Yes. And remember I want n°001.
The other day I completed my second single ended (just one channel) amp, but this time, instead of lateral Mosfet I used a V-Fet
Preliminary listening shows a distinct difference, and I would say advantage, for the V-Fet (2SK82) though a flea-weigth power. They are hard to get but worth it, IMHO.
Best wishes,
M.
....the other day I completed my second single ended (just one channel) amp, but this time, instead of lateral Mosfet I used a V-Fet
Preliminary listening shows a distinct difference, and I would say advantage, for the V-Fet (2SK82) though a flea-weigth power. They are hard to get but worth it, IMHO...
hi max,
if you like the vfets ( i do too ;-) ) then you should try the nelson's recent amp " sony vfet part2" and you will be more than positively surprised how nice a pp-amp can sound
best wishes
m.
hi max,
if you like the vfets ( i do too ;-) ) then you should try the nelson's recent amp " sony vfet part2" and you will be more than positively surprised how nice a pp-amp can sound
best wishes
m.
Sorry for O.T.
Yes, I am reading that thread now but have not found the final circuit and PCB etc.
Also a circlotron seems interesting.
Anyway, Single ended with good power supply and good coupling cap combination has something very special, paired with high efficiency transducers/speakers...
Sorry again.
Very very interesting, thanks John for the info and for the exciting developments!
(No buffer after the discrete dac? Why?)
Thanks,
Alex
(No buffer after the discrete dac? Why?)
Thanks,
Alex
Hi Alexandre,
Buffers add (temporal) distortion, so the output signal quality always gets worse. This is the result of imperfect components, it is not possible to construct perfect circuits with these imperfect components. At best one can fix one problem by creating a new one.
The Mosaic D/A converter core has relatively low output impedance of 750 Ohms and the new ternary logic circuits offer 2Vrms (5.65Vpp) output amplitude without the need for additional active circuits.
The single time domain operation, temporal dithering and the multi-stream digital audio pulse interface fix many issues with digital audio that kept us busy for the past decade.
The usual procedure is that we design a new DAC, test it for a while, then discover some minor issues with sound quality and start the next design. The main problems usually have to do with the digital audio receiver, timing and interference spectrum.
The final Mosaic DAC is different, it sounded remarkably good on the breadboard setup. It continued to sound good when located on the PCBs and it still sounds as good after months of listening. There is music, and after listening to it for 10 minutes or so it only keeps getting better and it becomes difficult to end the listening session.
Like all other DACs it still has some minor source dependency, this is impossible to block as this would require a zero bandwidth interface. So it remains important to use a good USB audio source with a good application. The final Mosaic DAC has built-in bit-perfect test, so it is easy to check if the USB audio source is configured correctly.
The audible differences with bit-perfect playback are caused by the interference riding on the USB data. It translates to an interference spectrum that changes the perceived sound by changing the output spectrum of the DAC. This degrading is based on demodulation of the interference spectrum. The demodulated interference that falls within the audio spectrum or reflects back in the audio spectrum then becomes audible.
Reclocking doesn't help here, it only adds yet another interference spectrum to the mix.
Isolators are not perfect either. Every isolator will add jitter, the best one's add at least 100ps of jitter. Some isolators contain on-chip oscillators and logic circuits that produce a powerful interference spectrum, making these even more unsuitable for digital audio.
So the best thing to do is to leave these isolators out and fix possible ground loop issues another way.
I used iTunes for years, but the final Mosaic DAC showed that iTunes fails to offer good sound quality, even when using the latest version on Sierra. Playback is bit-perfect so this isn’t the cause. When running another application on the same hardware and OS like VOX or Swinsian for example, sound quality is a lot better compared to iTunes.
For mac users, here is a link for the Swinsian application:
Swinsian — The Advanced Mac Music Player
Buffers add (temporal) distortion, so the output signal quality always gets worse. This is the result of imperfect components, it is not possible to construct perfect circuits with these imperfect components. At best one can fix one problem by creating a new one.
The Mosaic D/A converter core has relatively low output impedance of 750 Ohms and the new ternary logic circuits offer 2Vrms (5.65Vpp) output amplitude without the need for additional active circuits.
The single time domain operation, temporal dithering and the multi-stream digital audio pulse interface fix many issues with digital audio that kept us busy for the past decade.
The usual procedure is that we design a new DAC, test it for a while, then discover some minor issues with sound quality and start the next design. The main problems usually have to do with the digital audio receiver, timing and interference spectrum.
The final Mosaic DAC is different, it sounded remarkably good on the breadboard setup. It continued to sound good when located on the PCBs and it still sounds as good after months of listening. There is music, and after listening to it for 10 minutes or so it only keeps getting better and it becomes difficult to end the listening session.
Like all other DACs it still has some minor source dependency, this is impossible to block as this would require a zero bandwidth interface. So it remains important to use a good USB audio source with a good application. The final Mosaic DAC has built-in bit-perfect test, so it is easy to check if the USB audio source is configured correctly.
The audible differences with bit-perfect playback are caused by the interference riding on the USB data. It translates to an interference spectrum that changes the perceived sound by changing the output spectrum of the DAC. This degrading is based on demodulation of the interference spectrum. The demodulated interference that falls within the audio spectrum or reflects back in the audio spectrum then becomes audible.
Reclocking doesn't help here, it only adds yet another interference spectrum to the mix.
Isolators are not perfect either. Every isolator will add jitter, the best one's add at least 100ps of jitter. Some isolators contain on-chip oscillators and logic circuits that produce a powerful interference spectrum, making these even more unsuitable for digital audio.
So the best thing to do is to leave these isolators out and fix possible ground loop issues another way.
I used iTunes for years, but the final Mosaic DAC showed that iTunes fails to offer good sound quality, even when using the latest version on Sierra. Playback is bit-perfect so this isn’t the cause. When running another application on the same hardware and OS like VOX or Swinsian for example, sound quality is a lot better compared to iTunes.
For mac users, here is a link for the Swinsian application:
Swinsian — The Advanced Mac Music Player
Thanks John.
Any clue about first Mosaic production batch arrival?
While you are testing sources, could you test if Mosaic supports Bughead Infinity Blade player (works on latest windows OS) that can add different flavors like on an ice cream shop?
Bug head
(scroll down till "bug head 7.81" download icon)
Thank you very much.
M.
PS: soon pics of the latest mods to the DUAL-MONO DAC.
Any clue about first Mosaic production batch arrival?
While you are testing sources, could you test if Mosaic supports Bughead Infinity Blade player (works on latest windows OS) that can add different flavors like on an ice cream shop?
Bug head
(scroll down till "bug head 7.81" download icon)
Thank you very much.
M.
PS: soon pics of the latest mods to the DUAL-MONO DAC.
John, I can only say: Congratulations and keep up the great work!
(Have you ever tested the tda1543A? With the A suffix. I´d like to know if the outputs are simultaneous on this one. Don´t have any to test. On the tda1543 I have confirmed on the scope there is a delay between L and R output (while experimenting with balanced config). Tda1545 works.)
(Have you ever tested the tda1543A? With the A suffix. I´d like to know if the outputs are simultaneous on this one. Don´t have any to test. On the tda1543 I have confirmed on the scope there is a delay between L and R output (while experimenting with balanced config). Tda1545 works.)
Last edited:
MOSAIC DAC Relay Volume Control Module
Hi John,
when looking at the picture of the Relay Volume Control Module in Post #5747
it seems you changed to a serial volume control with up to 5 Relay contacts / SMD resistors in the signal path. Isn't this a step back compared to shunt volume control you used in the past.
Do thin film low noise precision resistors offer the same audio quality as the Rhopoint or the Welwyn W21 wirewound resistors you used before?
Is the use of thin film low noise precision resistors even the better choice (compact size, no lead wires, low stray inductance and capacitance) for a relay volume control.
Are ni-chrome resistors suited best?
Hi John,
when looking at the picture of the Relay Volume Control Module in Post #5747
it seems you changed to a serial volume control with up to 5 Relay contacts / SMD resistors in the signal path. Isn't this a step back compared to shunt volume control you used in the past.
Do thin film low noise precision resistors offer the same audio quality as the Rhopoint or the Welwyn W21 wirewound resistors you used before?
Is the use of thin film low noise precision resistors even the better choice (compact size, no lead wires, low stray inductance and capacitance) for a relay volume control.
Are ni-chrome resistors suited best?
sd player
the lessloss sd player seems finally to be for sale at 90k euro and they claim to be the world first of its kind , handling up to 32gb sd cards...
.... oh... world first? hmmm.... my 5 years old EC-design sd player works very fine with my old 64gb sd cards ;-)
Laminar Streamer SD player : High end SD player | Solid state memory player | Computer audio player
the lessloss sd player seems finally to be for sale at 90k euro and they claim to be the world first of its kind , handling up to 32gb sd cards...
.... oh... world first? hmmm.... my 5 years old EC-design sd player works very fine with my old 64gb sd cards ;-)
Laminar Streamer SD player : High end SD player | Solid state memory player | Computer audio player
I did not found the important: what DAC system it uses.
There will be plenty of HDD based players in the future...I'ḿ sure -EC- is considering one
Cheers,
M.
Last edited:
Nice screen. The problem is that Chinese re-sellers won't support you. For example, no one answered what the magnitude Vout for the I2S signals is?. If it is less than 1,2V, chances are that our beloved TDA1543 doesn't get exited, and if it is less than 1,5V maybe "I2S attenuator circuit" won't work.
IF I2S were higher, the player could be used with DUAL-MONO DAC. I attach some pictures of my DUAL-MONO in 4x2 version: two towers of 4 chips. Best midbass ever...
One of -EC- Vregs is visible on top.
Cheers,
M.
Attachments
Thanks Max, very useful info for the DIYers who are not prepared with 90 k /like myself/ and could easily get into troubles spearing the Christmas present budget for something looking affordable.
By the way does the above affect the TDA 1541 A as well ?
Regards,
Ignat
We are expectant...
Cheers,
M.
This could handle the job:
http://www.ti.com/lit/ds/symlink/sn74lvc3g34.pdf