Clock… again =(

[Drawmer]

Hi guys, it's me (again).
Again, I'm fighting the clock I can not synchronize with the signal AES.
It was like this:
192 (1) as the master clock - 8 channels adat ----> Mixer Ramsa DA-7
192 (2) as a slave clock - 8 channel TDIF ------> Mixer Ramsa DA-7
Two 192 interfaces are clocked via BNC
192 (1) clock output -------> 192 (2) clock input & 192 (2) clock output -------> 192 (1) clock input.
Everything worked perfectly!

When I decided add a third interface and get extra 8 AES channels to mixer, I could not do it
I clocked it like existing ones (loop clocking).
192 (1) clock output -------> 192 (2) clock input & 192 (2) clock output -------> 192 (3) clock input & 192 (3) clock output -------> 192 (1) clock input.
But no sound in mixer from AES channels. Adat & tdif sounds ok, adat and tdif slots marked "lock" in mixer's screen, but aes slot marked "unlock".
I think maybe my mixer need an extra BNC-sync especially for AES channels?
192 (1) clock output -------> 192 (2) clock input & 192 (2) clock output -------> Mixer clock input & Mixer clock output -------> 192 (1) clock input.
Now mixer is clocked through adat. If I plug it in to the clock loop BNC like interfaces? Now mixer synchronized through adat and sounds well.
Or plug mixer to x256 clock?

[JFreak]

You must not have any other boxes in the loop sync, besides Digidesign/Avid HD interfaces. That's a rule #1. Every other box needs to be clocked from the ext.clock output.

Now I'm not sure, because I don't use TDIF, but I suspect that setting one of the interfaces to TDIF makes the ext.clock run a 256x superclock. I don't know whether it makes *all* interfaces output superclock or just that one. But in any case, you need to know if your digital devices require normal (1x) clock or the superclock (256x). Once you've sorted that out, you know what to do:

1) run wordclock from ext.clock of the interface that uses TDIF to all boxes that require the 256x superclock
2) run wordclock from ext.clock of the interfaces that DO NOT USE TDIF to all boxes that require 1x normal clock
3) in case the boxes requiring normal clock don't get lock, split the loop sync which always outputs 1x normal clock. DO NOT put the other boxes to the loop, but instead use a passive T-splitter and "steal" the word clock from there.

Then report back.

[Bill Denton]

And don't forget termination...

[JFreak]

Originally Posted by Bill Denton

And don't forget termination...

Yep. Passive splitting and forgetting termination drop signal level so things just might barely work or barely not work because either or both of these... BUT, wordclock is a digital signal so as long as the signal is "good enough" things keep working -- just that you don't ever know how close to disaster you're working, it just either works or not.

[Drawmer]

Thank you!
I did everything as you wrote. Rechecked everything again, thought again, and decided that your diagram is correct. Everything SHOULD work. Adat and TDIF work fine, AES does not work.
I did a few more experiments with equipment and connections, and came to a simple conclusion: the mixer's AES-board is faulty. Some malfunction in board. I'm sure this on 95%!
I have a Solomonic decision: I just replace the wrong AES-board to the TDIF-board. This format is already running, there is no reason to reinvent the wheel).

Thank you very much for your advice!

[the dude]

Originally Posted by JFreak

Yep. Passive splitting and forgetting termination drop signal level so things just might barely work or barely not work because either or both of these... BUT, wordclock is a digital signal so as long as the signal is "good enough" things keep working -- just that you don't ever know how close to disaster you're working, it just either works or not.

Word clock is TTL not a digital audio signal

[JFreak]

Originally Posted by the dude

Word clock is TTL not a digital audio signal

I said digital signal, not audio.