Re: HIFIDIY.NET HDD APE player
Where did you find this one? Couldn't find it on hifidiy.net.
2A3SET said:
Where did you find this one? Couldn't find it on hifidiy.net.
SD player with i2s out
Look on this forum
Thread: (Digital) Ultimate Source (from 2006 !!)
Member: KOON3876
Onno
Look on this forum
Thread: (Digital) Ultimate Source (from 2006 !!)
Member: KOON3876
Onno
dear Onno,regarding your posting #2535 i have 1 question about it:
you stated there that DEM out clock (352khz) is obtained from pin12 on 74HC161,but that IC's pin 12 is labelled Q2, so if Q0 = 5,64mhz and Q1 = 2,82mhz that means Q2 has 1,4 mhz far far away from the desired 352khz value...
i also attached your diagram
you stated there that DEM out clock (352khz) is obtained from pin12 on 74HC161,but that IC's pin 12 is labelled Q2, so if Q0 = 5,64mhz and Q1 = 2,82mhz that means Q2 has 1,4 mhz far far away from the desired 352khz value...
i also attached your diagram
An externally hosted image should be here but it was not working when we last tested it.
Dem Clock
Dear Luxury54
The 74HC(T)161 in the timing module from the DI8M dac has as input BCK = 2822.4 Khz.
Q2 (pin 12) gives 2822.4 : 8 = 352.8 Khz !
Regards
Onno
luxury54 said:
Dear Luxury54
The 74HC(T)161 in the timing module from the DI8M dac has as input BCK = 2822.4 Khz.
Q2 (pin 12) gives 2822.4 : 8 = 352.8 Khz !
Regards
Onno
thanks for clarification,that explains a lot
i thought it was the 11.2896 mhz master clock fed into it 😀
i thought it was the 11.2896 mhz master clock fed into it 😀
Hi folks
Q for EC design
*
How to obtain optical spdif from mac power-book G4 ?
*
it exist at the headphones jack inside...
I used mini adaptor to get standard toslink cable.
but the system did not recognise change.
*
Maybe additional driver software I need?
Q for EC design
*
How to obtain optical spdif from mac power-book G4 ?
*
it exist at the headphones jack inside...
I used mini adaptor to get standard toslink cable.
but the system did not recognise change.
*
Maybe additional driver software I need?
Hi Zoran,
First go to "System Preferences > Sound".
Then select "Digital Out", and uncheck "mute" (square should be blank).
Sample rate can be set going to "Utilities > Audio MIDI Setup".
Set both "Default Output and System Output" to Built-in Audio.
Set source to "Digital Out".
Sample rate can be set to 32 ... 96 KHz using Format. Select either 2-ch 16 bits or 2-ch 24 bits.
Make sure that your mac has Toslink integrated in the Earphone jack, this is usually labelled "optical out". When the Toslink cable is inserted and everything functions well, the other side of the Toslink interlink should emit red light.
Unfortunately the (power) mac on-board Toslink output doesn't seem to provide bit-perfect playback, so equipment connected to this interface won't perform optimally.
The easiest way to go then, is to connect an Apple Airport Express module through a (wireless) network. This provides fixed 44.1/16 bit-perfect output, but there is still the source jitter problem.
Due to computer "issues" like failing to provide bit-perfect playback without extensive tweaking, power consumption, fan noise, and the problematic SPDIF / USB interface, I plan to go a completely different route.
The answer is to construct a small Audiophile SD-card player that's integrated in the DAC. Now we have a number of major advantages:
- Guaranteed bit-perfect playback.
- SD-card player clock = DAC master clock, master clock can be located close to the DAC chip.
- I2S interface (skipping SPDIF / USB).
- Completely silent operation.
- Very low power consumption.
- Multiple CDs or entire play-list on a single SD-card.
- Very small size (estimated size of the entire player, 2 x 10cm). This enables the player to be integrated in the DAC.
First go to "System Preferences > Sound".
Then select "Digital Out", and uncheck "mute" (square should be blank).
Sample rate can be set going to "Utilities > Audio MIDI Setup".
Set both "Default Output and System Output" to Built-in Audio.
Set source to "Digital Out".
Sample rate can be set to 32 ... 96 KHz using Format. Select either 2-ch 16 bits or 2-ch 24 bits.
Make sure that your mac has Toslink integrated in the Earphone jack, this is usually labelled "optical out". When the Toslink cable is inserted and everything functions well, the other side of the Toslink interlink should emit red light.
Unfortunately the (power) mac on-board Toslink output doesn't seem to provide bit-perfect playback, so equipment connected to this interface won't perform optimally.
The easiest way to go then, is to connect an Apple Airport Express module through a (wireless) network. This provides fixed 44.1/16 bit-perfect output, but there is still the source jitter problem.
Due to computer "issues" like failing to provide bit-perfect playback without extensive tweaking, power consumption, fan noise, and the problematic SPDIF / USB interface, I plan to go a completely different route.
The answer is to construct a small Audiophile SD-card player that's integrated in the DAC. Now we have a number of major advantages:
- Guaranteed bit-perfect playback.
- SD-card player clock = DAC master clock, master clock can be located close to the DAC chip.
- I2S interface (skipping SPDIF / USB).
- Completely silent operation.
- Very low power consumption.
- Multiple CDs or entire play-list on a single SD-card.
- Very small size (estimated size of the entire player, 2 x 10cm). This enables the player to be integrated in the DAC.
Hi John,
how far is this project with DI1M and SD player?
It is also practical idea to you SD card.
Thanks.
Bostjan
how far is this project with DI1M and SD player?
It is also practical idea to you SD card.
Thanks.
Bostjan
Hi jameshillj,
This still requires a standard interface, USB or (wireless) network, and a computer running during the time music is being played, drawing lots of power and making (lots of) noise. The idea was to make a low-power, noiseless stand-alone player that's so small, it can be easily integrated in the DAC. This way I2S output from the player can be routed to the DAC chip using short connections. By keeping the player physically small, and using as few components as possible, produced interference can be kept low. By running the DSP on the buffered DAC master clock, the DAC runs in sync with the player, and inter modulation between multiple crystal oscillators is avoided.
TS431 (Taiwan Semiconductor), max. noise level is below 60nV root Hz, average noise level is around 48nV root Hz, dynamic impedance is 0.22 Ohm typical.
LM336 (TI), max. noise level is below 225nV root Hz, average noise level is around 110nV root Hz, dynamic impedance is 0.27 Ohm typical.
It should be possible, but the bias current adds to the DAC output current, translating to a voltage across the I/V resistor. The D1M trans-impedance converter now runs without bias current sources.
I did add a resistor (approx 680 Ohms) in parallel with a small capacitor (approx. 100pF) between DAC output and reference voltage or GND (TDA1541A). This reduces HF noise, and reduces ac peak voltage at the DAC output when T1 stops conducting. The HF interference is highest with maximum DAC output current.
This still requires a standard interface, USB or (wireless) network, and a computer running during the time music is being played, drawing lots of power and making (lots of) noise. The idea was to make a low-power, noiseless stand-alone player that's so small, it can be easily integrated in the DAC. This way I2S output from the player can be routed to the DAC chip using short connections. By keeping the player physically small, and using as few components as possible, produced interference can be kept low. By running the DSP on the buffered DAC master clock, the DAC runs in sync with the player, and inter modulation between multiple crystal oscillators is avoided.
TS431 (Taiwan Semiconductor), max. noise level is below 60nV root Hz, average noise level is around 48nV root Hz, dynamic impedance is 0.22 Ohm typical.
LM336 (TI), max. noise level is below 225nV root Hz, average noise level is around 110nV root Hz, dynamic impedance is 0.27 Ohm typical.
It should be possible, but the bias current adds to the DAC output current, translating to a voltage across the I/V resistor. The D1M trans-impedance converter now runs without bias current sources.
I did add a resistor (approx 680 Ohms) in parallel with a small capacitor (approx. 100pF) between DAC output and reference voltage or GND (TDA1541A). This reduces HF noise, and reduces ac peak voltage at the DAC output when T1 stops conducting. The HF interference is highest with maximum DAC output current.
Hi Bostjan,
After some initial problems accessing the SD card (turned out to be yet another Microchip pin assignment problem with analogue I/O pins), the FAT32 file system is up and running, allowing support for large capacity memory cards. Master clock strategy is sorted out (DSP must be able to run on the DAC master clock). We now use 11.2896 MHz, this is divided-by-2 and multiplied by 16 inside the DSP, resulting in 11.2896 /2 * 16 / 2 = 45.1584 MHz system clock, providing approx. 22.5 MIPs. Most of this capacity is probably required to read some data from an SD-card, thanks to the "brilliant" FAT32 file system, but this is required for "compatibility". The I2S generation is handled by on-chip hardware, and doesn't put any load on the DSP.
The on-chip I2S hardware interface was successfully initialized (after some more problems), and is working fine now. Test pattern was output and I2S signals look fine.
Now it's time for the real work, reading data from the SD-card, and outputting it as an I2S stream.
We are currently using a 28-pin DSPIC from Microchip.
My brother is working very hard on the software for this player, and with the initial problems solved, I guess we have a working prototype soon.
The SD-card was chosen because of the simple straight-forward SPI interface. I don't want any connection between computer and DAC, so I use semiconductor memory. I also want to listen to the DAC without having the computer up and running all day
Meanwhile the D1M prototype has been further optimized using differential bit clock injection, new I2S attenuator circuits, and of course the new miniature low-noise shunt regulators. The trans-impedance converter was further optimized as well.
When the SD-card player is ready and integrated in the D1M, a lot of stuff can be removed (tracker, 2 x SPDIF receiver, and 2 voltage regulator modules). I expect much higher sound quality than ever before (using a perfect source and minimizing timing jitter).
After some initial problems accessing the SD card (turned out to be yet another Microchip pin assignment problem with analogue I/O pins), the FAT32 file system is up and running, allowing support for large capacity memory cards. Master clock strategy is sorted out (DSP must be able to run on the DAC master clock). We now use 11.2896 MHz, this is divided-by-2 and multiplied by 16 inside the DSP, resulting in 11.2896 /2 * 16 / 2 = 45.1584 MHz system clock, providing approx. 22.5 MIPs. Most of this capacity is probably required to read some data from an SD-card, thanks to the "brilliant" FAT32 file system, but this is required for "compatibility". The I2S generation is handled by on-chip hardware, and doesn't put any load on the DSP.
The on-chip I2S hardware interface was successfully initialized (after some more problems), and is working fine now. Test pattern was output and I2S signals look fine.
Now it's time for the real work, reading data from the SD-card, and outputting it as an I2S stream.
We are currently using a 28-pin DSPIC from Microchip.
My brother is working very hard on the software for this player, and with the initial problems solved, I guess we have a working prototype soon.
The SD-card was chosen because of the simple straight-forward SPI interface. I don't want any connection between computer and DAC, so I use semiconductor memory. I also want to listen to the DAC without having the computer up and running all day
Meanwhile the D1M prototype has been further optimized using differential bit clock injection, new I2S attenuator circuits, and of course the new miniature low-noise shunt regulators. The trans-impedance converter was further optimized as well.
When the SD-card player is ready and integrated in the D1M, a lot of stuff can be removed (tracker, 2 x SPDIF receiver, and 2 voltage regulator modules). I expect much higher sound quality than ever before (using a perfect source and minimizing timing jitter).
Hi John,
Indeed this seems to be the right way. Great to have a brother who does the software 😀
Question: how will the Human Interface look like? What functionality are you (or your brother) planning?
best regards
doede
Indeed this seems to be the right way. Great to have a brother who does the software 😀
Question: how will the Human Interface look like? What functionality are you (or your brother) planning?
best regards
doede
' morning, John.
The SD card system is looking very promising - I see now why you've used the TL431s and that o/p stage seems to be getting simpler all the time.
Thanks for the information, and your patience.
The SD card system is looking very promising - I see now why you've used the TL431s and that o/p stage seems to be getting simpler all the time.
Thanks for the information, and your patience.
No room for lossless compression in there? (I know... Would need a huge DSP and Linux on it.)
Hi EC-Designs
Would it be possible to start a new thread focused on computer interface issues so that this excellent thread remains mainly a repository and easy reference for the incredible info pertinent to the TDA1541A Dac?
Thanks for sharing your designs with us.
Rio
Would it be possible to start a new thread focused on computer interface issues so that this excellent thread remains mainly a repository and easy reference for the incredible info pertinent to the TDA1541A Dac?
Thanks for sharing your designs with us.
Rio
Hi John,
"Hi Builder Brad,
quote:
Do you have any pcbs for sale that would let me build one of the earlier DACs based on 4 x TDA1541 chips. I have a reclocked I2S signal, so would only be looking for the DAC chip modules, Interpolation module PS and I/V stage.
I still have the following PCBs that could be interesting for experimenting:
- DI8 mainboard
- TDA1541A PCB
- I/V converter PCB
- Interpolator PCB
I attached a photograph of the mainboard that can hold up to 8 TDA1541A modules, the interpolator, and 2 I/V converter PCBs.
"
I am interested in these modules - please let me know how much you would like for these modules inc. shipping to the UK.
Paypal OK for you?
Brad
"Hi Builder Brad,
quote:
Do you have any pcbs for sale that would let me build one of the earlier DACs based on 4 x TDA1541 chips. I have a reclocked I2S signal, so would only be looking for the DAC chip modules, Interpolation module PS and I/V stage.
I still have the following PCBs that could be interesting for experimenting:
- DI8 mainboard
- TDA1541A PCB
- I/V converter PCB
- Interpolator PCB
I attached a photograph of the mainboard that can hold up to 8 TDA1541A modules, the interpolator, and 2 I/V converter PCBs.
"
I am interested in these modules - please let me know how much you would like for these modules inc. shipping to the UK.
Paypal OK for you?
Brad
could somebody tell me what would be the best masterclock distribution so that this would be minimally afected by the loads that are attached to it (in my case the KWAK CLOCK 7)
i'm interested if some combinations of buffers and resistors in series with clock output could be benefical here
thanks !
i'm interested if some combinations of buffers and resistors in series with clock output could be benefical here
thanks !
Hi dddac,
I just started a new thread about this SD-card player project, as suggested by some diyAudio members.
The name of the thread is "Lossless SD-card player"
I will post all information there.
I just started a new thread about this SD-card player project, as suggested by some diyAudio members.
The name of the thread is "Lossless SD-card player"
I will post all information there.
Hi, luxury54
It's best to save the cleanest master clock for the DAC, this is the un-buffered master clock output. Logic gates / buffers added in series with the master clock signal will very likely increase existing master clock jitter amplitude.
It's best to use differential master clock distribution to the DAC chip, preventing return paths through GND as these will very easily pick-up interference.
I had best results (with the TDA154x) by making a small ring-core pulse transformer with approx. 15 turns on the primary winding and 1 turn on the secondary. This provides a galvanic insulated differential clock at the output. The bit clock amplitude for the TDA154x is around 400 ... 600mVpp. The primary is connected between master clock GND and to the master clock output, using a ceramic cap of around 2.2nF. The secondary winding is loaded with a 47 Ohm resistor.
The bit clock input requires a bias voltage of approx. 1.2V. I used a TLV431 reference diode (1.25V). One side of the secondary winding connects to this 1.25V reference voltage, the other goes to the bit clock input. This provides approx. 400 ... 600mVpp signal on a 1.25V bias voltage.
All other circuits requiring a bit clock could be connected through a clock buffer (74S125 for example), and suitable damping resistors (100R ... 1K) or resistive attenuators.
It's best to save the cleanest master clock for the DAC, this is the un-buffered master clock output. Logic gates / buffers added in series with the master clock signal will very likely increase existing master clock jitter amplitude.
It's best to use differential master clock distribution to the DAC chip, preventing return paths through GND as these will very easily pick-up interference.
I had best results (with the TDA154x) by making a small ring-core pulse transformer with approx. 15 turns on the primary winding and 1 turn on the secondary. This provides a galvanic insulated differential clock at the output. The bit clock amplitude for the TDA154x is around 400 ... 600mVpp. The primary is connected between master clock GND and to the master clock output, using a ceramic cap of around 2.2nF. The secondary winding is loaded with a 47 Ohm resistor.
The bit clock input requires a bias voltage of approx. 1.2V. I used a TLV431 reference diode (1.25V). One side of the secondary winding connects to this 1.25V reference voltage, the other goes to the bit clock input. This provides approx. 400 ... 600mVpp signal on a 1.25V bias voltage.
All other circuits requiring a bit clock could be connected through a clock buffer (74S125 for example), and suitable damping resistors (100R ... 1K) or resistive attenuators.
@ECdesign about powerbook G4 digital out...
*
in "System Preferences > Sound"
I do not have "Digital Out" option at all...
that is the problem.
*
what should i suppose to install to activate this option
because in 15in powerbook G4 manual clearly said
that the digital toslink input & output are standard
throught on board 3.5mm jack mic/headph. ,
but with a adaptor for tos link...
*
sory for the off topic please but I need the help
*
in "System Preferences > Sound"
I do not have "Digital Out" option at all...
that is the problem.
*
what should i suppose to install to activate this option
because in 15in powerbook G4 manual clearly said
that the digital toslink input & output are standard
throught on board 3.5mm jack mic/headph. ,
but with a adaptor for tos link...
*
sory for the off topic please but I need the help
