S/PDIF 1x Cinch-Out to 2x BNC-Out extension/modification

Exactly the opposite experience but my system probably is not revealing enough 🙂 The very best experience is without any external device with either USB or SPDIF interface. Just the same DAC chip as an internal DAC. It works.

3. Implement dedicated Line drivers for each BNC Output
This seems to be the optimal technical solution, although, the existing system did not have that and uses the TXP/TXN outputs with their 25Ohm source impedance - so this option seems "not required" to me.

Optimal not required.... It is not forbidden to make it technically sound certainly with 2 x 5 meters of cabling. I think I would draw a tiny PCB for it with buffers, their voltage regulator and 2 x 75 Ohm shielded PCB mount BNC connectors so just short supply and input cabling and easy integration. The nuts of the BNC connectors have proven themselves to be helpful with encasing. Also sturdy when someone trips over such a 5 meter long cable. Of course mounted IN the DSP thingie and not externally. Who needs that?
 
Last edited:
Sorry friends,

with all respect to the USB vs. SPDIF topic, this is an off topic discussion for my case: USB is neither an output option of the HDP-3 nor an input option of the DSP x-overs in my speakers.

Sorry for the blunt intervention... but thanks for your thoughts anyway!
Winfried
 
You wrote: "two BNC SPDIF Outputs may not always both be used together". This seems an important detail especially when one cable is possibly disconnected. Not wanting to be blunt but I think you could reconsider what it is you want exactly and adapt the situation to the most optimal outcome instead of keeping choices into account that maybe never occur. Or that always occur for that matter 🙂
 
Last edited:
Optimal not required.... It is not forbidden to make it technically better certainly with 2 x 5 meters of cabling. I think I would draw a tiny PCB for it with buffers, their voltage regulator and 2 x 75 Ohm shielded PCB mount BNC connectors so just short supply and input cabling and easy integration. The nuts of the BNC connectors have proven themselves to be helpful with encasing. Also sturdy when someone trips over such a 5 meter long cable.
Hello Jean-Paul,

all conceeded! There is sufficient space for a driver PCB directly behind the BNC connectors which I plan to mount on a small (prototype) board anyway. The needed buffering components could be TH or SMD. Tripping over cables is avoided by how the cabling is layed out in the room (one of the reasons for 5m length 😉 ).

Which buffer and regulator do you recommend?
What do you think about reclocking the TXR signal with two parallel HC74s with the ISCLK signal (driving CS8420 pin13) and the HC74 outputs drive the buffers?

You wrote: "two BNC SPDIF Outputs may not always both be used together". This seems an important detail especially when one cable is disconnected.
With that I meant to say that it is not guaranteed that both speaker may not always be connected at the same and therefore should be considered and seems to speak for two drivers as well.

Thanks and regards,
Winfried
 
Edit: you can invert an S/PDIF signal without changing anything, so you could just build a second instantiation of the present circuit and connect it to TXN. TXP then drives one output and TXN the other. The series resistor then has to be about 350 ohm, I don't understand why it is only 56 ohm in post #1.
This is the answer right here. The chip has a second output. If it is enabled, doing what MarcelvdG said here would be the way to go. Check the TXN pin for signal. If it's not there see how to enable it in the datasheet, then replicate the circuit that is connected to TXP. Done. Pretty interesting that the Manchester coding used by S/PDIF is not polarity dependent.

1744131702901.png
 
Unlikely to achieve 1ns output jitter with common opamps?

The S/PDIF has pretty loose implementation.

The input and output connectors can be RCA connectors. It is noi possible for an RCA connector to be a 75 ohm connector. It is a convention for the consumer. Consumers will grab any cable with an RCA, plug it and expect it to work. I bet the cable on most to all RCA to RCA connecting cables is 75 ohm cable, closer to 110 ohm.

It is a forgiving & tolerent implementation. As with many things in consumer electronics, conventions and standards are pretty sloppy in practical usage. You can "get away" with all kinds of things. I presented an experimental with some trial and error.

This is something I have experience, breaking all kinds of standards and specifications.0
 
Hello fellows,

before the discussion drifts off into "discussion on principles" space, Pros and Cons of different digital connection types are known and considered. Therefore I'm requesting to focus on solutions for my described case!

SPDIF is the connection of choice for my application and the question is how to best implement two parallel 75Ohm BNC outputs in the existing system I have. Technical implementation proposals with component identification are much appreciated and a few were kindly given already.

Thanks and Regards,
Winfried
 
Separated 75 Ohm BNC outputs.

BTW I noticed CS8420 to be still available at some sellers. I think it would be wise to have an extra one just in case.

FWIW I never heard any device with CS8420 performing OK and avoided it but heard the later ones are better. Maybe one of the DAC Meisters knows details. If this gossip is right you better find one of the last production years.
 
Last edited:
Have you considered switching to optical cable?
Yes Uriy,

thanks! I have considered and am actually building a Toslink option (but I don't have 5m optical cables yet) in parallel to the BNC 75Ohm version (for which I already have the BNC cables) discussed here.

Separated 75 Ohm BNC outputs.

... you better find one of the last production years.
OK Jean Paul,

to clarify: The CS8420 IS on the digital output board of the HDP-3, no choice here, I have what I have.

Looking into the CS8420 Datasheet, both TXN and TXP have identical output drive capability of 21mA. This matches the 5...6x 4mA capability e.g. the (often recommended) 74HCT04 chip would provide. So why not simply take the SPDIF output circuitry from TXP and duplicate on TXN as was suggested before and be done? ... or which better line driver type do you recommend instead because of the 5m cables?

Thanks and Regards,
Winfried
 
The same. Not parallel as written.

Yes you have what you have but you don’t have it anymore when the chip fails. I don’t know how sturdy it is with 2 outputs used but if it fails you may have an issue. That is why buffering could be a good idea just like having a spare IC.
 
  • Thank You
Reactions: wgh52
OK. Please have a look at this schematic. I have inluded Toslink transmitters as well, because I wanted to test both alternatives.
So, is this what's proposed as buffered BNC Outputs for my application?
Are the resistor values OK?
Are the CP & CN Capacitors required, nice to have or not needed?
Are any other components missing?

SPDIF Out extension 2.0.PNG


Thanks for your feedback!
Winfried
 
Looks OK to my amateur eyes but I have not kept up at all with extreme high performance stuff and picoseconds of jitter. Caps are necessary. CS8420 and 74HCT04 both are 5V logic but finding 5V Toslink transmitters like TOTX177 will be hard I think. Most of todays Toslink stuff is both non Toshiba and 3.3V.

FWIW I would give 74HCT04 and Toslink separate SMD regulators (if possible).

Are you sure you do not need SPDIF transformers? Both end devices are also mains fed aren't they?
 
Last edited:
Thanks for the expertise 🙂
.... Are you sure you do not need SPDIF transformers? Both end devices are also mains fed aren't they?
Yes, end devices are mains fed. Interestingly, though, the orignal Cinch SPDIF circuit (which I will disconnect from the CS8420) just does not have output transformers, unlike the (unused & disconnected) AES3 output circuits with PE-65612NL transformers. If transformers proove necessary, I may "abuse" the two available PE-65612NLs for the BNC connectors - hmm? I will reserve some board space for that.

Regards,
Winfried
 
My method is to have the PCB pads for the transformers cleverly placed so that either resistors/0 Ohm resistors or the transformers can be used. It can then be either without much penalty.

I think you want Murata DA101C. You could check if the devices at the other side have transformers. I never saw the point of having 2 of them.
 
Hi again,
You could check if the devices at the other side have transformers. I never saw the point of having 2 of them.
Ha! Actually this is a good point, especially as my driving DSP will only connect to my speaker DSP x-overs! ...and these have transformers in their coax input circuit. So, I'll leave the space for the transformers on the driver side and not populate them for now.

Regards,
Winfried
 
  • Like
Reactions: jean-paul