I just did it. I've got two boards for experiments and "voicing" the dac to my liking. While waiting for updated version. When comes, it's going to be easy to swap for clear mount on new ones, and no more messing with any experiments. As far as I know, upcoming boards will be 100% compatible with old main board, and with all voltages. Please correct me if I'm wrong on that.
Last edited:
Good move. Doede has confirmed that if and when he updates the DDDAC PCM1794 dac board design it will be backwards compatible with the other boards/mainboard.
Hi Chanh.
What are these resistors 10oHm on the bypas to the analog side? Are they MOX? Does it change anything with sound? Is it worth to experiment with those at all?
Regards
Hi,
They look like kiwame carbon film resistors. You can use these positions to tune the sound balance.
I use Riken ohm carbon film resistors in the power supply path to the analog shunts, this gives more color/texture to the sound.
Standard resistors are metal films, and these tend to sound thin and "metalic" (with one l ;-)
Regards,
They look like kiwame carbon film resistors. You can use these positions to tune the sound balance.
I use Riken ohm carbon film resistors in the power supply path to the analog shunts, this gives more color/texture to the sound.
Standard resistors are metal films, and these tend to sound thin and "metalic" (with one l ;-)
Regards,
Ok so you guys have discovered the analogue direct improvement but why not go the whole hog and have separate digital and analogue supplies? This gives a very noticeable improvement over the single split supply
Thanks supersurfer! Aaand another shopping in the name of great future :-D
I believe the effect of this solution in case of this particular dac is neglectable since digi and analog sides have common negative?
Last edited:
Can you clarify what power supply arrangement you swapped from and to for your digital stuff please?
The common ground doesn't prevent a decent improvement in SQ-there are many decent commercial Dacs that use t his arrangement. I started with the arrangement recently described, of 'direct' to the anaologue shunts, a nd compared this to separate supplies going direct to the the digital and analogue shunts with no resistors or anything else in between. Just an Oscon or Slimic on the cap input. The latter arrangement gave the best SQ
Thanks, but what power supply have you used to supply the power for the digital side? The same again or something different?
I see that the analogue side really benefits from a super low impedance
supply that's able to react quickly to a musically changing and dynamic demand, but I'm not certain what characteristics the digital side would like best. Maybe just as low noise as possible? Is it a problem to feed the digital regs with a heavily regulated supply or is it just a case of trying to keep things as low noise as possible?
Yes, they are kiwame 5W 10ohms resistors. I've chosen it because in ps, I would prefer precision/accuracy. The carbon resistors tend to have higher +- % tolerance. With my setup, being tubes/transformers all the way, with too many carbon resistors added too much coloration and certainly do not always provide the true speed/timing. So really, it is really about balancing your ingredients to suit personal taste and system's synergy.
Note - all my signals path, e.g I2S, are replaced by AudioNote 1W resistors. An expensive exercise there! 😀
Btw Stefan, I hope to also be part of the list receive the first batch of DSD prototype...! 🙂
Sent from my iPhone using Tapatalk
Last edited:
Just technical question that came to my mind doing my rookie logic. As we have these resistors on the path of power suply directly to the shunts, do we realy need them there? From the point of all trials and search for fastest psu, puting resistances on the way seems like small contradiction at he first glance, knowing that shunts will manage the vlotage from raw psu. Or they have also different task apart from sound balance?
I put these in as a form of risk assurance for voltage balance based on my Raw unregulated ps. You can have the lowest Z ps, but is it worth risking your Shunts and DAC? Note my raw unregulated ps does not have any resistors in its path, and my chokes are 0.23Ohms each. These 10-Ohms resistor brought the voltage down to shunts' acceptable voltage input. And I would prefer it as close to the Shunt as possible.
May be Stefan, can elaborate technically better? 😉
Sent from my iPhone using Tapatalk
May be Stefan, can elaborate technically better? 😉
Sent from my iPhone using Tapatalk
In the digital side you need to drop +/- 9V from the incoming 12V to get the required 3,3V.
The value of 10 ohm is as per Doede's schematic and can be applied for the Tent shunts as well. For the 3,3V shunts are required from 7V to 10V as input. In this case having 12V as input the Tent's spec require 6,4 ohm for each volt to drop.
Ok, I'm geting the point. It woild help with equal current distribution to all shunts/boards preventing from current picks and dips I persume.
Yes, I expressed myself poorly, I was thinking about analog part writening above. Thank you Enrico and thank you Chanh. Still in deep
learning process (deduction mode 🙂
PS for Enrico
I'm going to sleep just now 🙂
This would require very carefull thought. My experience is that if ground paths from different supplies cross (i.e. two shortest paths) they start interacting with each other with an unpredictable detrimental effect.
My research and listening tests all point to a low impedance loose psu being best for the digital section. Research suggests oversize is helpful too, but I havnt gone too mad in this area or tested bigger Txs than my current 80watt one. The current to the digital section may appear a steady stream, but is in fact very high frequency switching pulses. I have tried with choke and without choke and have a marginal preference for with choke.
Regarding earthing arrangements, my research leads to the conclusion that it is localised decoupling that is vital, particularly for the digital side. So I use an Oscon and Wima film /foil after as well as before my Belleson regs. I use a similar set up on the analogue side but with Silmics.
On my next build I will use Black Gates on the analogue side, Oscons on the digital but will try the Tent shunts. I will also experiment with / without the Oscon after the shunts.
It's not hard to try this yourself and simply let your ears decide!
Regarding earthing arrangements, my research leads to the conclusion that it is localised decoupling that is vital, particularly for the digital side. So I use an Oscon and Wima film /foil after as well as before my Belleson regs. I use a similar set up on the analogue side but with Silmics.
On my next build I will use Black Gates on the analogue side, Oscons on the digital but will try the Tent shunts. I will also experiment with / without the Oscon after the shunts.
It's not hard to try this yourself and simply let your ears decide!
Regarding grounding issue - I have a SSLV for a digital part which is further fed by shunts and 2 x SSLVs for analogue left and right parts of the circuit (supplies enter the circuit at 1 of 4 boards which are connected to each other). Is it ok from grounding point of view?
Great write-up, James, thanks 🙂
For a lurker like myself it´s a good check-list, if (when 🙂 I bite the bullit 🙂
- also looking forward to any news about I2S connection from a BBB or similar...
Cheers, Finn