Hi All,
Am planning to test the actual time synchronization in an AVB network - what I need to know is whether end-to-end Time Sync is happening or not, and how precise this is.
I have connected 2 xmos audio endpoints back to back and oscilloscope is connected to 1pps output pin.
How the 1pps pin in slave device is synced with respect to the master.
On what parameters basis the signal is generated on 1pps output pin ?
end-to-end Time Sync
-
- New User
- Posts: 3
- Joined: Wed Aug 12, 2015 6:25 am
-
- Experienced Member
- Posts: 66
- Joined: Fri Feb 05, 2010 12:34 pm
We have measured the synchronisation of LRCLK between two Endpoints connected to an AVB switch.
Here is the video that shows the test setup and the measurement on the Oscilloscope.
https://www.youtube.com/watch?v=bz27YdRFj4s
Here is the video that shows the test setup and the measurement on the Oscilloscope.
https://www.youtube.com/watch?v=bz27YdRFj4s
-
- XCore Expert
- Posts: 579
- Joined: Thu Nov 26, 2015 11:47 pm
As I noted here (http://www.xcore.com/forum/viewtopic.php?f=8&t=4388) the clocks lock but they can lose synch and then re-synchronize. When this happens the XMOS AVB talker channels are delayed with respect to each other by an integer number of sampling clock periods. I don't know if XMOS has recreated this issue or not or if there is any solution for it.
It is kind of a deal breaker for me because my application requires simultaneous sampling so if one channel is delayed with respect to the others it will introduce errors into the analysis.
It is kind of a deal breaker for me because my application requires simultaneous sampling so if one channel is delayed with respect to the others it will introduce errors into the analysis.