DAC progress

[xcortes]

You will have fun!

That DAC thing is getting serious. I noticed you're going up to 96k in resolution in your demos. Are you being restricted by the 0404 USB doing only 24/96 on its S/PDIF out?

[Doc B.]

To get started we have been using TOSLINK for the SPDIF in and out. The TOSLINK connection is limited to 96K. The DAC itself can handle 192K. We've all been quite busy the past few months and when the time allows a coax SPDIF interface will be added.

[Koda]

Have you tried a glass toslink "cable" yet? You might hear some differiences, as I found to be the case.
In my system there was a slight improvement over the typical plastic type, more dynamics and finer resolution, yada yada yada...
You can find them on the 'bay at good prices if you search for "glass toslink".

(Sorry, I didn't intend to send this off on another tangent)

[Doc B.]

Paul Stubblebine is going to send me some glass TOSLINK cables to try. It's one of those things that gets forgotten with the zillion other things we have going an all the time...

[arveedub]

Are you developing the DAC as a kit? If you are I hope you keep the TOSLINK.

[Doc B.]

Yes the goal is two versions, one a tube output 24/192K capable R2R DAC kit with TOSLINK and Coax SPDIF and possibly AES/EBU. The other product that may get developed is a music server project that would work with the DAC. This might take the form of a networked headless unit that allows one to use whatever OS and playback software they prefer on a slaved workstation, feeding it to the headless server over the network via Netjack. That would probably not be offered as a kit, as we wouldn't feel comfortable taking on the burden of OS and software installation tech support.

[Doc B.]

We spent yesterday afternoon working on the system setup in preparation for the Bottlehead holiday party this Sunday. We continue to knock down the noise floor, refine the speaker placement and listening position, and we did some minor level adjustment of the crossover as well. When all of this was finished we listened to some 24/96K files through the new DAC. It's getting pretty darned good, I wasn't that "shocked" going back and forth from digital to tape (sometimes the change from tape to digital makes me a little disoriented or ill). Our project today is to cook up a 5V linear supply for the EMU 0404 USB, which is running as our Async USB-SPDIF converter, reclocked by the DAC. Don't know if it will make much difference. But there is only one way to find out...

[Doc B.]

It turns out the linear supply on the 0404 USB makes a positive difference - even though the 0404 is functioning only as a USB to SPDIF converter. Seems to flesh things out a bit in the midrange and clean up the top end a bit. The 0404 seems to draw a bit less than .5A. We ended up using the 5V secondary of a surplus Paramount power transformer. This runs into a Schottky bridge, then a 1000uF cap, then an FC-1 filament choke and then a 12000uF cap. That puts out about 6.5VDC at load, which we are regulating with an LM1085IT-ADJ to 5.03V.

I would definitely recommend trying a linear supply if you are using an EMU 0404 USB. It could possibly make even more difference than we heard if you are using the DAC and the audio section in the 0404 rather than an external DAC like we are using.

[BNAL]

I have found that power supplies can make the biggest difference to sound, even with ok circuits and where you think that no change will happen. I have even built power filters for my equipment, because I feel that the cleaner the power the audio circuit gets the better the sound, in most cases.

[Jim R.]

All sounds good to me.  Can I ask how high a sampling rate you can go with the 0404?  There are all kinds of opinions and statements that say that it can only do 24/96 and others who claim it works fine at the 24/192.  I'm not sure who to believe, except somebody who is actually using it in the way you are -- bidirectional spdif.

There may also be some other fairly easy and inexpensive little tweaks to get a bit cleaner signal out of the 0404 too, and hopefully I'll be working on that some in the new year.

-- Jim

[Doc B.]

I can't confirm or deny 192k operation yet, because we are still limited to 96k due to the TOSLINK. IIRC you can only do up to 96k with a Mac and you can do 192k with a PC - but only with the Creative driver, I think? I'm not married to the 0404USB for this particular project. It was just a very simple way to get started. If I can get my act together with the headless Linux box I bought I may eventually try a different headless box that can handle a PCI-E sound card. At this point I am hung on the headless setup until I find a Linux savvy person to help me sort out why the instructions I have for setup aren't working.

[Jim R.]

I have one of those headless linux boxes in the project queue too.  But mine doesn't have any provision for any add-in cards except wireless cards.  No spdif either, just USB 2.0 spec ports and 3 ethernet ports.  Better suited to working with a usb dac, which was what I intended it for.  Hopefully that will feed the HRT MusicStreamer+ into the sex amp for the bedroom system.

Any rough ideas what this headless linux box may cost?


-- Jim

[Doc B.]

That sounds like the same Alix headless box I have. My original plan was to do the MPD server that nycparamedic describes on Head Fi. This sets up a separate Linux workstation as an NFS server and the MPD player is actually on the headless box. So the USB and LAN ports are just the right setup with a USB DAC. John Swenson has suggested I look into running netjack on the headless box instead and use a Jack friendly player on a Mac, PC or Linux workstation, so that the headless server isn't setting the limit on what player software one might want to use.

Right now I am just trying to get the headless box set up to talk to a NFS, and I'm thinking that I might try to set my PC workstation up for NFS to serve the files and go ahead and try MPD. I seem to have run into a problem with the Linux code on the headless box, and I'm busy enough this time of year that I may have to call in a Linux guru to help me out.

[paba]

Hello,

we are mid way in Q1, is it rude to ask for an update (any direction changes) and a best guess on time to market...Q2, Q3, Q4?
I'm holding back on my slow entry into HDD based music to see how this one plays out.

thanks
paba

[Doc B.]

I think we have worked out what we want to do for a tube stage, and now we are concentrating the best way to get from the PC to the DAC. John has come up with a neat USB interface suggestion that would involve another manufacturer and thus may take a little time to pursue. Between the fact that this is a new arena for us and that rapid changes are happening in high end server development right now, I feel that we need to take our time and do it right the first time.  So it's a little hard to say when it will be ready. I know that's not what folks want to hear, but I think that is better than rushing a product to market that might not be as good as we could ultimately make it.