Optical SPDIF<->AES Converter, am I doin' right? Circuit check.

[Yansi]

Hi!

it's me again and my dirty prototypes. This time, a situation has come so I need an (optical) SPDIF to AES converter, to test a device with an AES/EBU interface. (SPDIF being the virgin interface on my sound card - now coming handy for the first time!)

Could you please check that my approach is correct and validate the circuit? Description of circuit is following, although there is not much to it.

I have selected a pair of TOSLINK connectors from CLIFF, should not expect issues there (maybe apart from the footprints).  The  TOSLINK provides (or is provided) a standard 5V TTL signal.

For the AES, I use the PE-65612 transformers, that are specifically designed for this purpose, so I do not expect problems here either.  There is just the simple interface with an optional attenuator on the input side, series RC network for optional emphasis (cable correction). The output side is just AC coupled transformer and simple resistive impedance match for the transmitter.

What may be wrong is my use of the classic SN75179 (datasheet here).  I think it probably could be used for this purpose, as long as it is fast enough. (Otherwise I would need to use his expensive brother SN65HVD179 (datasheet here)). I think it also fulfills the requirement for the voltage levels involved, including the 200mV minimum signal level (the 65HVD179 may be even better). I dunno of any different neat solution for this, other than using off the shelf RS422 transceiver.  The purpose of R12 and R13 is to establish reasonable common mode voltage on the input.

The circuit is powered by a simple 5V linear regulator. Nothing special. I try to be generous on the supply decoupling side of thing.

Is there anything plain wrong, did I miss something important? Or do you think my ideas are valid?




What the circuit certainly lacks is some ESD protection on the XLR side and a bit of RFI suppression on the output. Maybe a simple SOT23 sized TVS should do just fine and a pair of ferrite beads on the output too.

Thank you for any recommendations and help,
Y.

[Yansi]

Never mind, I have already sent the PCB to be manufactured.

[ptrn]

Out of curiosity, did the boards work as intended?

[Yansi]

Hello, sorry for not responding any quicker, I did not have much time left to even check the forum regularly during this year.  Yes, the converter is definitely working, tested with about for sure at least 50 meters of random audio XLR cables (so not even impedance matched high frequency cables). Yet it still worked.

Although, I do not think the transmit side have been done correctly, with the 485 line driver and the series termination of R8, R9, in an attempt to create a ~100 ohm source impedance. If I would design something like this next time, I would definitely think about improvements there. Maybe, it is enough to drive the line directly,  relying only on the termination on the load side (receiver). No idea, afaik I could not even loacate any reliably looking AES3 driver schematics for reference.

Receiver part could also get improvements, for example adding a passive high frequency boost, to counteract high frequency attenuation in a long cable. Just an idea.

Regards,
Y