Hello
Wen doing a reclock and wen we use a receiver chip (WM8804) who use a different xtal frequency than the dac chip reclock xtal, does it do a problems like drop bit or else ?
Thank
Bye
Gaetan
Wen doing a reclock and wen we use a receiver chip (WM8804) who use a different xtal frequency than the dac chip reclock xtal, does it do a problems like drop bit or else ?
Thank
Bye
Gaetan
Asynchronous reclocking always causes missing/duplicate bits every n seconds. Whether it is audible, is a different question.
Hello
Here's some details.
I've just buy two WM8804, one to mod a friend's cd player, and the other one for my diy dac to replace a CS8414.
For my dac it's like this:
WM8804 (12 mhz crystal) --> SAA7220 (11.2896 mhz clock) --> 74VHC175 reclock (11.2896 mhz clock) --> TDA1541A
Thank
Bye
Gaetan
Here's some details.
I've just buy two WM8804, one to mod a friend's cd player, and the other one for my diy dac to replace a CS8414.
For my dac it's like this:
WM8804 (12 mhz crystal) --> SAA7220 (11.2896 mhz clock) --> 74VHC175 reclock (11.2896 mhz clock) --> TDA1541A
Thank
Bye
Gaetan
Last edited:
If I correctly understand your question then the answer is, no, the additional crystal will not cause any more issues than are already caused by asynchronous re-clocking. The WM8804 utilizes a crystal as an external frequency reference for it's internal PLL clock-recovery circuit operation. This second crystal is not used to alter the the embedded clock domain of the received S/PDIF signal, nor will it cause the WM8804 to drop bits. The Wolfson receiver chips offer superior jitter rejection performance.
I suggest that you drop the whole asynchronous re-clocking scheme, as it actually increases total jitter. Instead, synchronously re-clock using the low jitter MCLK output signal of the 8804. This should dramatically reduce total jitter relative an asynchronous re-clocking scheme.
I suggest that you drop the whole asynchronous re-clocking scheme, as it actually increases total jitter. Instead, synchronously re-clock using the low jitter MCLK output signal of the 8804. This should dramatically reduce total jitter relative an asynchronous re-clocking scheme.
Last edited:
Well, in this case you have to use the CLKOUT as the clock source (programmed to 11.2896 MHz) for the SAA7220, not the 12 MHz crystal. You can use CLKOUT also for reclocking BCK.
Hello
I will use the WM8804 in harware mode only.
I will look in the data sheet to try to see how to get a 11.2896 mhz clock out from the WM8804 to feed the SAA7220 and the 74VHC175.
Thank
Bye
Gaetan
I will use the WM8804 in harware mode only.
I will look in the data sheet to try to see how to get a 11.2896 mhz clock out from the WM8804 to feed the SAA7220 and the 74VHC175.
Thank
Bye
Gaetan
Last edited:
You should be able to obtain a synchronous, low-jitter, 11.2896MHz clock directly from the MCLK pin as the WM8804 receives a 44.1KHz sample rate input signal. MCLK must be configured to output a 256Fs mode clock. Check the datasheet hardware master mode configuration options for MCLK.
Last edited:
Hello
I will have two DAC using WM8804, one for my cd player digital output, and the other one will be use for the digital output of my JVC DVD player. So, one DAC will receives a 44.1KHz sample rate input signal, and the other one will receives a 48KHz sample rate input signal.
Does it worth it to use an external 12 mhz clock for the WM8804 ?
Thank
Bye
Gaetan
I will have two DAC using WM8804, one for my cd player digital output, and the other one will be use for the digital output of my JVC DVD player. So, one DAC will receives a 44.1KHz sample rate input signal, and the other one will receives a 48KHz sample rate input signal.
Does it worth it to use an external 12 mhz clock for the WM8804 ?
Thank
Bye
Gaetan
Last edited:
The 12MHz xtal, in receiving mode, is used just to measure the incoming signal data rata and signal it further to a DSP... Also it is used to generate the internal "free runing" frequency, but the PLL loop is the one that ultimatelly decides the jitter and output data - by locking and following the input signal. The Xtall is NOT used for any "reclocking".
But you HAVE to use the 12MHz xtall, cannot "skip it", because you are talking about the hardware mode...
But you HAVE to use the 12MHz xtall, cannot "skip it", because you are talking about the hardware mode...
Last edited:
Hello
Does the WM8804 MCLK pin will give a 11.2896MHz clock (for the reclock) if there is a 48KHz sample rate input signal ?
Thank
Bye
gaetan
Does the WM8804 MCLK pin will give a 11.2896MHz clock (for the reclock) if there is a 48KHz sample rate input signal ?
Thank
Bye
gaetan
Last edited:
Hello
Maby I found another way.
Looking at the SAA7220 data sheet, the crystal frequency at the crystal input pin can be up to 12,42 mhz
How about using one 12 mhz crystal oscillator and 3 buffers to feed the WM8804 and SAA7220 crystal inputs pin and also one 74VHC175 to reclock the SAA7220 ws, bck and data output, would it work ok ?
Thank
Bye
Gaetan
Maby I found another way.
Looking at the SAA7220 data sheet, the crystal frequency at the crystal input pin can be up to 12,42 mhz
How about using one 12 mhz crystal oscillator and 3 buffers to feed the WM8804 and SAA7220 crystal inputs pin and also one 74VHC175 to reclock the SAA7220 ws, bck and data output, would it work ok ?
Thank
Bye
Gaetan
Last edited:
Work what? You are not clear in what are you trying to acomplish...
SAA7220 generates the signals that drive the SAA7210 and ultimatelly controlls the optical pickup. Pins 22 and 23 are providing the servo information (control loop).
8804 regenerates the clocks from the incoming SPDIF signal, not from the xtall. It does not "reclock" nothing, is just a SPDIF receiver, not a ASRC/SRC chip...
The MCLK/BITCLK/LRCLK pins of the 8804 are derived directly from the incoming SPDIF, not from the 12MHz clock.
SAA7220 generates the signals that drive the SAA7210 and ultimatelly controlls the optical pickup. Pins 22 and 23 are providing the servo information (control loop).
8804 regenerates the clocks from the incoming SPDIF signal, not from the xtall. It does not "reclock" nothing, is just a SPDIF receiver, not a ASRC/SRC chip...
The MCLK/BITCLK/LRCLK pins of the 8804 are derived directly from the incoming SPDIF, not from the 12MHz clock.
Hello
I'm not at my best with the english language, so I may have not explained correctly what I was want to do.
Here is it: it would be stand-alone dac, what I was asking is if the SAA7220 and the 74VHC175 would work ok with the 12mhz crystal oscillator ?
Like in the drawing I done and I include.
Only the 74VHC175 would do the reclock job, the SAA7220 are a bit noisy so the 74VHC175 would reclock the SAA7220 signals before going in the TDA1541A.
Thank
Bye
Gaetan
I'm not at my best with the english language, so I may have not explained correctly what I was want to do.
Here is it: it would be stand-alone dac, what I was asking is if the SAA7220 and the 74VHC175 would work ok with the 12mhz crystal oscillator ?
Like in the drawing I done and I include.
Only the 74VHC175 would do the reclock job, the SAA7220 are a bit noisy so the 74VHC175 would reclock the SAA7220 signals before going in the TDA1541A.
Thank
Bye
Gaetan
Attachments
Last edited:
SAA7220 needs 11.2896MHz clock. The 8804 needs a 12MHz in hardware mode - I guess because of internal dividers.
In software mode you can use a 24MHz or a 27MHz xstall clock and get the 11.2896MHz MCLK needed for the 7220.
I doubt that 74VHC175 will "clean" or "reclock" the jitter comming from 7220. Jitter will just propagate through as such.
Most of the people use DIR9001 because doesn't require an external clock.
In software mode you can use a 24MHz or a 27MHz xstall clock and get the 11.2896MHz MCLK needed for the 7220.
I doubt that 74VHC175 will "clean" or "reclock" the jitter comming from 7220. Jitter will just propagate through as such.
Most of the people use DIR9001 because doesn't require an external clock.
Hello
Looking at the data sheet, as you can see in my new image, I was under the impression that the SAA7220 could use a 12MHz crystal oscillator.
If a reclock with the 74VHC175 will not "clean" or "reclock" the jitter comming from the SAA7220, what can clean that jitter ?
I only have the SAA7220 chip for digital filtering. I have two WM8804. And I have six TDA1541A, so I would use them.
I can use the WM8804 in hardware mode only, I'm a bit limited in my knowledges.
Thank you
Bye
Gaetan
Looking at the data sheet, as you can see in my new image, I was under the impression that the SAA7220 could use a 12MHz crystal oscillator.
If a reclock with the 74VHC175 will not "clean" or "reclock" the jitter comming from the SAA7220, what can clean that jitter ?
I only have the SAA7220 chip for digital filtering. I have two WM8804. And I have six TDA1541A, so I would use them.
I can use the WM8804 in hardware mode only, I'm a bit limited in my knowledges.
Thank you
Bye
Gaetan
Attachments
Last edited:
You are correct, it can use 12MHz. But its the clock master (256fs) in the system - what audio do you have to play out at 46.875kHz?
Of course the 74VHC175 will help clean up the jitter. But jitter may well not be the only problem as regards sound quality from the 7220. There could be a high level of common-mode RF too, induced on to the power supply. The 7220 is rather a noisy chip and draws almost 1W of power.
Hello
Well, it seem that the SAA7220 really need a 11.2896MHz clock.
To clean the rf noise coming from it, I would do a separated and filtered power supply for the SAA7220.
I will have two stand-alone DAC using WM8804, SAA7220 and TDA1541A. One for a CD player and one for a DVD player, and since one of those dacs will receives a 48KHz sample rate input signal from a DVD player, the WM8804 MCLK pin may not give a 11.2896MHz clock for the SAA7220 and the 74VHC175.
Maby I should go back to the first ideas of using one 12MHz crystal for the WM8804 and one 11.2896MHz crystal oscillator for the SAA7220 and the 74VHC175.
If the reclock of the SAA7220 output do clean up the jitter just a bit, maby it more simple to not use a reclock.
Thank
Bye
Gaetan
Well, it seem that the SAA7220 really need a 11.2896MHz clock.
To clean the rf noise coming from it, I would do a separated and filtered power supply for the SAA7220.
I will have two stand-alone DAC using WM8804, SAA7220 and TDA1541A. One for a CD player and one for a DVD player, and since one of those dacs will receives a 48KHz sample rate input signal from a DVD player, the WM8804 MCLK pin may not give a 11.2896MHz clock for the SAA7220 and the 74VHC175.
Maby I should go back to the first ideas of using one 12MHz crystal for the WM8804 and one 11.2896MHz crystal oscillator for the SAA7220 and the 74VHC175.
If the reclock of the SAA7220 output do clean up the jitter just a bit, maby it more simple to not use a reclock.
Thank
Bye
Gaetan
Last edited:
If you plan to use 48kHz samplerate too, you should know that 7220 was never thought to work with that samplerate.
In my opinion, TDA1541 and SAA7220 are dinosaures. There are better DAC's today, with OC filter included, easier to implement.
Look at TI, AD, CS portofolio. For example PCM1794 is a great product. hardware mode, way better than 1541... I say that because I have a TDA1541 and few PCM1792.
In my opinion, TDA1541 and SAA7220 are dinosaures. There are better DAC's today, with OC filter included, easier to implement.
Look at TI, AD, CS portofolio. For example PCM1794 is a great product. hardware mode, way better than 1541... I say that because I have a TDA1541 and few PCM1792.
Sure enough, but its pretty close. Its within the max clock rate on the sheet.
I agree on the 7220, but the jury's out on the TDA for me until I try it with a decent OS filter.
Can't see how the PCM1794 is easier to implement for a DIYer than a TDA1541. Its a tiny package for one thing. How is it better? I can see it has more bits, but what else makes it better than TDA?