(06-Nov-2014, 10:48)Mikeeo Wrote:

(06-Nov-2014, 09:17)mauidan Wrote: I don't understand how using a 75 ohm cable into an input that should be optimized for 110 ohms is optimal for sound quality.

I mailed Thierry several times on this topic and either I had a nonsense answer or none at all. The closest Thierry came was probably when he wrote that the Devialet proprietary solutions take care of such issues.
As I said before on this forum, I do think that their recommendation for digital XLR has been one of quality (the plugs) but as you say a 110 ohm designated input with 75 ohms ???

It would be good for once if Devialet could announce clearly what they see the best solution. However I do not think they know anything at the moment as they are busy running for a bigger market.
From my standpoint I would like to link my units with AES/EBU - AES/EBU (but for some reason the AES/EBU is an input only function).

Time will tell (I hope)/Mike

(06-Nov-2014, 11:09)GuillaumeB Wrote:

(06-Nov-2014, 10:48)Mikeeo Wrote:

(06-Nov-2014, 09:17)mauidan Wrote: I don't understand how using a 75 ohm cable into an input that should be optimized for 110 ohms is optimal for sound quality.

I mailed Thierry several times on this topic and either I had a nonsense answer or none at all. The closest Thierry came was probably when he wrote that the Devialet proprietary solutions take care of such issues.
As I said before on this forum, I do think that their recommendation for digital XLR has been one of quality (the plugs) but as you say a 110 ohm designated input with 75 ohms ???

It would be good for once if Devialet could announce clearly what they see the best solution. However I do not think they know anything at the moment as they are busy running for a bigger market.
From my standpoint I would like to link my units with AES/EBU - AES/EBU (but for some reason the AES/EBU is an input only function).

Time will tell (I hope)/Mike

Perhaps the Devialet needs more than one AES/EBU port? There you go that's the wish list for the next gen Devialet started!

Yes I had similar conversations with Thierry myself. If I recall his explanation was the presence of a transformer on the AES input which makes it qualitatively better.

At least they have always been consistent in saying that the AES/EBU input is the best digital input. That's why I connect my totaldac server via AES/EBU. For the connection from master to slave I chose the best S/PDIF I could find.

I must say I am quite curious to try this Audioprana cable which at 260 euros sounds like a bargain.

Guillaume

(06-Nov-2014, 09:17)mauidan Wrote: I don't understand how using a 75 ohm cable into an input that should be optimized for 110 ohms is optimal for sound quality.

(16-Nov-2014, 23:29)IanG-UK Wrote: My recollection is that in the early days (D-Premier) the master S/PDIF connection to the companion AES/EBU was the only way to operate dual mono. We then got to the stage where Devialet said that that connection was the preferred one; although perversely the default link on the configurator is S/PDIF to S/PDIF.

However, if one's configuration is of the former connection type, and is inevitably defined through the configurator, then this presumably includes the necessary transformer, as has already been documented by Thierry, so surely using the Audioprana and including a further balun (if that is a material thing) is either a waste of €260 or, even worse, a mistake.

Audioprana do not claim that this is a Devialet-compliant cable, merely "Câble de symétrisation sortie source rca/spdif 75 Ohms vers entrée AES/EBU 110 Ohms". Of course, one can try it, and there is a statutory right to return items within 7 days, which Audioprana perversely extends to 8 days but, in the context of something like this, that is a short period.

As this type of component connection is unique in the industry (as far as I know), I cannot imagine that Devialet got it wrong, and if they initially did it has surely been solved through the great flexibility and processing power which their components possess.

(17-Nov-2014, 12:19)Mikeeo Wrote: Devialets solution for linking 2 units with a RCA-XLR set-up only follows the classic way of connecting RCA-XLR ie hot (signal) to pin 2 and then ground (return) to pin 1 and 3, but signal would still be 75 ohm in the case of connection 2 units as output from Dev is 75 ohm (or is it not for some reason?).

As I see it Dev may have their own solution in form of a transformer with auto detect conversion if it is 75 or 110 ohm or it is simply a compromise but due to the transformer it is still a more stable handling of the signal hence the better solution!?

If Dev has a proprietary solution I guess it is at least pat pending and the time gone since such invention would have been filed it should be in the public domain. So until Dev show anything on this matter I am sceptic.

Cheers/Mike

(17-Nov-2014, 17:37)IanG-UK Wrote:

(17-Nov-2014, 12:19)Mikeeo Wrote: Devialets solution for linking 2 units with a RCA-XLR set-up only follows the classic way of connecting RCA-XLR ie hot (signal) to pin 2 and then ground (return) to pin 1 and 3, but signal would still be 75 ohm in the case of connection 2 units as output from Dev is 75 ohm (or is it not for some reason?).

As I see it Dev may have their own solution in form of a transformer with auto detect conversion if it is 75 or 110 ohm or it is simply a compromise but due to the transformer it is still a more stable handling of the signal hence the better solution!?

If Dev has a proprietary solution I guess it is at least pat pending and the time gone since such invention would have been filed it should be in the public domain. So until Dev show anything on this matter I am sceptic.

Cheers/Mike

Hi Mike

As you say, the physical link RCA to XLR still keeps the output specification in "RCA mode" at the companion XLR socket: hot to pin 2 and ground to pins 1 and 3; then behind the XLR socket is presumably a balun activated by the configuration definition when in dual-mono RCA/XLR mode. I can't imagine this is patentable as its simply a combination of a balun and a switch, so there is nothing new.

So I'm quite happy to accept Devialet's word that the cable can be one of standard formation without any intermediate "box" - and I'm really not willing to speculate €260 when Devialet suggest that that is unnecessary. If I did doubt them I think I would try the Neutrik device first (for £30) since if there is a pertinent effect that will primarily, IMO, come through the insertion of the balun rather than through the linking wires. I'm a cable sceptic particularly in relation to digital transmission, at least until one of the magazines get some statistically significant blind listening results to validate any differences - but we have been waiting years for that to happen ... and I suspect we will continue to wait years in the future ...

(18-Nov-2014, 00:42)Mikeeo Wrote: Hi Ian,

Makes sense what you say about the balun on the AES/EBU input as long as the settings acknowledge that it is a link between Devs (75 ohm) or a non-link digital source in (110 ohm).
But for those like me that would like to use a 75 ohm dig out and the 110 ohm in from a digital source I guess the system will not know anything else than a 110 ohm as I haven't seen any setting for the input AES/EBU that would give me the choice of 75 or 110 ohms, or does this exist?
If not, at least I would need my connection from my Unitiserve to change impedance before entering the Devs 110 ohm input.
Well it is not hard to make your own balun as it takes one Lundahl 1574 transformer between a RCA/BNC and XLR, voila! maybe I should test that first and see. I can always cut out the transformer and compare or the opposite

However buying all the diy stuff will be halfway the cost of Audioprana, so...

Cheers/Mike

(18-Nov-2014, 10:27)IanG-UK Wrote:

(18-Nov-2014, 00:42)Mikeeo Wrote: Hi Ian,

Makes sense what you say about the balun on the AES/EBU input as long as the settings acknowledge that it is a link between Devs (75 ohm) or a non-link digital source in (110 ohm).
But for those like me that would like to use a 75 ohm dig out and the 110 ohm in from a digital source I guess the system will not know anything else than a 110 ohm as I haven't seen any setting for the input AES/EBU that would give me the choice of 75 or 110 ohms, or does this exist?
If not, at least I would need my connection from my Unitiserve to change impedance before entering the Devs 110 ohm input.
Well it is not hard to make your own balun as it takes one Lundahl 1574 transformer between a RCA/BNC and XLR, voila! maybe I should test that first and see. I can always cut out the transformer and compare or the opposite

However buying all the diy stuff will be halfway the cost of Audioprana, so...

Cheers/Mike

Hi Mike

Yes, my reference is purely to the configuration setting which, through the link, can use RCA out 75 ohms to XLR in 110 ohms and my conjecture, seemingly confirmed by Devialet, that they include a compensatory balun put in circuit when the configurator is activated in this way.

To the best of my knowledge, there is no way to get a digital XLR output from the Devialet, and I have not tried tricking the unit into providing a digital RCA output by configuring for dual mono but then not linking - I suspect it would not work. If one wants to use the XLR digital input on the master, then I guess that is 110 ohms as the industry standard to match the output from your source which should be to the same standard.

But as far as I can see the Unitiserve digital output is S/PDIF 75 ohms, BNC or Toslink, so you would not want the XLR digital input anyway? Likewise my CD Xt signature is optical or coaxial (no XLR). My older transport, an Audio Synthesis Transcend, now in storage, has all five serial digital interfaces fitted - 75 ohm BNC and also RCA, 110 ohm XLR, glass optical AT/AT&T (eliminating impedance matching requirements) and Toslink; plus some proprietary codings for which one needs the partnering Audio Synthesis DAC. At some stage I might dig this out and be able to try immediate comparisons ...

Regards,

Ian

(18-Nov-2014, 10:54)Mikeeo Wrote: Ian,

For the linking duties I am in no doubt then that Dev, as the claim, has a balun taking care of this due to config, However Thierry also recommended me to use the AES/EBU input on the master hence it is a better input. From this connection I can't see how the system could recognize the Unitiserve 75 ohm output without an auto detect function!? So I guess using the AES/EBU is a compromise that is still better than the S/PDIF input. Here I definitely see a possibility for the 75 to 110 conversion.

Cheers/Mike