I see. I hope yours doesn't regard loudspeakers as climbing material, like our Donny does; it's the reason why I sold my ESL63s.
Bubble (and her sister, Squeak, who we lost to a car - BTW Bubble & Squeak is a traditional British breakfast dish) used to be terrible - they once completely shredded the foam surround of one of my Lowther units. She's now in sedate middle-age; the mouths of my horn speakers, which point into the corners of the room, are one of her favourite hiding/sleeping places but she still has to come and check out whatever I'm doing so I have to be careful about exposed wires and the like and is why I generally don't do breadboard builds.
To me the first question is: does anyone mind having to hand solder at most three 0805 SMD resistors (2 mm by 1.2 mm)? It doesn't matter when you accidentally short them with a solder bridge, as they are 0 ohm anyway. In fact using blobs of solder instead of official 0 ohm resistors will also do the trick.
I don't know if SMD pin headers would fit, there is not much space (otherwise I would have used 1206 0 ohm resistors). If I would be able to squeeze in SMD pin headers and hold on to the idea of keeping digital data processing parts on the bottom of the board, two of them will stick out from the bottom of the board.
Ray originally wanted U.FL connectors and later asked for an additional Amanero-style input. I remember there was not much space left to fit it in. I wanted to include a way to disconnect the Amanero-style header as it spoils the shielding between data and clocks that the U.FL connectors provide. Besides there was a need to be able to choose between bit clock and master clock. Hence the 0 ohm resistors.
I don't know if SMD pin headers would fit, there is not much space (otherwise I would have used 1206 0 ohm resistors). If I would be able to squeeze in SMD pin headers and hold on to the idea of keeping digital data processing parts on the bottom of the board, two of them will stick out from the bottom of the board.
Ray originally wanted U.FL connectors and later asked for an additional Amanero-style input. I remember there was not much space left to fit it in. I wanted to include a way to disconnect the Amanero-style header as it spoils the shielding between data and clocks that the U.FL connectors provide. Besides there was a need to be able to choose between bit clock and master clock. Hence the 0 ohm resistors.
I have seen Post 486 is referencing a Projected Build Cost of £500 as a Minimum if the Chassis is a Basic Model.
These costs for such a unique Electronics Design are stimulating, especially now I am becoming serious about creating a Digital Source to be used in a
Second System for the Main Room in the home and leave my CDP > Valve Output DAC in the HiFi Rooms System.
As a side, the Second System will be a Slagle AVC > Low Watt Valve Amp( yet to be decided), High Efficiency OB Speakers.
This Valve DAC design as part of the source, I am feeling will work very nicely.
These costs for such a unique Electronics Design are stimulating, especially now I am becoming serious about creating a Digital Source to be used in a
Second System for the Main Room in the home and leave my CDP > Valve Output DAC in the HiFi Rooms System.
As a side, the Second System will be a Slagle AVC > Low Watt Valve Amp( yet to be decided), High Efficiency OB Speakers.
This Valve DAC design as part of the source, I am feeling will work very nicely.
Maybe I'm stating the obvious now, but Ray's estimate and group buy are for the raw DSD version (DSD-only version). If you want to play CDs over it, you will need something to convert the PCM data from the CD player into DSD, like the AK4137 board Steve and Ray were discussing. You could also build the original valve DAC, but that is considerably more expensive and it is not part of Ray's group buy.
Last edited:
This morning I've spent a couple of hours listening to the Valve DAC through my new 13E1 SE-OTL amplifier, which is now complete apart from a couple of cosmetic parts (like a front panel).
I had two objectives, the easy one was to just run the 13E1 for a prolonged period to start burn it in and check for any problems. The amp behaved impeccably, no misdemeanours or hums/buzzes. Sounds very good too.
The second objective was to continue tweaking the Valve DAC; having spent quite a bit of time last weekend listening to 'silence' in order to minimise noise by balancing the valves. I started by playing the solo flute pieces that I've become quite familiar with and encountered an interesting situation as I was hearing noise that correlated to music playing - this was mentioned as a possible scenario previously;
https://www.diyaudio.com/forums/dig...-linear-audio-volume-13-a-66.html#post6311094
Anyway, I've now got the 'AMSDM7 512+fs' modulator selected in in HQP (at DSD256) and that addresses the noise and sounds good with the filter options I've currently settled on. To all intents and purposes (i.e. measuring with my ears sitting in the listening seat) the only noise the DAC/Amp produces is the music.
Although the modulator seems to be a fix for the noise I'm wondering if the situation is telling me there's more work to do with tweaking the timing potentiometers (RV3/4) or should I consider switching from using the masterclock to the bit clock - any thoughts Marcel?
I had two objectives, the easy one was to just run the 13E1 for a prolonged period to start burn it in and check for any problems. The amp behaved impeccably, no misdemeanours or hums/buzzes. Sounds very good too.
The second objective was to continue tweaking the Valve DAC; having spent quite a bit of time last weekend listening to 'silence' in order to minimise noise by balancing the valves. I started by playing the solo flute pieces that I've become quite familiar with and encountered an interesting situation as I was hearing noise that correlated to music playing - this was mentioned as a possible scenario previously;
https://www.diyaudio.com/forums/dig...-linear-audio-volume-13-a-66.html#post6311094
Anyway, I've now got the 'AMSDM7 512+fs' modulator selected in in HQP (at DSD256) and that addresses the noise and sounds good with the filter options I've currently settled on. To all intents and purposes (i.e. measuring with my ears sitting in the listening seat) the only noise the DAC/Amp produces is the music.
Although the modulator seems to be a fix for the noise I'm wondering if the situation is telling me there's more work to do with tweaking the timing potentiometers (RV3/4) or should I consider switching from using the masterclock to the bit clock - any thoughts Marcel?
Yes, I agree Marcel, the post referenced actually says;
"ignoring the upstream infrastructure to deliver DSD data to the DAC input... ...without valves... ...but with perhaps a very cheap, basic chassis I think the project will cost at least £500 and likely a bit more... ...BTW, I haven't used boutique caps/resistors etc."
Just to illustrate the point further, If you choose the 'no compromise' source route of using HQ Player, a licence will cost you around £230 and you'll need a powerful computer to run it on too - an old laptop just won't cut it.
John, as you know, we're both members at Audio-Talk so I'm aware of some of your system developments and I'm really not sure this is a project for you. PM me if you want to discuss.
Trying to minimize the effect by tweaking RV3 and RV4 is definitely worth trying.
Changing to the bit clock could also help for DSD256 and below, but there is a potential problem. The resynchronizer may or may not cause a hold time issue when you use the bit clock. Hold time violations could result in the loud noise you had at the beginning reappearing and disappearing at seemingly random moments.
Does anyone know whether the raw DSD data coming out of the Beaglebone Black change on the rising or the falling bit clock edge? If not, does anyone have a Beaglebone Black and an oscilloscope to measure it?
Changing to the bit clock could also help for DSD256 and below, but there is a potential problem. The resynchronizer may or may not cause a hold time issue when you use the bit clock. Hold time violations could result in the loud noise you had at the beginning reappearing and disappearing at seemingly random moments.
Does anyone know whether the raw DSD data coming out of the Beaglebone Black change on the rising or the falling bit clock edge? If not, does anyone have a Beaglebone Black and an oscilloscope to measure it?
I have both BBB and an oscilloscope as well as TPA Hermès/Cronus stack... if you give me details on what how exactly to measure it - I’d be happy to.
Thank You Marcel
I did not know the CD Replay was not an option and the AK4137 Board would be a solution to produce the CD Replay.
I am not a builder of any electronic devices, much of the tech talk is not comprehended by myself, and I would not be able to contribute to these discussions.
I am intending on discussing this build with a EE who has done builds for me throughout many years, at present the EE is not available, I have to wait a little longer to see if they are able to support me with the build of this one.
What ever the outcome, the Design has my attention, and picking up on the feedback of the builds will prove interesting to follow.
I did not know the CD Replay was not an option and the AK4137 Board would be a solution to produce the CD Replay.
I am not a builder of any electronic devices, much of the tech talk is not comprehended by myself, and I would not be able to contribute to these discussions.
I am intending on discussing this build with a EE who has done builds for me throughout many years, at present the EE is not available, I have to wait a little longer to see if they are able to support me with the build of this one.
What ever the outcome, the Design has my attention, and picking up on the feedback of the builds will prove interesting to follow.