Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
 Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
 |
« Reply #60 on: May 02, 2011, 11:09:22 PM »
|
|
|
mtx, and orienz, I just sent you two another test release.
I reduced the fast init inter byte time from 20ms to 13ms. I think the previous releases that used the 20ms inter byte time may have been causing problems because it was right at the max time. If this doesn't work, then we will probably need to try again at 19ms or something.
I'm not sure what could be causing the slow init problem. We are reading the sync byte now because I made it wait longer than the spec requires to receive it, but we are receiving the wrong sync byte. I removed a redundant change in the USB time outs that I am hoping may remove some potential noise in the signal when reading the sync byte.
|
|
|
|
|
Logged
|
|
|
|
|
mtx-electronics
|
|
« Reply #61 on: May 03, 2011, 06:38:47 AM »
|
|
|
Here are the logs.
|
|
|
|
|
Logged
|
|
|
|
Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
|
« Reply #62 on: May 03, 2011, 10:48:02 AM »
|
|
|
You're awesome MTX. Thanks for taking the logs.
Slow init appears to work, except that the ECU appears to be sending the 0x55 sync byte with a baud rate of 9732 instead of 10400. According to the spec you are supposed to use the 0x55 sync byte to determine the baud rate. I haven't gotten my baud rate detection code to work yet though. I will take a look at it again and maybe I can at least hard code some specific combinations.
Fast init does not appear to get a response with the 12ms inter byte timings. I will bump that back up to 19ms for the next test release.
|
|
|
|
|
Logged
|
|
|
|
Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
|
« Reply #63 on: May 03, 2011, 01:48:10 PM »
|
|
|
Uwe just informed me that it is normal for the EDC15 to respond to a slow init with a 9600 baud sync byte. I will add a special case to support this baud rate in the next test release.
|
|
|
|
|
Logged
|
|
|
|
|
mtx-electronics
|
|
« Reply #64 on: May 03, 2011, 03:34:03 PM »
|
|
|
I'm not surprised to see soch behaviour... It seems that these EDC15 ecu's don't have very elastic kwp2000 protocal specs.
I've got that Fiat EDC15C7 ecu that will not communicate with interbyte time over 2,5ms and find this very strange.
|
|
|
|
|
Logged
|
|
|
|
Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
|
« Reply #65 on: May 03, 2011, 03:57:02 PM »
|
|
|
I've got that Fiat EDC15C7 ecu that will not communicate with interbyte time over 2,5ms and find this very strange.
So it won't respond to the StartCommunication request with an interbyte time longer than 2.5ms? |
|
|
|
|
Logged
|
|
|
|
|
mtx-electronics
|
|
« Reply #66 on: May 03, 2011, 04:28:41 PM »
|
|
|
Yah, I know it's stange  I also logged a popular flashing tools wakeup procedure and they are using 1-2ms on this ecu too. The cool thing is that with timming params. & init. diag. mode I can set the ecu to 125,000bps without interbyte and download flash content in 1,5min which is not back considering that my Galep4 programmer read's the AM29F400BT in parallel mode in 58sec. |
|
|
|
|
Logged
|
|
|
|
Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
|
« Reply #67 on: May 03, 2011, 04:56:42 PM »
|
|
|
Is this the same ECU that requires 20ms inter byte time to do fast init?
|
|
|
|
|
Logged
|
|
|
|
|
mtx-electronics
|
|
« Reply #68 on: May 03, 2011, 05:01:49 PM »
|
|
|
Wait a minute.. Let's not get confused  This ecu I'm talking about is EDC15C7 that is used on Fiat/ALfa/Lancia. The 20ms on fast init. is a EDC15P+ used on Seat/Skoda/Audi/VW. I was using this ecu as a comparison to show how much different these two EDC15's are between each other. |
|
|
|
|
Logged
|
|
|
|
Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
|
« Reply #69 on: May 03, 2011, 05:59:06 PM »
|
|
|
I thought you were talking about two different ECUs, just wanted to be sure.
I really don't want to require the user to input the correct timing settings into the program to get it to connect, but ultimately that may be the only way to support all of them.
|
|
|
|
|
Logged
|
|
|
|
Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
|
« Reply #70 on: May 03, 2011, 10:31:45 PM »
|
|
|
I just sent orienz, and MTX another test release.
This one has the fast init timings using a default of 19ms for the inter byte timings.
Also, the slow init connection will fall back to 9600 baud if 10400 baud fails to read the sync byte.
|
|
|
|
|
Logged
|
|
|
|
|
mtx-electronics
|
|
« Reply #71 on: May 04, 2011, 04:38:01 PM »
|
|
|
Ok, so now slow init. finally wakeup the ecu: 05/mag/2011 12.10.44.729: USER: KWP1281 connect info: ?38906019NE 1,9l R4 EDC 0000SG 6119 ?) That's my car Cool, great work Tony... Instead for fast init. there is still some problem... Seems that the ecu starts to reply with negative response before you finish sending the crc byte. I did notice one very small detail though, if you look at the working log I did using my software for some strange reason my timing between when I sent out 0x81 and 0x01 is only 17ms while after that it correctly uses 19ms... So maybe 19ms is a little to long and 17ms would be better, take a look and tell me what you think? So where getting close to the solution now. BTW after slow init. nefmoto disconnects why is that? This is the first time I use nefmoto so I don't know what to expect... P.S. I'm supposing that even if the protocol from here onward is the similar to ME7 specs. what about the secury key?? I would guess that they used a different algo for these ecu's right? |
|
|
|
|
Logged
|
|
|
|
Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
|
« Reply #72 on: May 04, 2011, 05:49:35 PM »
|
|
|
Is it possible to get a longer log for the slow init? After trying two slow inits with address 0x01, I try a slow init on address 0x11. The logs say that there was no sync byte received for that slow init, but the logic analyzer logs didn't record that far. I am wondering if the ECU switched baud rate or something.
With the fast init, I don't know why the ECU responds before receiving the checksum byte at the end of the message. It may be caused by an inter byte time out, and like you suggest 17ms could work better. The ECU also responds to my fast init by sending a message to address 0x00, when the tester address the start communication message comes from is set to 0xF1. I will try changing the inter byte time to 17ms, and also try disabling the dumb mode test 0x55 byte I send before doing any fast or slow inits.
|
|
|
|
|
Logged
|
|
|
|
nitroCZ
Newbie
Karma: +3/-0
Offline
Posts: 4
|
|
« Reply #73 on: May 05, 2011, 05:54:46 AM »
|
|
|
Are you tested in car or on board?When will you trying to connect to EDC ECU fast pull down and up supply voltage,maybe help it
|
|
|
|
|
Logged
|
|
|
|
Tony@NefMoto
Administrator
Hero Member
Karma: +132/-4
Offline
Posts: 1389
2001.5 Audi S4 Stage 3
|
|
« Reply #74 on: May 06, 2011, 11:56:41 AM »
|
|
|
orienz, and mtx, I just sent you two another test release.
Aside from a few bug fixes, this release also has another tab to allow you to adjust the communication settings. This way you can try out any P4 inter byte time you want. These settings only take effect when a connection is started, and will have no effect on a connection that is already open.
I would recommend testing different P4 inter byte times, as well as testing with "verify cable in dumb mode" turned off.
Please let me know if this helps.
|
|
|
|
|
Logged
|
|
|
|
|
