Page 2 of 2
Posted: Mon Aug 21, 2006 9:08 am
by shaodome
coolio
Posted: Tue Aug 22, 2006 5:50 am
by Fastest95PGT
Little status update.
Good news and bad news. The code change I had Mad Prof make seems to work well for him as mentioned already. However, when I got home I tried it with an older BT/USB adaptor I have (the original one I tried and never got to work) and the fix does not work.
I haven't had time to debug it with some com logging yet, but I hope to soon. Once I get my older adaptor to work (assuming I can), I'll have Mad Prof test the final code change on his.
Mad Prof has confirmed that his current code changes still work with the standard serial port cabled method as well.
Posted: Wed Aug 23, 2006 5:08 am
by Mad Professor
As: Fastest95PGT has said we have been playing about with the source code, and for my bluetooth link it now seems that the RTS & DTR are not the problem or well looks like it is not.
We have found in the msComm.cpp around line 242 if we change a line,
From:
Code: Select all
timeout.ReadTotalTimeoutConstant = _totalReadTimeout; // 25 for MS-I, 250 for MS-II
To:
Code: Select all
timeout.ReadTotalTimeoutConstant = 200;
That MegaTune stops jumping about, and I not longer get any resets, ect, like I was getting in my
MegaSquirt & Bluetooth Post.
I my self don't know C++, I am more QB & VB Based, So I don't know if this is the best way of getting around my program or not.
Thanks for your time.
Posted: Wed Aug 23, 2006 5:17 am
by Fastest95PGT
The way I see it. As long as the change I had you make works for you, then that's what matters. Posting it here allows Eric/anyone the oportunity to review/comment accordingly on how it will affect other communications.
This change does not seem to affect normal RS232 serial comms at least with a straight through serial cable.
MT appears to write and read on a "per character" basis whereas MTX does it in chunks. This could be the reason why MTX did not have the same issues MT did.
As for my older adaptor, I'm in the process of upgrading it's drivers (the ones I had were from '03) in hopes that will work with this modified code. If so, the fix will be proved as more than just a "fluke" that it worked for Mad Prof above. I will also then test the code with a straight through serial cable, USB/Serial, and on board USB.