USB reference design and SPDif

Technical discussions related to any XMOS development kit or reference design. Eg XK-1A, sliceKIT, etc.
User avatar
RemcoW
Member
Posts: 10
Joined: Tue Jan 15, 2013 9:00 pm

Post by RemcoW »

Yes but in SPDif mode only. It does not happen in USB out mode. I can change sampling frequencies with impunity.

I did add a channel between the decouple and clockgen thread but it doesn't do anything right now. It is just there, not being used.


User avatar
RemcoW
Member
Posts: 10
Joined: Tue Jan 15, 2013 9:00 pm

Post by RemcoW »

By way of follow up, I just pulled that extra unused channel out. It did not help.
User avatar
RemcoW
Member
Posts: 10
Joined: Tue Jan 15, 2013 9:00 pm

Post by RemcoW »

I have added a fair bit of code to the design that performs calculation to figure out what to set the oscillator to. That code is in C as it has to perform a fair bit of floating point calculation, which is obviously pretty slow. The calculation is done on the fly, based on how deep the spdif bucket has been filled up and very slowly makes adjustments to the oscillator. It yields a very low jitter output.
It being slow in calculation and having no real time requirement, it was offloaded to its own thread.

A while back, it occured that perhaps that was the source of my trouble so I hard coded the oscillator parameters to known settings and took the code that made the slow adjustments out for testing purposes. Our SPDif source is an Audio Precision analyzer so I can create a very low jitter data stream and everything worked well.
It would still lock up after 7 times, though. Btw, the code that detects the stream and figures out its speed is not locked up. It keeps running.
User avatar
Ross
XCore Expert
Posts: 962
Joined: Thu Dec 10, 2009 9:20 pm
Location: Bristol, UK

Post by Ross »

The normal cause of these kinds of issues is that there are too many routes open between the two tiles causing the communication to lock up.

Is this running on the XMOS L2 board (i.e. something I could potentially re-create?)
User avatar
RemcoW
Member
Posts: 10
Joined: Tue Jan 15, 2013 9:00 pm

Post by RemcoW »

No, unfortunately that would not be a quick thing to try. It runs on our board and the clocking mechanism to the DAC and to SPDif is quite different.

That is a thought, though. Let me see if I can get my code to run on the L2 board. It will take a bit of disconnecting of stuff and not breaking things in the process.

Please stay tuned until I get that up - I really appreciate your thoughts and help.
Rem
User avatar
RemcoW
Member
Posts: 10
Joined: Tue Jan 15, 2013 9:00 pm

Post by RemcoW »

After some messing around with my project, would it be possible I send you one of my boards instead? I'll modify it such that it runs off of USB power and it hooks straight into the regular debug tool.
Post Reply