[phusis]

Is the new 192K asynchronous USB input module of the DAC-9 galvanically isolated?

Any respons to this - Jason?

[rbbert]

Is the new 192K asynchronous USB input module of the DAC-9 galvanically isolated?

I also wonder about this, as well as whether it would sound better to use a USB>SPDIF convertor with a BNC-BNC cable rather than the USB in?

[nuforce-casey]

The USB portion is optically isolated, while the SPDIF signals are galvanic isolated.  There is no reason why a longish cable would sound better than a direct PCB socket connection however, the cable could add some coloration and help 'tune' the sound, if so preferred.

[phusis]

Thanks for your reply, Casey.

I wrote this over on your(Nuforce's) Facebook wall:

...
What is the reason for not implementing a digital volume control in the DAC-9 as showcased in the DAC-100? In theory a high-bit digital volume control seems preferable, but I could be wrong. I find the (supposed "in the analogue domain") internal volume control of my DAC-9 to be more natural sounding than going be the internal digital volume control of my software mediaplayer JRiver MC17, but that's probably not a fitting comparison to the potential of an in-DAC high-bit digital volume control over an analogue ditto.

[nuforce-casey]

The analog volume comes from a real preamp circuit in the DAC9 with a bypass = positive gain up to 8V.  DAC-100 digital 'attenuation' can only bring down the voltage and no positive gain, so it can't be used as a preamp.

[shkorc]

I'm trying to setup my DAC-9 on a Linux machine, but haven't been able to find any information as to which chip is used in the 192k USB module, to assist with locating appropriate drivers. Any inpu on this would be really appreciated, thank you.

[shkorc]

Opening the DAC-9 and examining the USB module, the USB transceiver chip says NuForce NFUSB192S 12W12, which does not help much in term of drivers. I received some unconfirmed information that it might use a Tenor 8802. Can someone please confirm?

[nuforce-casey]

Hi,
The 192kHz chip is not certified to work in Linux.   I believe some customers were able to have 96kHz working with ALSA, but nothing further up.

[shkorc]

Thank you for your reply Casey.

Could you provide some more info on 96k and ALSA? Is the brand and model of the chip something Nuforce would prefer not to share with others?

[ohlins]

I've just sent in my DAC 9 to upgrade to the 192 board ....

Just curious why is soldering required ....

[nuforce-casey]

The original optical isolater used to isolate the USB input has a speed-grade that is not fast enough to pass 192kHz signal, hence the need to replace the isolator which is on the main PCB.

[nuforce-casey]

The USB chipset manufacturer has no official Linux support.  Nevertheless, it was reported by some it could be working in Alsa 1.0.25 with Ubuntu.

[shkorc]

I can confirm it is working with Alsa 1.0.25 in Ubuntu 12.04.1, this is good news for others that might be interested in using their DAC-9 on a Linux machine.

The reason why I am inquiring about the chipset is I'm trying to connect the DAC-9 to a SB Touch, which uses another Alsa driver and Linux kernel version. Others have managed to get it working, but their DACs have native Linux support. It is possible to patch the system to make it work, but some information about the chip are required to locate the necessary patches.

I'd really like to use the async USB in pair with this network streamer, so I can get rid of the noisy computer, that's ruining my listening experience, lower power consumption etc.

[Ern Dog]

Since my cdp source does not have a BNC output, but does have a digital coax output, would there be any advantage to using a 75 ohm cable with an RCA connector on one end (from cdp) and a BNC connector on the other end (to the DAC-9)?

[nuforce-casey]

Since my cdp source does not have a BNC output, but does have a digital coax output, would there be any advantage to using a 75 ohm cable with an RCA connector on one end (from cdp) and a BNC connector on the other end (to the DAC-9)?

Based on my experience, using BNC on the DAC-end and RCA output from the CDP end is still beneficial; with the exact same cable but using BNC termination, even 1 BNC termination sounded better.

[rbbert]

Since my cdp source does not have a BNC output, but does have a digital coax output, would there be any advantage to using a 75 ohm cable with an RCA connector on one end (from cdp) and a BNC connector on the other end (to the DAC-9)?

If you're still using an Oppo, you're probably better off with optical than RCA.

[phusis]

Based on my experience, using BNC on the DAC-end and RCA output from the CDP end is still beneficial; with the exact same cable but using BNC termination, even 1 BNC termination sounded better.

Casey --

I'd be very grateful for your (as well as others) assessment here:

Currently I'm using my DAC-9(the "older" one with 96K USB input) via its XLR AES/EBU input fed from an HTPC-based RME HDSP 9632 soundcard and its XLR AES/EBU balanced output. Would you instead recommend one of the following:

- That I upgrade my USB module to the asynchronous 192K ditto and use this USB-connection to my HTPC, or..

- that I instead buy a so-called Word Clock Module(WCM) expansion board with BNC outputs to my RME 9632(for about the same amount of money as the 192K USB module) and connect BNC-BNC? What do you believe will be the sonically superior connection of all, incl. my current one?

(link to the WCM board, at the bottom of the page: http://www.rme-audio.de/en_products_hdsp_expansion_boards.php)

Best,
Mikael

[nuforce-casey]

Hi,
First of all, I believe BNC connection is a better link than AES/EBU, which the real reason is for VERY LOOONG cable drive, but in short distance, offers no meaningful advantage, in fact, could be disadvantage, because the signal level from AES/EBU is huge (8Vrms), and that has repercussions in generating more ripples and more jitters, ceteris paribus, plus the fact that XLR connector itself are not true 110-ohm impedance.

I read the manual of the World Clock module, and if what RME stated is accurate, it should offer a significant clock improvement.  I would think since with your RME investment, and the sophistication of the RME system, my suggestion is to stick with the RME.

In general, the PCI bus offers far better interface accuracy and option, but with a high complexity for both the design side and customer side (opening up PC, can't use with laptop, etc).   I don't think you should give up the RME and start all over with the new USB module given the same cost.  The World Clock module is plug-and-play and has no impact with your other settings.

[phusis]

Casey --

Thank you very much for your thorough reply, also in reading the manual to the WCM to make your assessment - I really appreciate it!

I'll go ahead then and order the WCM expansion board, and have myself a proper BNC cable (either DIY or pre-assembled) with true 75 ohm connectors.

I'm hoping it will make a difference for better connecting BNC-BNC this way, and in any case the expense is rather small so even subtle differences will be appreciated.

If you could describe it in your own words, Casey, how would you come about the sonic difference a BNC-BNC connection typically makes over the other S/PDIF inputs?

Best,
Mikael

[rbbert]

FWIW, NuForce makes a digital cable available with BNC connectors.