I didn't compare directly with the nichicons in the dddac supply. The black gates were lying around and I had some good results with these in super E-cap ( they are non-polars) configuration with a previous dac, so I just put them in.
I have not tweaked with other things because I am still optimising with the shunt regulators and sowters, also some salas shunt boards are lying around that I will try first.
The supply and dac boards are difficult to desolder because of the big ground plane (it dissipates all the heat of the soldering iron) so exchanging components is not an easy task. Even when I use a desoldering iron with suction pump
Btw: The nichicons muse that Doede selected are pretty good also!
building project on Audio Creative
Hi Doede,
There is a nice building project of your PCM1794 DAC on Audio-Creative:
Lot's of pic's as well
We gaan verder met Deel 2 van de bouw van de DDDAC 1794 NOS DIY DAC waarin we gedetailleerd ingaan op de bouw er van en mogelijke tweaks. - Audio-Creative
Hi Doede,
There is a nice building project of your PCM1794 DAC on Audio-Creative:
Lot's of pic's as well
We gaan verder met Deel 2 van de bouw van de DDDAC 1794 NOS DIY DAC waarin we gedetailleerd ingaan op de bouw er van en mogelijke tweaks. - Audio-Creative
Choke loaded supply
Hi Supersurfer,
Ref your question whether choke loading is necessary with a regulated supply. The basic approach I have found useful even for solid state is to get the raw supply ripple as smooth as possible. With the choke loaded supply I have used, the out put ripple in the raw supply is in the mv range and is a smooth sine wave of 100hz frequency(Checked this on an oscilloscope). Without choke loading the sharp charging currents have harmonics which you can only reduce with a regulator, not eliminate. This does impact sound quality.
An example of my experience.When upgrading my Revox B77 open reel tape machine, I used a Jung type super regulator (developed by AL Weekes) to replace a 7820 type regulator. This step resulted in a big improvement in sound. When I added a Gyrator to smoothen the ripple from the raw supply, another jump up in the quality. Finally, last week I added a common mode choke after the raw supply as I had concerns about rfi pollution of mains in my location. This resulted in another huge jump in sound quality. Now. listening to 4 track reel tapes from the 60s and 70s, I am able to realise the inherent high quality of recording in some them, when listening through headphones. Did not expect this to happen. Hence, my approach re PSU.
To summarise, a well smoothed raw supply with only 100hz ripple in the raw supply may enable the regulator to perform at its optimium level and help in obtaining the best possible sound from the DAC.
Sorry for the long message. Hope, this is of assistance.
Cheers!
Hi Supersurfer,
Ref your question whether choke loading is necessary with a regulated supply. The basic approach I have found useful even for solid state is to get the raw supply ripple as smooth as possible. With the choke loaded supply I have used, the out put ripple in the raw supply is in the mv range and is a smooth sine wave of 100hz frequency(Checked this on an oscilloscope). Without choke loading the sharp charging currents have harmonics which you can only reduce with a regulator, not eliminate. This does impact sound quality.
An example of my experience.When upgrading my Revox B77 open reel tape machine, I used a Jung type super regulator (developed by AL Weekes) to replace a 7820 type regulator. This step resulted in a big improvement in sound. When I added a Gyrator to smoothen the ripple from the raw supply, another jump up in the quality. Finally, last week I added a common mode choke after the raw supply as I had concerns about rfi pollution of mains in my location. This resulted in another huge jump in sound quality. Now. listening to 4 track reel tapes from the 60s and 70s, I am able to realise the inherent high quality of recording in some them, when listening through headphones. Did not expect this to happen. Hence, my approach re PSU.
To summarise, a well smoothed raw supply with only 100hz ripple in the raw supply may enable the regulator to perform at its optimium level and help in obtaining the best possible sound from the DAC.
Sorry for the long message. Hope, this is of assistance.
Cheers!
Hi Bravi, fully agree, the Regulation does it Thing and less garbage in, is less garbage out
Indeed, great article. Even if you don't speak dutch, the many Pictures are telling a good Story as well!
Hi Bravi,
I understand your approach. I have used a shunt regulator in a tube amp that had a tube rectifier and two chokes in front of the shunt regulator. This worked fine except for one thing, the power mosfets were very sucseptable to voltage peaks and I got tired of exchanging them after every break down. So I finally got rid of the shunt regulator and used a simple unregulated supply with: cap,choke,cap,choke,cap,choke,cap. This worked perfectly and I never had the urge to use a supply regulator again.
To put this in perspective; the amp was a single ended type so everything that comes out of the supply will be amplified and heard.
This got me wondering if it would not be an idea to use an unregulated choke supply on the dac. The current draw would probably be fairly steady because the dac chips run in balanced mode.
Can you confirm this Doede?
Selection of shottky diodes and use of resistors around the diodes will also help in stopping inrush peaks. However not as effevtively as a choke, but it is much more compact and cheap.
Regards,
I understand your approach. I have used a shunt regulator in a tube amp that had a tube rectifier and two chokes in front of the shunt regulator. This worked fine except for one thing, the power mosfets were very sucseptable to voltage peaks and I got tired of exchanging them after every break down. So I finally got rid of the shunt regulator and used a simple unregulated supply with: cap,choke,cap,choke,cap,choke,cap. This worked perfectly and I never had the urge to use a supply regulator again.
To put this in perspective; the amp was a single ended type so everything that comes out of the supply will be amplified and heard.
This got me wondering if it would not be an idea to use an unregulated choke supply on the dac. The current draw would probably be fairly steady because the dac chips run in balanced mode.
Can you confirm this Doede?
Selection of shottky diodes and use of resistors around the diodes will also help in stopping inrush peaks. However not as effevtively as a choke, but it is much more compact and cheap.
Regards,
Last edited:
yes, I can confirm... current is very constant, just varying like 5-10mA per Board when 44 or 192 is playing or no Signal is applied...
Choke loaded supply
Hi Supersurfer,
Hope you have downloaded the Power supply modeler from Duncan amps site. A very useful tool to model Power supply behaviour before you build.
Choke loading without regulation may need to be checked to ensure that in start up mode, you do not exceed the max voltage spec of the DAC board. This is particularly important when using SS diodes as rectifiers as you do not have the slow ramp up of voltage that is possible with some tube rectifiers.
Cheers!
Ravi
Hi Supersurfer,
Hope you have downloaded the Power supply modeler from Duncan amps site. A very useful tool to model Power supply behaviour before you build.
Choke loading without regulation may need to be checked to ensure that in start up mode, you do not exceed the max voltage spec of the DAC board. This is particularly important when using SS diodes as rectifiers as you do not have the slow ramp up of voltage that is possible with some tube rectifiers.
Cheers!
Ravi
Hi Ravi,
I use the Duncan tool sometimes. Have you tried your supply without the regulator?
Coolchoke loaded power supply! Have you compaired with the stock power supply design?
I recieved some Salas shunt supply boards today, will try them when the 4 dac boards and sowters are installed and burned in.
@tuyen: please report your findings if you try out the cap modification.
Regards,
Any news about the 4 DAC boards & Sowters yet Supersurfer?
Right, good hint...Choke-Input Need a bit more knowledge to design right...some of this knowledge is hidden in books of the 40s...so, it makes a lot of sense to introduce some R's with wsilicon diadoes in the Bridge plus a shunt-R which consumes from the beginning on current, so that the choke can work immediately and has not this high voltage upswing.
Hi Ravi,
That's correct but keep in mind that the dac boards still have their own regulators so the raw supply voltage is not so critical.
I had another thought: I have some electronic chokes from tentlabs laying around. The MEC 100 can deliver max. 160mA so using two in parallel would probably work for 4 boards. They can be put in place of the first power resistor on the dddac supply board, this makes for a nice choke input filter!
Only thing to keep in mind is that it drops 9v from in- to output so you need to adjust the supply upwards.
I will check the current draw to be sure.
The two extra boards have not arrived so I will refrain from judging the sowters with only two boards (missing some weight in the low and low mid registers, but does sound very dynamic and detailed in the highs)
Last edited:
I would checkwith tentlabs first. If you read the description of the e-choke well, you will see that you Need a strong big filter cap before the e-choke. It Needs a pre-filtering. The e-choke is good for ripple, but nothing for energy storage by the way. A real choke ismuch better, soundwise in my view. I have just exchanged my e-choke against an amorphous core choke and the Advantage in Sound is evident.
Last edited:
Hi Blitz,
A real choke is always better but it's bulky in comparison. You are right, you cannot use it as choke input, than it should be mounted behind the first cap. But I measured current and one board is doing 260mA with shunt regulators mounted so the idea will not work anyway.
The two last boards have arrived this afternoon so tomorrow they will be mounted ( with shunts) and judging of the sowters can start finally!
Than the next step will be to try Salas regulator supplies.
Regards,
A real choke is always better but it's bulky in comparison. You are right, you cannot use it as choke input, than it should be mounted behind the first cap. But I measured current and one board is doing 260mA with shunt regulators mounted so the idea will not work anyway.
The two last boards have arrived this afternoon so tomorrow they will be mounted ( with shunts) and judging of the sowters can start finally!
Than the next step will be to try Salas regulator supplies.
Regards,
this is very cool
I thought about it, but decided I had no time
could you please post circuit and partnumbers?
Sorry for delay, steep learning curve with Eagle PCB software.
I have enclosed design notes including LCD schematic, BOM as requested plus Arduino sketch code.
I don't pretend this is the most elegant solution, but it works.
Best Regards