My DDDAC story
Hi,
I am since around one year in posession of the Doede Dac. I thought it was time to write a review so here it goes:
My setup consists of:
- 4 dac boards
- modified on board shunt regulators (Tent labs), 16 in total
- allen bradley IV resistors, will be changed for Audio Note Tantalum next week
- salas shunt regulators for the dac and wave IO
- balanced connection to input transformer of my tube amps (audio consulting silver volume transformers)
- alix computer board with mpd pup software and linear power supply (tried raspberry pi first but it is not powerfull enough)
- streaming from a NAS
- all controlled with Mpad ipad interface (Super Cool and easy!)
As former designer and current owner the high end CD player of Bow Technologies, (modified with pcm 1794 dac with tubed output), I have been looking for a nice streaming setup without the illusion of bettering my beloved CD player.
Listening to several commercial streaming devices, including very expensive ones, they always sounded less detailed, less organic, less analogue, less real.
However my computer music database was expanding and I opted for the dddac design as a second music source.
After the basic build was ready I was surprised by the sound quality, it came near my CD player and was clearly better than all the commercial stuff I knew.
This was promising.........
Changing the power supplies, building a nice USB cable and connecting balanced to my amps without a capacitor made it clearly better than my CD player and made me very happy!
(This has lead to a lot of work, making bit perfect rips of my complete CD library, but now I can be a couch potato when listening to music with my ipad................
)
Now, after almost one year my CD player is covered with dust and ready for retirement.
Some notes on high resolution audio streaming:
High resolution audio failed to give me the satisfaction it should be able to. This has several causes.
1) there are 24/192 issues that are really worse than good a good red book remaster. Some Eagles re-issues are famous for clipping for instance
2) there are also upsampled re-issues, you will understand that upsampling does not really provide higher resolution
3) on the other side: a lot of good 24/192 issues are really bad musical performances I call this elevator music
4) 99% of the best music is not available in high resolution, so we will have to do with red book
I got my hands on some good high resolution issues from Patricia Barber and Miles Davis that are really better sounding than my own MoFi gold disk issues, so it is possible!
So, how do we persuade the music industry to improve the quality?
For the time being, I am very happy playing all my music, being red book or better through my DDDAC.
I want to thank Doede for a very nice design and for his support. Cheers Doede
I have build a second dac for a friend and probably more will follow. I also expect the sound quality will develop even more with Lucian designing a new wave IO and daughter board and mpd pup software being updated for higher performance.
This design shows that open cource and non-commercial design by enthousiasts betters what the industry can do
Regards,
Hi,
I am since around one year in posession of the Doede Dac. I thought it was time to write a review so here it goes:
My setup consists of:
- 4 dac boards
- modified on board shunt regulators (Tent labs), 16 in total
- allen bradley IV resistors, will be changed for Audio Note Tantalum next week
- salas shunt regulators for the dac and wave IO
- balanced connection to input transformer of my tube amps (audio consulting silver volume transformers)
- alix computer board with mpd pup software and linear power supply (tried raspberry pi first but it is not powerfull enough)
- streaming from a NAS
- all controlled with Mpad ipad interface (Super Cool and easy!)
As former designer and current owner the high end CD player of Bow Technologies, (modified with pcm 1794 dac with tubed output), I have been looking for a nice streaming setup without the illusion of bettering my beloved CD player.
Listening to several commercial streaming devices, including very expensive ones, they always sounded less detailed, less organic, less analogue, less real.
However my computer music database was expanding and I opted for the dddac design as a second music source.
After the basic build was ready I was surprised by the sound quality, it came near my CD player and was clearly better than all the commercial stuff I knew.
This was promising.........
Changing the power supplies, building a nice USB cable and connecting balanced to my amps without a capacitor made it clearly better than my CD player and made me very happy!
(This has lead to a lot of work, making bit perfect rips of my complete CD library, but now I can be a couch potato when listening to music with my ipad................
)Now, after almost one year my CD player is covered with dust and ready for retirement.
Some notes on high resolution audio streaming:
High resolution audio failed to give me the satisfaction it should be able to. This has several causes.
1) there are 24/192 issues that are really worse than good a good red book remaster. Some Eagles re-issues are famous for clipping for instance
2) there are also upsampled re-issues, you will understand that upsampling does not really provide higher resolution
3) on the other side: a lot of good 24/192 issues are really bad musical performances
I call this elevator music4) 99% of the best music is not available in high resolution, so we will have to do with red book
I got my hands on some good high resolution issues from Patricia Barber and Miles Davis that are really better sounding than my own MoFi gold disk issues, so it is possible!
So, how do we persuade the music industry to improve the quality?
For the time being, I am very happy playing all my music, being red book or better through my DDDAC.
I want to thank Doede for a very nice design and for his support. Cheers Doede
I have build a second dac for a friend and probably more will follow. I also expect the sound quality will develop even more with Lucian designing a new wave IO and daughter board and mpd pup software being updated for higher performance.
This design shows that open cource and non-commercial design by enthousiasts betters what the industry can do
Regards,
Last edited:
Hello Supersurfer
Fine Review...i have ordered one..so this is good reading
How much do the shunt powersupply give in sound quality,,,,where on the boards give it most ?--(analog on the pcm1794 ,,or the digital on pcm1794)
can see you change to pcm1794 on the Bow cd,,,,have you change from pcm1702 ??
best Bjarne
Fine Review...i have ordered one..so this is good reading
How much do the shunt powersupply give in sound quality,,,,where on the boards give it most ?--(analog on the pcm1794 ,,or the digital on pcm1794)
can see you change to pcm1794 on the Bow cd,,,,have you change from pcm1702 ??
best Bjarne
Hi Bjarne,
I made a mistype, the Bow was mofified with 1704 dac chip and is running non oversampling.
As the 1704 is regarded by most to be one of the best multibit dacs ever made it is quite an archivement that the doede dac sounds better. This however may not directly be caused by the dac because there are more differences like the source being a cd loader instead of a computer.
I did not experiment with trying on digital or analog for the shunts, I have extended experience with other amp and dac designs that a shunt regulator (or no regulator) is always better than a series regulator.
But if you want to try I asume its best doing analog stages first; just my thoughts.
Regards,
I made a mistype, the Bow was mofified with 1704 dac chip and is running non oversampling.
As the 1704 is regarded by most to be one of the best multibit dacs ever made it is quite an archivement that the doede dac sounds better. This however may not directly be caused by the dac because there are more differences like the source being a cd loader instead of a computer.
I did not experiment with trying on digital or analog for the shunts, I have extended experience with other amp and dac designs that a shunt regulator (or no regulator) is always better than a series regulator.
But if you want to try I asume its best doing analog stages first; just my thoughts.
Regards,
Last edited:
Hello
Yes the pcm1704 sound good ..i have tried it several times..
Now i use a Rakk dac mk3---it also sound good ,,but it is a little sharp in some array--
maybe i will try the shunt on the analog part..i will think it is a bit difficult to do that ,when i have make it all together---so i must do it before
i will use my AMB.org supply for the 12v
Best Bjarne
Yes the pcm1704 sound good ..i have tried it several times..
Now i use a Rakk dac mk3---it also sound good ,,but it is a little sharp in some array--
maybe i will try the shunt on the analog part..i will think it is a bit difficult to do that ,when i have make it all together---so i must do it before
i will use my AMB.org supply for the 12v
Best Bjarne
No real arrangement as you are the first to ask
I am pretty sure you can do this directly. just send me pm and we will get it arranged...
Doede, thank you for the reply. I'll PM you later this week.
I need to spend a little time more time debugging before pointing fingers. I think the problem is on the hardware side, rather than software. I did play a track at each sample rate and verified via the OS (on Linux: "cat /proc/asound/L20/stream0") that the track was being played back at the sample rate I expected. As I said before, play a 44k1 track and the 44k1 LED flashes at the start of the track and then goes dark after a fraction of a second. Likewise, for each sample rate, the relevant sample rate LED flashes and then switches off. If I suspect anything, it's hardware, relating to the 8mA current source(s) driving the LED's on the WaveIO board side. Nothing to do with player software, I think. Anyway, it's no big deal. (If the board is returned to Lucien for the firmware update maybe he can take a quick look.) If the I2S output stopped working, I'd have an issue. Sample rate display, no issue.
Hi friends,
I read on the Luckit.biz page tht the WaveIO supports 32bits 384kHz when the revision ID is >= 0x0331.
But I do not a "revision ID" number anyware on the card. The XMOS chip is marked: "GT1244L1" and "PKH293.09".
I have received the WaveIO card 3 weeks ago (not yet installed), so it should be a recent one, but I would like to know if it is 384kHz capable ?
In addition, is there an indication that it is decoding a 384kHz stream, because the LEDs indicate 44.1 up to 192kHz, but I do not see anything related to 384kHz ...
Thanks
I read on the Luckit.biz page tht the WaveIO supports 32bits 384kHz when the revision ID is >= 0x0331.
But I do not a "revision ID" number anyware on the card. The XMOS chip is marked: "GT1244L1" and "PKH293.09".
I have received the WaveIO card 3 weeks ago (not yet installed), so it should be a recent one, but I would like to know if it is 384kHz capable ?
In addition, is there an indication that it is decoding a 384kHz stream, because the LEDs indicate 44.1 up to 192kHz, but I do not see anything related to 384kHz ...
Thanks
Last edited:
I hope Doede will not mind to answer to few questions here!
@clivem please do PMs with Doede and see if the problem is from software or hardware. As for me, I can only guarantee that if the issues is from hardware then I'll solve it as best as I can. Just speak with Doede first since he is the one who will decide what you have to do next regarding your WaveIO.
@Nounours18200: Check my first post here for details on how to find the revisionID of your card!
Cheers,
Lucian
@clivem please do PMs with Doede and see if the problem is from software or hardware. As for me, I can only guarantee that if the issues is from hardware then I'll solve it as best as I can. Just speak with Doede first since he is the one who will decide what you have to do next regarding your WaveIO.
@Nounours18200: Check my first post here for details on how to find the revisionID of your card!
Cheers,
Lucian
Update on my noise problem:
After installing the 4.7nF caps the quantisation noise has reduced but not gone completely, and there is still a hissing sound coming from the tweeters when no music is playing.
This hissing is a problem on quiet music late at night, and quite painful if the ear lines up directly with the tweeter but during the day with background noise it is not really noticeable. Unfortunately I am currently using the hifi to do some noise reduction work for a client, and part of this involves removing hiss from recordings. More hiss added by the DDDAC is not a good thing here so I am keen to begin to understand the problem better in order to find a solution.
I have yet to reduce the output level of the DAC, i.e. turn down the hiss, because I would like to think it possible to remove the noise at the source.
Also still a little confused about the output stage. Supersurfer above mentions the use of zero output caps because the amps have input transformers. I am currently using 4x basic 2.2 uf mundorfs supplied with the kit. I have not tried to connect the DAC directly to amps without any caps, but the Hypex modules do allow for DC offset adjust. Maybe for me the best way is to adjust the offset and connect the DAC directly. This will of course mean that I cannot use other sources in future without recalibrating the offset. Thoughts?
P.S. This is still the best DAC I have heard.
After installing the 4.7nF caps the quantisation noise has reduced but not gone completely, and there is still a hissing sound coming from the tweeters when no music is playing.
This hissing is a problem on quiet music late at night, and quite painful if the ear lines up directly with the tweeter but during the day with background noise it is not really noticeable. Unfortunately I am currently using the hifi to do some noise reduction work for a client, and part of this involves removing hiss from recordings. More hiss added by the DDDAC is not a good thing here so I am keen to begin to understand the problem better in order to find a solution.
I have yet to reduce the output level of the DAC, i.e. turn down the hiss, because I would like to think it possible to remove the noise at the source.
Also still a little confused about the output stage. Supersurfer above mentions the use of zero output caps because the amps have input transformers. I am currently using 4x basic 2.2 uf mundorfs supplied with the kit. I have not tried to connect the DAC directly to amps without any caps, but the Hypex modules do allow for DC offset adjust. Maybe for me the best way is to adjust the offset and connect the DAC directly. This will of course mean that I cannot use other sources in future without recalibrating the offset. Thoughts?
P.S. This is still the best DAC I have heard.
Last edited:
Hi Thommy,
think I have the same setup as you have, a 1 board DAC and Hypex NC400 modules.
This is what I did against the hissing problem:
1. I lowered the gain of the NC400 modules by removing the R141 resistor (see: http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-104.html#post2837086)
2. Installed the 4.7nF caps on the DAC
3. As IV resistors I used two 68 ohm in series and tapped the signal from the center.
With this modifications I can hear the hissing only with my ear directly at the tweeter.
Because I use a balanced connection and digital volume control no caps or transformers are needed between the DAC and the NC400. Sound is excellent with this direct connection!
Regards,
Jürgen
think I have the same setup as you have, a 1 board DAC and Hypex NC400 modules.
This is what I did against the hissing problem:
1. I lowered the gain of the NC400 modules by removing the R141 resistor (see: http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-104.html#post2837086)
2. Installed the 4.7nF caps on the DAC
3. As IV resistors I used two 68 ohm in series and tapped the signal from the center.
With this modifications I can hear the hissing only with my ear directly at the tweeter.
Because I use a balanced connection and digital volume control no caps or transformers are needed between the DAC and the NC400. Sound is excellent with this direct connection!
Regards,
Jürgen
Hi guys,
Thankyou both very much for your comments, very encouraging. I guess from this that this hissing is to do with connecting a source to such a powerful amp with no attenuation, rather than an inherent fault with my build somewhere.
I have already done the 4.7nF cap modification, but have yet to split the resistors to lower the output gain so maybe this will help. At the moment gain is reduced by 12dB on the WaveIO, this is not the same as the software volume control in my streamer software so I still have that too. It would however make much more sense to lower the gain just after the DAC rather than just before it.
Jurgen, you must have very sensitive speakers. Removing the R141 drops input gain by 14dB, and splitting the IV resistor in the manner you suggest will also affect the gain (reducing it by 50%?). This would be too quiet for me, with speakers of 93dB sensitivity.
Are you both saying that I can connect the 3 pins on each channel of the DAC directly to the differential input on my amplifiers, and that this type of input would remove the DC offset and therefore the need for any capacitors in the signal path?
I have not tried this yet because every DDDAC build I have seen has output transformers or capacitors, so I made the assumption that it simply must be a necessary part of the design.
Thankyou both very much for your comments, very encouraging. I guess from this that this hissing is to do with connecting a source to such a powerful amp with no attenuation, rather than an inherent fault with my build somewhere.
I have already done the 4.7nF cap modification, but have yet to split the resistors to lower the output gain so maybe this will help. At the moment gain is reduced by 12dB on the WaveIO, this is not the same as the software volume control in my streamer software so I still have that too. It would however make much more sense to lower the gain just after the DAC rather than just before it.
Jurgen, you must have very sensitive speakers. Removing the R141 drops input gain by 14dB, and splitting the IV resistor in the manner you suggest will also affect the gain (reducing it by 50%?). This would be too quiet for me, with speakers of 93dB sensitivity.
Are you both saying that I can connect the 3 pins on each channel of the DAC directly to the differential input on my amplifiers, and that this type of input would remove the DC offset and therefore the need for any capacitors in the signal path?
I have not tried this yet because every DDDAC build I have seen has output transformers or capacitors, so I made the assumption that it simply must be a necessary part of the design.
Last edited:
Hi Thommy,
I use an input transformer in my amps, that is the same as an output transformer; this blocks any DC.
But DC will saturate the transformers I use so a good balancing helps.
You can try for yourself by measuring the DC on the positive and negative output pin. Than connect your amp without speaker and measure if there is any DC on the output of your amp.
You should keep the resistor value as close to 33 for each board as possible. Otherwise the sound quality will suffer.
Regards,
I use an input transformer in my amps, that is the same as an output transformer; this blocks any DC.
But DC will saturate the transformers I use so a good balancing helps.
You can try for yourself by measuring the DC on the positive and negative output pin. Than connect your amp without speaker and measure if there is any DC on the output of your amp.
You should keep the resistor value as close to 33 for each board as possible. Otherwise the sound quality will suffer.
Regards,
Having measured the DC between the positive and ground, I found 2.7v - the same between negative and ground. There is no measurable DC between positive and negative. Is that what you mean?
I will try your suggestion of connecting the XLR pins directly to the DDDAC output pins and then measure the voltage at the speaker outputs. Thankyou.
I will try your suggestion of connecting the XLR pins directly to the DDDAC output pins and then measure the voltage at the speaker outputs. Thankyou.
Last edited:
Yes Thommy no transformers or caps. If you use the positive lead of your multimeter on the outputs from the dac and the common to ground, you should notice the dc changes polarity, i.e. 2.7v positive to ground and -2.7 negative to ground. This cancels out the dc on the output, the offset you need to look at is the difference between the positive and negative voltages, if between the pos and neg outs, the voltage is 0.02v, that 20mv could affect the amp. Check the output to the speakers and see the offset there.
Make sense?
Drew.
Make sense?
Drew.
Last edited: