Hi swak,
The Mosaic UV DAC has no output buffer (passive output). There are fixed outputs (750 Ohms) and a variable outputs (0 - 5K).
Ternary logic switches switch between -2V8, 0V and +2V8. This logic drives a passive resistor matrix that has 750 Ohms output impedance. The output is DC-coupled since it swings symmetrically around GND, so no coupling caps in the signal path either.
The +2V8 is provided by a discrete quatrode voltage regulator, the tracking -2V8 is derived from a second discrete quatrode regulator and a servo circuit that compares both plus and minus voltages and corrects the negative voltage for tracking. This way there is no DC on the outputs.
The built-in relay volume control consists of relays and resistors (also passive circuit) and controls volume in 32 calibrated steps of 2dB (0.1% thin film precision resistors).
It is possible to lower the volume control impedance to say 5K or 2K5 easily, however, this would also reduce the output voltage because of the 750 Ohms converter output impedance that is loaded by the volume control.
The relay volume control can be controlled by IR remote control or through the host (USB). The digital (software) volume control on the host should always be set to maximum for bit-perfect playback.
The Mosaic UV DAC has no output buffer (passive output). There are fixed outputs (750 Ohms) and a variable outputs (0 - 5K).
Ternary logic switches switch between -2V8, 0V and +2V8. This logic drives a passive resistor matrix that has 750 Ohms output impedance. The output is DC-coupled since it swings symmetrically around GND, so no coupling caps in the signal path either.
The +2V8 is provided by a discrete quatrode voltage regulator, the tracking -2V8 is derived from a second discrete quatrode regulator and a servo circuit that compares both plus and minus voltages and corrects the negative voltage for tracking. This way there is no DC on the outputs.
The built-in relay volume control consists of relays and resistors (also passive circuit) and controls volume in 32 calibrated steps of 2dB (0.1% thin film precision resistors).
It is possible to lower the volume control impedance to say 5K or 2K5 easily, however, this would also reduce the output voltage because of the 750 Ohms converter output impedance that is loaded by the volume control.
The relay volume control can be controlled by IR remote control or through the host (USB). The digital (software) volume control on the host should always be set to maximum for bit-perfect playback.
Hi Fellas!
I am wondering if any of you here have ever tried to run a TDA1541 or a TDA1541A from a 192kHz 24bit source with I2S NOS mode, directly coming from an optically isolated USB to I2S converter.
I would like to build a simple DAC, and i thought i might ask some more experienced folks.
I am wondering if any of you here have ever tried to run a TDA1541 or a TDA1541A from a 192kHz 24bit source with I2S NOS mode, directly coming from an optically isolated USB to I2S converter.
I would like to build a simple DAC, and i thought i might ask some more experienced folks.
Hi neptunes4you,
First of all, best wishes for 2018.
For 192/24 I2S source it is most likely to have 64 bits / frame. This means that we need 64 bit clock pulses to clock in the stereo data into the TDA1541A through I2S. The required bit clock rate for 192 KHz sample rate would be 192,000 * 64 = 12.288 MHz
The TDA1541A can only handle half of that. So it would be limited to 96 KHz.
When using a receiver or converter that offers I2S with 32 bits / frame (16 bit stereo), the required bit clock frequency would be 192,000 * 32 = 6.144 MHz and the TDA1541A could handle that.
I personally would advise to run the TDA1541A in simultaneous mode (connect pin 27 to -5V) as this offers a number of advantages like lower jitter (dedicated LE sample timing signal) and higher sample rate support (L+R data can be clocked in simultaneously).
With TDA1541A running in simultaneous mode it could support 384/16.
isolators have to offer certain bandwidth for the data and clock signals to pass. Interference that falls within this isolator bandwidth will simply pass as well. So isolators do not remove interference, they are primarily intended for safety (isolation) and breaking up ground loops.
Isolators will add significant jitter (100ps and up), and noise. Compared to a direct coupling, performance will be reduced (more jitter, more noise).
First of all, best wishes for 2018.
For 192/24 I2S source it is most likely to have 64 bits / frame. This means that we need 64 bit clock pulses to clock in the stereo data into the TDA1541A through I2S. The required bit clock rate for 192 KHz sample rate would be 192,000 * 64 = 12.288 MHz
The TDA1541A can only handle half of that. So it would be limited to 96 KHz.
When using a receiver or converter that offers I2S with 32 bits / frame (16 bit stereo), the required bit clock frequency would be 192,000 * 32 = 6.144 MHz and the TDA1541A could handle that.
I personally would advise to run the TDA1541A in simultaneous mode (connect pin 27 to -5V) as this offers a number of advantages like lower jitter (dedicated LE sample timing signal) and higher sample rate support (L+R data can be clocked in simultaneously).
With TDA1541A running in simultaneous mode it could support 384/16.
isolators have to offer certain bandwidth for the data and clock signals to pass. Interference that falls within this isolator bandwidth will simply pass as well. So isolators do not remove interference, they are primarily intended for safety (isolation) and breaking up ground loops.
Isolators will add significant jitter (100ps and up), and noise. Compared to a direct coupling, performance will be reduced (more jitter, more noise).
Hi!
Happy new year to you as well!
I was thinking about using an ADUM1400 isolator, it seems like a very capable and fast IC.
Regarding the I2S output, i am using Foobar2000 with the WASAPI plugin. It is set to output 32bits. My converter is a CM6631A based Chinese PCB. I have an oscilloscope at home, i will try to make measurements. It would be really cool to get the TDA1541 to play 192kHz 16bits through I2S. Also, i am thinking about making a TDA1543 based DAC as well, with the same 192kHz "dream".
I -EC-. Happy New year for you and for all DAC building lovers.
I forgot to write about my first New Year party with the Mosaic.
It was a joy of undistorted, analog sounding, detailed and rhythmic (tropical; pop) music.
Sudenly (it may have been the drinks...) I realized that I was listening to probably the finest digital-to-analog converter on the planet!
Then I followed: my Stevens&Billington's TVC is one of the finest "passive" volume controls there is...and my Tannoy Autograph enclosures are probably the cleverest enclosures ever designed, though my passive crossover and one of my big Beyma coaxials let a lot to be desired...and my 8 Watt Tokin VFET single ended Class A amp is not bad at all, capable of driving my Autographs to Disco party levels with ease...boy, I was (still am) a happy man!
Nowadays I moved to build and mod amplifiers, after understanding that I cannot beat John in DAC arena...seriously guys, buy this almost perfect DAC before the price climbs to stratospheric Hi End level.
Cheers,
M.
I forgot to write about my first New Year party with the Mosaic.
It was a joy of undistorted, analog sounding, detailed and rhythmic (tropical; pop) music.
Sudenly (it may have been the drinks...) I realized that I was listening to probably the finest digital-to-analog converter on the planet!
Then I followed: my Stevens&Billington's TVC is one of the finest "passive" volume controls there is...and my Tannoy Autograph enclosures are probably the cleverest enclosures ever designed, though my passive crossover and one of my big Beyma coaxials let a lot to be desired...and my 8 Watt Tokin VFET single ended Class A amp is not bad at all, capable of driving my Autographs to Disco party levels with ease...boy, I was (still am) a happy man!
Nowadays I moved to build and mod amplifiers, after understanding that I cannot beat John in DAC arena...seriously guys, buy this almost perfect DAC before the price climbs to stratospheric Hi End level.
Cheers,
M.
Hi RonSSNova,
We have few Mosaic UV DACs left, these are the last one's. We won't be starting up a new batch. Manufacturing these fully handcrafted DACs is problematic and very time consuming.
Another problem are the external factors we have no control over like mains power quality and digital audio source properties. Other DACs also suffer from these external factors.
So we plan to redesign the entire DAC once again in order to eliminate the influence of these external factors and simplify the design so manufacturing becomes easier and faster.
Prototypes are up and running, showing this is indeed possible.
We have few Mosaic UV DACs left, these are the last one's. We won't be starting up a new batch. Manufacturing these fully handcrafted DACs is problematic and very time consuming.
Another problem are the external factors we have no control over like mains power quality and digital audio source properties. Other DACs also suffer from these external factors.
So we plan to redesign the entire DAC once again in order to eliminate the influence of these external factors and simplify the design so manufacturing becomes easier and faster.
Prototypes are up and running, showing this is indeed possible.
Can you tell us a little more?
For mains power I think I have solved the problem using a good isolation transformer upfront. On the digital source side, I am quite happy with an Allo USBridge. What type of input are you planning and what further benefits do you forsee?
I confirm that the TDA1541a runs beautifully at 192k in I2S NOS mode. It sounds even better in simultaneous mode. I can endorse from personal experience the following chain. BBB to AKO S03 isolator/reclocker to IanCanada PCM to simultaneous converter to Ryanj 1541a PCB. All are topics on DIYAUDIO and readily available. I started with USB. Good but eliminating USB is a big step up so the BBB or a RPi directly outputting I2S is great. Most has been positively influenced by the knowledge in this thread.
Sounds like Mosaic would be a step forward again.
Hi John, are you keeping USB as the main and only input for your new version?
Happy New Year everyone! 
Yes, I can confirm that the TDA1541A supports a stream of 352 kHz / 16 bits. I used the TDA1541A in a simultaneous mode with a digital filter SM5813APT. This DF supports 8x mode, i.e. 44,1*8=352,8 kHz.
Yes, I can confirm that the TDA1541A supports a stream of 352 kHz / 16 bits. I used the TDA1541A in a simultaneous mode with a digital filter SM5813APT. This DF supports 8x mode, i.e. 44,1*8=352,8 kHz.
Hi ernesternest,
Yes, we keep the proven USB (UAC1, adaptive) as one and only digital audio interface.
The Mosaic II prototype offers a -very- high degree of source immunity (the source basically doesn't matter anymore as long as it's bit-perfect).
It is battery powered, this will simply eliminate mains power supply and voltage regulator related issues.
Yes, we keep the proven USB (UAC1, adaptive) as one and only digital audio interface.
The Mosaic II prototype offers a -very- high degree of source immunity (the source basically doesn't matter anymore as long as it's bit-perfect).
It is battery powered, this will simply eliminate mains power supply and voltage regulator related issues.
That is THE best way to eliminate the mains problem. Using LiPo4 batteries makes sure the powersource is very quiet. At a time I even had USB reciever -fifo -dac preapm and Power amp running on battery power. And plenty of power for the speakers. UcD 700 with a discrete input stage on 2 x 96 V batteries.
Sounded really good. Only problem I love my 6C33 OLT tube amplifier, and it will be very hard to run that one on batteries!
Hi John, thank you; interesting. So basically USB input and PSU are being modified. The DAC itself stays in principle and will be produced more efficiently.
Will you be able to keep the price point where it is now?
Cheers, Ernst
Being able to use any source USB device would be brilliant. There is so much time, money and energy wasted on trying to optimize usb at the source... Can you tell us a little more on how your DAC would be "source immune"?