECU. So, these guys were even worse off than we are now with the Evo ECU. We can do what they did and have the advantage of looking at the stock timing maps. Even so, we can't completely say when and how much timing is being pulled, until we get a definitive load formula and do map tracing. I hope to be able to do this for the Evo ECU soon, once we can log mass airflow.
But, this has already been done, and more, with DSMLink for the DSM ECU. There is no guessing here. The pocketlogger guys took a shot, and came up with what they though was going on. I commend them for the efforts. But, the DSMLink guys disassembled the ECU (like is happending now with the Evo ECU) and told us EXACTLY what is going on. And, to top it off, we can log it for proof.
It's really quite straightforward...I don't know why you have such a hard time understanding. I know you don't have access to the DSMLink forums, so you have to trust what I say, but just research DSMLink a bit, if you don't believe me.
Eric
You don't have to believe me....I'm not forcing anything on you. I'm just telling your the truth. Take is as you may.
But, if you put your thinking cap on her for a second and think logically, you can come to the conclusion that I am correct, without even knowing me.
The guys that wrote the pocket logger basically had the ability to log timing and knock counts. That's all they had to determine how much timing was being retarded by the stock ECU. The pocketlogger is just a logger....it doesn't display timing maps or anything like that in the ECU. So, these guys were even worse off than we are now with the Evo ECU. We can do what they did and have the advantage of looking at the stock timing maps. Even so, we can't completely say when and how much timing is being pulled, until we get a definitive load formula and do map tracing. I hope to be able to do this for the Evo ECU soon, once we can log mass airflow.
But, this has already been done, and more, with DSMLink for the DSM ECU. There is no guessing here. The pocketlogger guys took a shot, and came up with what they though was going on. I commend them for the efforts. But, the DSMLink guys disassembled the ECU (like is happending now with the Evo ECU) and told us EXACTLY what is going on. And, to top it off, we can log it for proof.
It's really quite straightforward...I don't know why you have such a hard time understanding. I know you don't have access to the DSMLink forums, so you have to trust what I say, but just research DSMLink a bit, if you don't believe me.
Eric
I went to the DSMlink website and found no mention of what you mentioned about how many knock sums are needed to pull 1 degree of timing. I looked and looked but could not find the info you mentioned.
I really do not know who you are as you do not know who I am. You need to post supporting material on this. Simply asking me to trust you does not work for me.
I posted where my info came from, now please give me a link so I can examine this on my own.
Aug 3, 2006, 06:51 PM
#50
nh1266,
OK, just to try to satisfy you, which this probably won't, I looked through my old DSMLink logs that has some knock in them. Here are a couple screenshots of the logs with only a few parameters checked so to help clean up the log. The first screenshot shows the cursor at an area of 1 knock count and the second shows and area of 7 knock counts. The first shows the timing being pulled as .4* (client rounds to the nearest tenth) for the 1 knock count and 2.5 * for the 7 knock counts (7 x .35*).
If you want, download and install the DSMLink client and I can send you a bunch of logs.
Edit: This site doesn't let you upload files over 100K, so the pictures look pretty crappy. Download the client and you can see them full screen.
Eric
OK, just to try to satisfy you, which this probably won't, I looked through my old DSMLink logs that has some knock in them. Here are a couple screenshots of the logs with only a few parameters checked so to help clean up the log. The first screenshot shows the cursor at an area of 1 knock count and the second shows and area of 7 knock counts. The first shows the timing being pulled as .4* (client rounds to the nearest tenth) for the 1 knock count and 2.5 * for the 7 knock counts (7 x .35*).
If you want, download and install the DSMLink client and I can send you a bunch of logs.
Edit: This site doesn't let you upload files over 100K, so the pictures look pretty crappy. Download the client and you can see them full screen.
Eric