15 USB/SPDIF converters shootout

[Bob Stern]

I read what Dan [Lavry] posted. Let's just say I have not seen many sources that have a rise time of 10 nSec, or slower.

Can't speak for what others do, but a lot of folks use the LVC logic family. I can assure you it is much faster than 10 nSec. Faster signals allow those of us who design this stuff to see more of the effect of reflections. That is just plain RF engineering.

Some of us can measure the effects of cable length, and reflections, on the recovered clock. (Getting rid of SPDIF would solve that problem.)

 

Two comments:

 

1. Even if the rise time of the S/PDIF signal is 1 ns, even a 1-inch long connector would have a propagation delay of only 1/6 this amount, so its characteristic impedance should not be important.

 

2. What counts is not whether you can see on a scope distortion of the S/PDIF waveform due to reflections, but whether ameliorating such reflections actually reduces jitter in the recovered clock. Both the S/PDIF input receiver and the clock recovery phase-locked loop include low-pass filtering and interpolation that reduce the effects of distortion components at frequencies orders of magnitude above the clock frequency. To determine whether such effects remain, one would have to actually measure the jitter spectrum of the recovered clock with different S/PDIF cables and connectors.

[Bob Stern]

My preceding post is not intended to contradict ar-t's report that he actually did measurements to prove that the jitter spectrum of the recovered clock is affected by differences in the reflections caused by different S/PDIF cables and connectors.

 

I merely intended to supplement ar-t's post with some engineering background for the possible benefit of non-engineers here.

 

I would be interested to know whether ar-t is referring to measurements he did years ago with S/PDIF input receivers and clock recovery PLLs that are no longer state of the art, or whether even today's very best clock recovery circuits remain sensitive to differences among S/PDIF cables and connectors.

[Timothy Hoang]

Regarding your Berkeley Audio Design Alpha USB®, in the rear output of BNC SPDIF, It is stated on their website to have 75 ohm impedance, but why does it have a center insulator?

 

The last time i checked BNC Jacks with central insulators were 50 Ohms.

 

Tim

Artemis

[andrex99]

Audiobyte Hydra: High-End USB interface | One man's journey into audio world

 

looks like an intresting new converter for reasonable price..quality clocks, battery powered aswell external usb.cannot find Much comparitions though.

[ar-t]

Two comments:

 

1. Even if the rise time of the S/PDIF signal is 1 ns, even a 1-inch long connector would have a propagation delay of only 1/6 this amount, so its characteristic impedance should not be important.

 

So you say................

 

2. What counts is not whether you can see on a scope distortion of the S/PDIF waveform due to reflections, but whether ameliorating such reflections actually reduces jitter in the recovered clock. Both the S/PDIF input receiver and the clock recovery phase-locked loop include low-pass filtering and interpolation that reduce the effects of distortion components at frequencies orders of magnitude above the clock frequency. To determine whether such effects remain, one would have to actually measure the jitter spectrum of the recovered clock with different S/PDIF cables and connectors.

 

And you assume we can't?

[ar-t]

Regarding your Berkeley Audio Design Alpha USB®, in the rear output of BNC SPDIF, It is stated on their website to have 75 ohm impedance, but why does it have a center insulator?

 

The last time i checked BNC Jacks with central insulators were 50 Ohms.

 

Tim

Artemis

 

Some 75 ohm BNCs have a very thin central insulator, on the female end.

[ar-t]

WBT 110 are ideal for digital being 75 ohm to 1GHz. The WBT 102 is 75 ohm to 200MHz.

 

And you believe them, just because it is in their data sheet?

[ar-t]

My preceding post is not intended to contradict ar-t's report that he actually did measurements to prove that the jitter spectrum of the recovered clock is affected by differences in the reflections caused by different S/PDIF cables and connectors.

 

I merely intended to supplement ar-t's post with some engineering background for the possible benefit of non-engineers here.

 

I would be interested to know whether ar-t is referring to measurements he did years ago with S/PDIF input receivers and clock recovery PLLs that are no longer state of the art, or whether even today's very best clock recovery circuits remain sensitive to differences among S/PDIF cables and connectors.

 

Well, now that you put it that way..........

 

We have much better gear, to measure with today. At some point, I may measure the latest gizmo that Crystal makes. They still use a Schmitt trigger, so my guess is "yes". Maybe not as much, as they claim the recovery circuit is better.

 

But, to no one's surprise, I remain skeptical.

 

What no one seems to understand is we are not talking about tons of difference, between the two. If you are trying to make the best that can be, it makes no sense to overlook such a small detail. The amount of jitter in the recovered clock will always be much worse than the source. Even if you do fix the front-end.

 

We can, so we do. Simple as that.

 

Others will claim it doesn't exist, then it exists, but isn't important (or can't work). At least until they decided that they invented it, and incorporate it in all their so-called products.

[cpvniii]

Yes, but it's a firewire Dac called the Weiss 202!

Are there any DACs at this time, at any price, whose USB input is not bettered by a first-rate external USB to S/PDIF converter?

[bhobba]

Are there any DACs at this time, at any price, whose USB input is not bettered by a first-rate external USB to S/PDIF converter?

 

Yea the Playbnack Designs. Its a matter of opinion if it's improved. I think it is slightly improved by an Off-Ramp but the difference is so slight I can't be bothered using it and some think its internal is slightly better. Although everyone agrees there is hardly anything in it.

 

Another is a PDX - again very close - hardly anything in it.

 

Thanks

Bill

[bhobba]

Yes, but it's a firewire Dac called the Weiss 202!

 

Sorry mate - to my ears my Off-Ramp clobbered it. But it was better than a Stello.

 

But then again I am not a Wiess 202 fan - its a bit flat and reserved for my tastes (those that know me know I like a forward in your face sound) - it may grow on you with time though - I only spent a small amount of time with it at a dealer and not in a system I am familiar with.

 

Thanks

Bill

[cpvniii]

The off ramp? Into what DAC? The question I responded to was "sans" USB converter ! The Weiss 202 is anything but flat!

Maybe you are thinking of the Minerva or the Dac 2 ?

[Audio_ELF]

Sorry ... deleted

[DanRubin]

Yea the Playbnack Designs. Its a matter of opinion if it's improved. I think it is slightly improved by an Off-Ramp but the difference is so slight I can't be bothered using it and some think its internal is slightly better. Although everyone agrees there is hardly anything in it.

 

Bill

Anyone know if the USB inputs built into the best DACs (Meitner, MSB, dCS, PBD, Bricasti) convert to S/PDIF or do they go straight to I2S? If the former, the superiority of external converters can be attributed to isolation. But if the latter, it seems inconceivable to me that the the built-ins are not regularly trouncing the external converters.

 

Dan

[Superdad]

Anyone know if the USB inputs built into the best DACs (Meitner, MSB, dCS, PBD, Bricasti) convert to S/PDIF or do they go straight to I2S? If the former, the superiority of external converters can be attributed to isolation. But if the latter, it seems inconceivable to me that the the built-ins are not regularly trouncing the external converters.

 

Dan

 

No DAC with USB input that I know of converts to S/PDIF internally since that would then require converting back to I2S (or other word alignments depending upon the DAC chip's input). It would be ridiculous if some designer did that!

 

I don't have time at the moment to go into it, but there are a host of reasons that decently done external USB>S/PDIF converters often sound better that a DACs own USB input.

 

Yet it is true, a well done internal USB>I2S board should (and often does in the DIY circles) sound better that an external USB>S/PDIF. It is kind of obvious: USB>I2S>DAC chip versus USB>S/PDIF>S/PDIF receiver>I2S>DAC chip.

And of course clocks and jitter and power supplies and isolation all involved in all cases.

[DanRubin]

I don't have time at the moment to go into it, but there are a host of reasons that decently done external USB>S/PDIF converters often sound better that a DACs own USB input.

Perhaps when you have more time. I'd really like to know what those reasons are and what it takes to overcome them.

[mav52]

No DAC with USB input that I know of converts to S/PDIF internally since that would then require converting back to I2S (or other word alignments depending upon the DAC chip's input). It would be ridiculous if some designer did that!

 

.

I could be wrong, but I thought the little Concero did that. CONCERO | Resonessence

[bhobba]

The off ramp? Into what DAC? The question I responded to was "sans" USB converter ! The Weiss 202 is anything but flat!

Maybe you are thinking of the Minerva or the Dac 2 ?

 

The Off-Ramp into a Wiess 202 compared to its Firewire. Its Firewire is good - better than a Stello we tried - but the Off-Ramp was better again.

 

Mate how DAC's sound varies considerably between systems. My personal reference DAC, a PDX, is very hyper detailed, dynamic and in your face - so much so in some very revealing systems some find it un-listenable - describing it like ultra-violence:

NAD M51 Digital Direct DAC – initial impressions | Digital Audio Review by John Darko

 

On my system it's very detailed and in your face but not to the point anyone says they don't like it. I have a NAD M51 as well and everyone that heard it in my system easily prefers the PDX over the M51.

 

I took the PDX around to some other guys I know to compare it to a Metrum on their systems. One was a home made open baffle Seas Excell system and another SGR active speakers. On those systems the PDX was dull, slow and boring. Normally I give the PDX the edge over the Metrum - but on those systems the Metrum was clealy better. Exactly why - who knows. But the fact is how DAC's sound shows enormous variation depending on system synergies.

 

That said I am not the only one to say what I did about the Wiess - and indeed I heard it on a very unfamiliar system at a dealer so could easily change my mind hearing it on other gear.

 

Thanks

Bill

[bhobba]

No DAC with USB input that I know of converts to S/PDIF internally since that would then require converting back to I2S (or other word alignments depending upon the DAC chip's input). It would be ridiculous if some designer did that!

 

Errrr - exactly what DAC's you know 100% for sure do I2S internally?

 

I hate to tell you but most in fact do SPDIF internally. They use an inexpensive chip that is available that has both analogue and SPDIF out - but no I2S and simply feed it into the SPDIF. USB in the past was an after thought which is why external converters were usually universally way above internal implementations. That is slowly starting to change - but as of now don't expect any but the very best USB DAC's to convert to I2S.

 

Thanks

Bill

[bhobba]

I could be wrong, but I thought the little Concero did that. CONCERO | Resonessence

 

You aren't wrong - most do it that way. The reason is its easy to implement with a readily available off the shelf chip.

 

Thanks

Bill

[bhobba]

Anyone know if the USB inputs built into the best DACs (Meitner, MSB, dCS, PBD, Bricasti) convert to S/PDIF or do they go straight to I2S?

 

Hi Dan

 

Most don't - they use a cheap chip that converts to SPDIF.

 

And indeed that is a reason external converters sound better.

 

The other thing to realize is many of the big makers have engineers that think the stuff audiophiles like us go on about is in cloud cuckoo land and it doesn't make any difference anyway.

 

Thanks

Bill

[junker]

I suppose as much as I believe you... Do you have a legitimate reason to not?

 

http://www.wbtusa.com/pdf/0110.pdf

 

And you believe them, just because it is in their data sheet?

[dallasjustice]

Rca connectors are not 75 ohm connectors. You would need good bnc connectors. Yes, Pat has measured and posted them in various places.

I suppose as much as I believe you... Do you have a legitimate reason to not?

 

http://www.wbtusa.com/pdf/0110.pdf

[junker]

I'd think that you would have to know that their specs are inaccurate, otherwise I suppose it could be considered libel if those assertions were untrue...

 

The 110s have a thin conductor though polymer components to maintain 75 ohms through 1GHz. The metal bodied units only maintain this characteristic impedance through 200MHz, I believe is what they are spec'ing.

[junker]

Also, here is the international patent on the design:

 

Espacenet - Bibliographic data