• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

Pushing the limits of TDA1543 NOS DAC

If you refer to the datasheet (fig5, format of input signals) you'll see that everything is 'seen' by the chip only when there's a clock edge. So only when a clock rising edge follows the WS falling edge is the current sample converted. In fact there's a delay of a clock cycle after that edge before all the serial data has been received so its impossible for the DAC to convert a sample on the WS transition.

Hello, may I ask one more thing:
Using a 2.8224Mhz clock connected to the dac achieves low jitter and can easily translate to 44.1Khz by means of a 74HC4040 clock divider.
But should I need 22.5792Mhz on another part of the circuit what IC can I use to multiply the clock?
( The specific case is an sd card reader that I want to slave that uses the 22.5792Mhz)
Thanks :)
 
Whilst clock multiplication is possible - you'd use something like a 74HC4046 or HC7046, I'd not recommend it. Rather use a 22.5792MHz oscillator and divide it down to the frequencies you need. Use a synchronous counter rather than a ripple counter for lowest jitter. Note the HC4040 is a ripple counter so you'd best not use it to create any clock that needs to be lowest jitter.
 
Whilst clock multiplication is possible - you'd use something like a 74HC4046 or HC7046, I'd not recommend it. Rather use a 22.5792MHz oscillator and divide it down to the frequencies you need. Use a synchronous counter rather than a ripple counter for lowest jitter. Note the HC4040 is a ripple counter so you'd best not use it to create any clock that needs to be lowest jitter.

I was wondering how well a ripple counter and jitter go along...you have answered that...!
22.5792Mhz ÷ 44.1Khz = 512 = 2^9
That would make it a 10 bit synchronous counter, any IC you could suggest ?
Thanks B.
 
Disabled Account
Joined 2002
Many still have them in stock. I have them in stock. In fact I will be selling my complete BG stock.

* Just an anecdote: I designed a DAC with another member here. Many were built and I only built one with exclusively BG NX HiQ (in several values). This DAC was given away as a present. I recently listened to this DAC and was blown away by the difference in sound quality. The best sounding of the many versions I built.
 
Last edited:
Disabled Account
Joined 2002
How many do you need ? Without reading this complete thread I can tell I use solid aluminium caps as an alternative but those are out of production too. Easier to find though. I also stock these ;-)

There are not many alternatives except the obvious attempts to sell stuff that is supposedly just as good (for the wallet of the seller).
 
Last edited:
I do not have an exact number just now. But I would still prefer an alternative that was currently in production. What about Acrylic SMD capacitors like the ones recommended by ec-designs as decoupling for the TDA1541? Has anyone tried these as replacement for the non polarized 0,1 uF BG nx types? You got PM
 
THis is an interesting chip as it is easy to connect, however for high end or even average quality it simply will be blown away by the cirrus 24 bit 192khz chip in the iphone.

In theory maybe. But in practice absolutely no.
There are still some highly regarded commercial projects out there actually using the 1543, like CAD, Audio Note and BorderPatrol, just to name a few.