S/PDIF in general and interfaced with ADAU DSPs

[DannerD3H]

Hello guys!
I'm back with another problem that confuses an electronics noob like me.. Its called SPDIF..

I have tried really hard to understand this signal form. Mainly the RCA/Coax version has been tricky for me. However, I think I finally understand it. I just need a sanity check from all the smart heads in here.

I'm trying to make and SPDIF interface module for my DIY ADAU1466 DSP.
The module has:

1. Toslink and RCA inputs.
2. The inputs can be selected with a switch sticking through the rear panel between the two connectors.
3. There are two LED's used to indicate which input is live.
4. The switch can also be placed in a center position(or not installed at all) to allow the DSP or a uC to select the input type.
5. RCA output.
6. It uses transformers for galvanic isolation on both RCA connectors.
7. Will be used to accept inputs from my PC(toslink) or streamers like WIIM and so on(RCA and Toslink).
8. The RCA output is mainly intended to daisy chain multiple dsp units or to connect a subwoofer.

This is what my schematic looks like. I have also added some textboxes to try and describe what (i think) is going on. Please let me know if I messed up big time or if it should work ok, Thanks!

The SPDIF_IN goes directly from here to a connector and then to the DSP boards SPDIF Input. The Output likewise to the DSP boards SPDIF Output.

The DSP core board has the ADAU's SPDIF_RX connected direcctly to a connector and the SPDIF_TX pin is terminated with an 47R resistor as it exits the DSP and then straight to the connector like this:

I have been very much in doubt if I need a buffer of sorts on the RCA input/output. But according to Analog Devices the ADAU1466 should be able to both read the input and drive the output directly with its pins. This is also confirmed by one of their EVAL boards:

Im just not sure how a 0-0.6V signal can trigger the input when it is centered at 1.65V? But I guess there is more going on behind the scene that I have no clue about?


Thanks in advance guys!

 

[DannerD3H]

This is what the module looks like BTW.

Still not 100% done with the routing so components may move and silkscreen isnt touched at all.

 

[altor]

Im just not sure how a 0-0.6V signal can trigger the input when it is centered at 1.65V? But I guess there is more going on behind the scene that I have no clue about?

Because of a series capacitor between SPDIF transformer (or input) and ADAU's input.
BTW, relay control (Q202) is totally wrong,

P.S. Why do you remove the resistor and capacitor from the transformer input and CHASSYS in the first schematic? This is useful.
C215 in the second schematic is too low.

Alex.

 

[DannerD3H]

Because of a series capacitor between SPDIF transformer (or input) and ADAU's input.

How does that work? I understand the cap blocks dc, but how can a 0.6Vpp signal get higher than 1.65v?

BTW, relay control (Q202) is totally wrong

Damnit.. still trying to get the hang of transistors, they confuse me every time..

P.S. Why do you remove the resistor and capacitor from the transformer input and CHASSYS in the first schematic?

I may as well just place them, but for a start the chassis is plastic. But will add the same parts between the chassis (from the RCA) and gnd.

C215 in the second schematic is too low.

That's the eval board. I use 100nF instead. Is that okay?

 

[altor]

How does that work? I understand the cap blocks dc, but how can a 0.6Vpp signal get higher than 1.65v?

Chip input is biased to 1.65V (w/o signal).
0.6Vpp=+/-0.3Vampl signal changes the input voltage around the bias, from (1.65-0.3)=1.35 to (1.65+0.3)=1.95V

I may as well just place them, but for a start the chassis is plastic.

Connect to GND.
It is not necessary, by w/o this RC, spdif input may be sensitive to the environment noise (for example - you turn on/off the light in the room, and hear the click in you speakers).

I use 100nF instead. Is that okay?

Yes, usually it is OK.

Alex.