HLLY USB SPDIF/ I2S CONVERTER
 

Hi
I need help to identify the 5v supply point of this board to mod.
Where is the regulator for the 5V ?
I have replaced the 3.3 V line with a salas shunt . The 3.3 v line was derived from the Lm 317 and have good result.
thanks

kp93300

I only see the LM317 & the other 3.3V derived from LM317, the unit is USB powered?

The AMS117 is one 3.3V regulator and LM317 probably was the second 3.3V one (separate powers?).
Supply comes from the jack in right (next to big capacitors and LED). I really doubt that is anything to gain from modding the power supply on a digital-only board.

Hi
Thanks for the reply.
The unit be powered by USB or external dc of 8 to 12 V.
The 3.3 V line is derived from the LM 317 and it has been replaced.
The TCXO works on 5V and I am trying to replace it.

thanks

kp93300

Well, then I don't understand. If you have 5V, LM317 is the 5V regulator then...

A lot of improvement gained by using a good psu for dac, so what's your explanation behind you view here? Care to share..

Why don't power directly the TXCO? when I powered the BII clock with Salas BiB I heard a big improvement so let's go;)

My explanation is that this is not a DAC. It is just a digital converter and since no analog signals are generated, there is not too much to gain in a super clean power supply in the low frequency domain (below 100kHz, like the DAC needs).
At the frequencies involved (12 MHz) and fast slew rates of digital fronts (equivalent of hundreds of MHz), the inductances of the traces between the power supply and consumers (digital cips) basically isolates those two in AC domain. Any "improvement" done in the power supply's HF response will not translate in better voltage for the digital chip (in the frequency range that is relevant here). In DC domain the digial cips are happy even with +/- 10% variation, so is nothing to gain there either.
Actually, that is the reason why not a single comercial regulator that can operate well over 1MHz. Because... is no need of that, the traces will cancel all that "performance".

Because of the high frequancy local capacitors are 100 times more important than the power regulator in digital domain, and that board seems to have very good local decoupling - solid tanatalum with ceramics on every cip supply.

Sure, people claim "big improvements" but none backs that with a real measurements (like the cristal jitter). Actuall those improvements are because the cristal does not inject as much RF in the analog portion of the DAC, not because the new PS somehow did something usefull for the cristal. Nobody showed mesurements to prove improvement of jitter because of a fancy PS (as opposed to just a regular decent IC one).

Hi Merlin,
This is what i intend to do and looking for ideas to do it neatly. i wish to place the clock as close as possible to the board.
There is a constant noise heard through the speaker and coming from the HLLY board.
I hope the new power supply can solve the problem.
Otherwise the SQ is very good .

The 3.3 V is a diy version of the BIB salas shunt and is off board. Nice improvement.
regards
kp93300

Hi Korben

Answer in red:

Cheers
Felipe