Replies to This Discussion

Permalink Reply by Andrew Radford on December 21, 2011 at 12:43pm

There is only the one checksum for a Mavlink message - there is no other inside the payload if that's what you mean?

I assume you are actually discarding failed crc packets?

• ▶ Reply

Permalink Reply by Nicolas Brieger on December 21, 2011 at 2:58pm

Hey Andrew,

thanks for your reply!
It was my understanding that the discarding of failed crc packets was handled by the mavlink builtin function mavlink_parse_char (inside protocol.h)? There is explicitly a section inside which drops packets that fail the crc check (under the MAVLINK_PARSE_STATE_GOT_CRC1 case). The return value of this function is only true if a complete message could be decoded successfully and thats the only case where I inspect the payload.
Thus, I modify my original question: Does the official mavlink implementation handle the crc check and I have a hardware failure or do I have to explicitly do more checksum testing?

~Nicolas

• ▶ Reply

Permalink Reply by Andrew Radford on December 21, 2011 at 6:09pm

If you are using the generated implementation of mavlink then it should be - although I've no personal experience with the C version. 

I've been writing a C#/.net version. I've got this working with silverlight, and have written as proof of concept a mavlink 'debugger' which allows easy inspection of mavlink logs. It may be of use to you. The website for this is here:

http://mavlinkpacketviewer.codeplex.com/

I've a test deployment here: 

http://mandrolic.com/mavlink/MavlinkDebuggerTestPage.html

However what's in the repo is more recent.

• ▶ Reply

Permalink Reply by Nicolas Brieger on December 23, 2011 at 4:03am

Your project really looks interesting. I checked the logs generated by QGroundControl and today, they show other issues... I am coming to think that indeed my hardware is broken.

Thanks for your help, anyway!

• ▶ Reply