Another digital transformer/interface question

[ble0t]

I've been digging through some old threads looking for information on how to 'clean up' a coaxial SPDIF connection and came across this post as well as this post.

I am currently using my computer as my transport (yea yea...I know), so modifying things at that end is not really a great option. Would it be beneficial to combine these two circuits at the DAC end? My thought was to use a transformer (such as this one) immediately followed by the differential circuit.

Any thoughts/suggestions/criticisms?

 

[ble0t]

I suppose it's worth mentioning my DAC uses a CS8414...it's very similar to Yeo's Monica 2 DAC.

 

[lucpes]

I've been digging through some old threads looking for information on how to 'clean up' a coaxial SPDIF connection and came across this post as well as this post.

I am currently using my computer as my transport (yea yea...I know), so modifying things at that end is not really a great option. Would it be beneficial to combine these two circuits at the DAC end? My thought was to use a transformer (such as this one) immediately followed by the differential circuit.

Any thoughts/suggestions/criticisms?

Get a better sound card and start from there, period. Starting with a cleaner transport in the 1st place is best. Clock on the transport side is critical for the overall final result.

There are LOTS of things to try inside the PC...

Replacing connectors (even on the PC side) with BNC ones, adding output transformer if you don't have any on the PC side, tracing the digital output on the soundcard AFTER the pulse transformer if it has one and bridging any cheap X7R SMD caps in the path, replacing all the capacitors in the digital-related supplies on the soundcard with SANYO OSCON SP or with the new SEP - sounds better IMO.

Replacing 5V&3.3V filtering capacitors on the mainboard with OSCONS also helps keep the switching PS noise out of the equation.

Check your PC power supply, most capacitors inside fail after a couple of years of continous use, change them for say Panasonic FC. Add extra polypropylene caps bypasses on the underside of the supply's PCB while you're at it.

Use a soundcard that supports ASIO and a corresponding player.

I'm fiddling with digital output from the PC with various DACs for two years now and so far the most improvement came from buying a better sound card (eg. EMU 1212m ~$200). It will save you from lots of headaches and mistrials in this regard

 

[Jocko Homo]

Well, you found a good transformer, but Schott is no longer around. There are second source vendors, not sure that they make that exact one. Pulse Engineering makes similar stuff; not as good, but usable. I hate Scientific Conversions.

And yes, on the RX side, use a linear differential circuit between it and that crappy Crystal RX chip.

Jocko

 

[ble0t]

lucpes: Modifying the sound card itself would be fairly difficult seeing as how it is of the built-in variety (based on the ALC850). I'll see what I can do for the other components onboard, but I'm always hesitant to mess around too much with the motherboard (especially after I finally got everything stable/working well). I'll definitely look at improving the PSU though.

jocko: Schott did get bought out, but it appears the product is still the same? They appear to still be around in some form (http://www.schottcorp.com)?

As for the RX chip....*sigh*....my only excuse is I built the DAC before I knew any better

 

[Jocko Homo]

I use a different Schott transformer. The "new" version seems to be a little bit better. If better is measured by lower leakage inductance. In my world, it is.

I don't know the story with Schott and their site. A buddy said he sent them an e-mail, and someone answered 3 months or so later.

As for the RX chip........no one seems to know. But now you do, so it is easy to drop something in front of it.

Jocko

 

[gmarsh]

after reading all the hate for the "Crystal RX chips", is there a better 192KHz-compatible chip out there than the CS8416?

TI doesn't make the 96KHz DIR series anymore, and I can't find any places in north america that sell AKM stuff in small quantities...

 

[Jocko Homo]

Charles Hansen uses some AKM stuff in his products. He might know of a source.

Jocko

 

[Elso Kwak]

I've been digging through some old threads looking for information on how to 'clean up' a coaxial SPDIF connection and came across this post as well as this post.

I am currently using my computer as my transport (yea yea...I know), so modifying things at that end is not really a great option. Would it be beneficial to combine these two circuits at the DAC end? My thought was to use a transformer (such as this one) immediately followed by the differential circuit.

Any thoughts/suggestions/criticisms?

Hi Omit C2 & C3 in that logic frontend for the CS841*.
Hope you get it working as I was not too happy with it.
You can also try a comparator. The diodes are only for static protection.

 

[ble0t]

Yes...in further reading through those threads, it seems that the general consensus (if there ever is one ) is to leave out those coupling caps.

Perhaps I'll do a little test on both circuits....gotta get to Digikey and order some parts anyways...thanks!

 

[ble0t]

Here's a shot of the circuit I've come up with. I've set all the unused logic gates to VDD and left all the unused outputs open. Does this look OK or is there anything I should do differently?

 

[Jocko Homo]

Oy.........

Why IC1B? Why the XOR gates? You already have differential signal; you don't need to reinvent it with XOR gates.

Hint: If you insist on using IC1B, move it, and ditch the XOR gates.

BTW..........you need HCU04s....................and while you are it...........get some picogates.

Jocko.........not advancing the casue of anything new and revolutionary.