Hi ECdesign,
Please forgive me if i'm asking a stupid question here,
because most of the details discussed here are way over my head.
I understand that the cd-image is bit perfect copied on
the SD card to be played with a maximal limitation of jitter. Since the transfer fro cd to SD card is made on computer, can this perfect image be transformed with something like BruteFIR DRC, without loosing the benefits your after (since if I understand it the Brutefir is
still in digital domain) ?
This way a DSP between CDplayer and speaker can be ommited and the need for a computer (with the associated noise and or interference) can be ommited during playback.
Looking forward for your insight,
Edwin
Please forgive me if i'm asking a stupid question here,
because most of the details discussed here are way over my head.
I understand that the cd-image is bit perfect copied on
the SD card to be played with a maximal limitation of jitter. Since the transfer fro cd to SD card is made on computer, can this perfect image be transformed with something like BruteFIR DRC, without loosing the benefits your after (since if I understand it the Brutefir is
still in digital domain) ?
This way a DSP between CDplayer and speaker can be ommited and the need for a computer (with the associated noise and or interference) can be ommited during playback.
Looking forward for your insight,
Edwin
EdwinR3 said:
What you can do. Run your files through brutefir and save them as 16/44.1 files. These files
you save to the SD card. This way you can accomplish DRC, partial speaker linearization and time allignment of the speakers even with such a player. To apply a better dither algorithm you could run the brutefir output (without dithering e.g. 32bit float) into Sox to apply a more advanced dither algorithm and some other goodies.
You need to compare of course what'll be the better choice after all.
The filtering and dithering of brutefir with all its benefits might sound worse then playing the original data.
In any case I presume the sound will be considerably better than a standard realtime setup with brutefir and direct soundcard output.
John,
Please forgive an old question - I've got left behind (again!) and appearing dumb seems better than staying stupid about this
this ...
As I follow your summary (post 180, above) of the current 1543 o/p stage, the o/p pins are maintained at +3.3 volts via a rather good quality 2 stage supply and the 2mA o/p current develops the O/P signal across the 680R resistor with another high quality + volts, and dc isolated with a high quality cap,yes?
I assume that the "jitter feedback" network is still connected to the o/p pin as well?
So, now there aren't any transistors to the o/p stage at all?
Does this same system apply to the 1541A? [just when I thought the O/P stage was completed!]
Sorry for all the questions - hoping that all your development work can be applied to the 1541A chip later on.
Please forgive an old question - I've got left behind (again!) and appearing dumb seems better than staying stupid about this
this ...
As I follow your summary (post 180, above) of the current 1543 o/p stage, the o/p pins are maintained at +3.3 volts via a rather good quality 2 stage supply and the 2mA o/p current develops the O/P signal across the 680R resistor with another high quality + volts, and dc isolated with a high quality cap,yes?
I assume that the "jitter feedback" network is still connected to the o/p pin as well?
So, now there aren't any transistors to the o/p stage at all?
Does this same system apply to the 1541A? [just when I thought the O/P stage was completed!]
Sorry for all the questions - hoping that all your development work can be applied to the 1541A chip later on.
Hi jameshillj
I added a screen shot of the SD-player output stage schematics.
I2S DATA signal passes a I2S attenuator / power limiter FB13, R32, R24, R29, N13.
I2S WS signal passes a I2S attenuator / power limiter FB12, R31, R28, R23, N12.
I2S BCK is derived from ultra high-speed single D flip-flop U8, and passes dynamic jitter attenuator D3, D4, D5, R5, R6, R9, C16 and C17.
TDA1543 outputs drive I/V resistors R16 and R17. Outputs are AC coupled, C19, and C20. The strange symbol for the output caps was required to indicate to my prehistoric DOS CAD system that 4 sets of pads were used to support different types and shapes of output coupling caps.
3VDR reference voltage is derived from a cascaded LED shunt regulator.
The "high quality" cap that performed almost equally well as a Vcap tin / teflon cap in this specific design, was a cheap miniature 1uF WIMA polyester cap (pitch 2.5mm). Later attempts to replace this cap with "better" polypropylene caps failed, these WIMAs performed amazingly well, perhaps it's the small size.
Jitter feedback didn't provide the expected results (too critical). I now use an extreme low jitter 3-crystal clock with amplitude stabilization (XO on the schematic) with a modified sine wave output, and the dynamic jitter attenuator circuit.
No transistors, no OP-amps, no tubes, eliminating all problems these parts may cause.
The TDA1541A is rather picky in this regard. I was able to squeeze maximum amplitude with minimum distortion out of these DAC chips using an external bias voltage circuit that was integrated with the passive I/V resistors. This provided approx. 550mVpp, if you could drive your pre- or power amp with this signal, then yes, you could use the TDA1541A without any transistor, OP-amp or tube too, and benefit from the advantages.
I added a screen shot of the SD-player output stage schematics.
I2S DATA signal passes a I2S attenuator / power limiter FB13, R32, R24, R29, N13.
I2S WS signal passes a I2S attenuator / power limiter FB12, R31, R28, R23, N12.
I2S BCK is derived from ultra high-speed single D flip-flop U8, and passes dynamic jitter attenuator D3, D4, D5, R5, R6, R9, C16 and C17.
TDA1543 outputs drive I/V resistors R16 and R17. Outputs are AC coupled, C19, and C20. The strange symbol for the output caps was required to indicate to my prehistoric DOS CAD system that 4 sets of pads were used to support different types and shapes of output coupling caps.
3VDR reference voltage is derived from a cascaded LED shunt regulator.
The "high quality" cap that performed almost equally well as a Vcap tin / teflon cap in this specific design, was a cheap miniature 1uF WIMA polyester cap (pitch 2.5mm). Later attempts to replace this cap with "better" polypropylene caps failed, these WIMAs performed amazingly well, perhaps it's the small size.
Jitter feedback didn't provide the expected results (too critical). I now use an extreme low jitter 3-crystal clock with amplitude stabilization (XO on the schematic) with a modified sine wave output, and the dynamic jitter attenuator circuit.
No transistors, no OP-amps, no tubes, eliminating all problems these parts may cause.
The TDA1541A is rather picky in this regard. I was able to squeeze maximum amplitude with minimum distortion out of these DAC chips using an external bias voltage circuit that was integrated with the passive I/V resistors. This provided approx. 550mVpp, if you could drive your pre- or power amp with this signal, then yes, you could use the TDA1541A without any transistor, OP-amp or tube too, and benefit from the advantages.
Attachments
Hi EdwinR3,
Yes, the uncompressed WAV file (44.1/16) is simply copied to the SD-card with some restrictions to file name (must start with 2-digit number to indicate track number), and specified directory structure (all WAV files belonging to one CD must be placed in the same directory. The directory name must also start with a two-digit number, indicating disk number.
The extreme low jitter was achieved by the 3-crystal master clock (I added a photographs of the latest version). It runs on very low power (approx. 6mW), has amplitude stabilization (increases PSRR with up to factor 200), and has output properties that no longer require the use of clock buffers or comparators, eliminating the extra jitter these circuits would add.
Low jitter is maintained by using suitable clock signal distribution scheme.
Jitter is further reduced by the dynamic jitter attenuation circuit.
DAC chip on-chip jitter is minimized by using I2S attenuators / power limiters, and a discrete low noise power supply.
You could risk ending up with poorer performance like soundcheck already mentioned.
I prefer to use the original audio data as-is, in the bit-perfect form.
Yes, the uncompressed WAV file (44.1/16) is simply copied to the SD-card with some restrictions to file name (must start with 2-digit number to indicate track number), and specified directory structure (all WAV files belonging to one CD must be placed in the same directory. The directory name must also start with a two-digit number, indicating disk number.
The extreme low jitter was achieved by the 3-crystal master clock (I added a photographs of the latest version). It runs on very low power (approx. 6mW), has amplitude stabilization (increases PSRR with up to factor 200), and has output properties that no longer require the use of clock buffers or comparators, eliminating the extra jitter these circuits would add.
Low jitter is maintained by using suitable clock signal distribution scheme.
Jitter is further reduced by the dynamic jitter attenuation circuit.
DAC chip on-chip jitter is minimized by using I2S attenuators / power limiters, and a discrete low noise power supply.
You could risk ending up with poorer performance like soundcheck already mentioned.
I prefer to use the original audio data as-is, in the bit-perfect form.
Attachments
Hi Hisatugo,
I intend to sell the complete product with or without housing for starters. This way I can guarantee sound quality, and the user can directly experience the performance this concept can offer.
The SD-card player is not a simple building block that solves all problems associated with digital audio sources.
Everything must still be maticulously tuned and matched. I can change one component in the SD-player and the magic is gone, so it's still highly critical. But feel free to experiment after listening to the original SD-player version.
By using PCB-only design without wiring, and an external power supply, I barely managed to construct a product with reproducable sound quality (sound quality between the 4 SD-players I already built, matches very closely).
I attached a photograph of 3 SD-players that were used for direct comparison and duration tests. The other one is at my brother's place for software development.
Other development is testing and comparing cryo-treated TDA1543 chips (thanks goes to a333bt), they were selected from, and compared with untreated DAC chips from the same batch. All I can say for now is that the cryo-treated DAC chips do sound different.
I intend to sell the complete product with or without housing for starters. This way I can guarantee sound quality, and the user can directly experience the performance this concept can offer.
The SD-card player is not a simple building block that solves all problems associated with digital audio sources.
Everything must still be maticulously tuned and matched. I can change one component in the SD-player and the magic is gone, so it's still highly critical. But feel free to experiment after listening to the original SD-player version.
By using PCB-only design without wiring, and an external power supply, I barely managed to construct a product with reproducable sound quality (sound quality between the 4 SD-players I already built, matches very closely).
I attached a photograph of 3 SD-players that were used for direct comparison and duration tests. The other one is at my brother's place for software development.
Other development is testing and comparing cryo-treated TDA1543 chips (thanks goes to a333bt), they were selected from, and compared with untreated DAC chips from the same batch. All I can say for now is that the cryo-treated DAC chips do sound different.
Attachments
This looks like a fantastic idea.
I won't pretend to understand all the details but overall it makes very good sense.
I currently use my pc as my main source but never liked the extra noise associated with this even though i use headphones.
I love the idea of one sd card per album or playlist as it would be just like using my cd's again.
When theyre ready ill buy one please
I won't pretend to understand all the details but overall it makes very good sense.
I currently use my pc as my main source but never liked the extra noise associated with this even though i use headphones.
I love the idea of one sd card per album or playlist as it would be just like using my cd's again.
When theyre ready ill buy one please
Traxmod Open Source SD player split to:http://www.diyaudio.com/forums/showthread.php?s=&threadid=146233
as requested.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.