But achtung, it's not because you see a chineese in a suit than he's Gary Cooper ! Caps are all different if your system is clear enough to notice it by ears. And indeed a generic chineese cap for instance with non great ESR value will work far better than a low esr Panasonic FR after a 78/79xx regulator. However near the load you need a good decoupling. Often length of the traces between these old regs made things easier. I remember very few cases where a good cap after these regs -I mean at the output pin leg- were improving things. I'm curious though to try an output cap with a higher value than the smoothing cap. In the PSU1 example 10 000 uF in spite of the 4700 uF as the smoothing cap is 6800 uF. Ok a a sim on PSUD can say it to you but often you may have surprise in real world after a listening test.
Anyway for the curious I will try first a bad 220 uF cap at the outpin of the reg,a good Silmic 2 at input pin + its 0.1 uF MKP. near the load, anythinh between 0.1 uF to 100 uF of a good ESR
.
Anyway for the curious I will try first a bad 220 uF cap at the outpin of the reg,a good Silmic 2 at input pin + its 0.1 uF MKP. near the load, anythinh between 0.1 uF to 100 uF of a good ESR
.
First thing is to respect the pitch between cap legs the pcb dictates for good reasons. After it is often a question of taste and about compensation according the rest of your hifi,the regs, their embeded caps, etc. I would say just avoid Elna Silmic II in the analog stages (12V) and start from the BOM.
That is interesting
If you find some progress or a technique how to power it, share it please 
PSU 3 - a discrete one?
Will be there interest for a discrete PSU?
Quite a simple connection, where the output transistor is biased via JFET and stabilized with feedback - transistor and zener diode (like LM4040 for better stability).
Diode bridge + filtration capacitors can be on a separated PCB or on this PCB.
Will be there interest for a discrete PSU?
Quite a simple connection, where the output transistor is biased via JFET and stabilized with feedback - transistor and zener diode (like LM4040 for better stability).
Diode bridge + filtration capacitors can be on a separated PCB or on this PCB.
Attachments
Miro - I'd say there would be interest for existing builders if the existing PSU1/2 could be used before a discrete regulator? Maybe thats just not practical though.
If this isn't a runner - then I'd vote for rectification and filter caps on the one board but maybe with the option to separate the boards so that those who use a 2 box solution can leave the raw DC in another box.
If this isn't a runner - then I'd vote for rectification and filter caps on the one board but maybe with the option to separate the boards so that those who use a 2 box solution can leave the raw DC in another box.
IMHO no needs to use a PSU1 or 2 before... it mays kill all the goodness of such simple but good sounding and dynamic transcient of such shematic.
The MJE is quite famous but EUVL talked as well of some modern Toshibas that are fine to him IIRC for such regs.
The usual CRC could be swapped due to the low needs of current for the dac chip and oap for smd coil : CLC. We could as well use carbon comp resistor to in serie after the diode bridge before the first C. as showed talentuous designers already.
It's also easy to split the board in two for ones needing + 5V for instance for the USB streamers, etc : use two bridged at the output plug and you have a - 0 + reg with two boards near each others. Simple but not simplier
The MJE is quite famous but EUVL talked as well of some modern Toshibas that are fine to him IIRC for such regs.
The usual CRC could be swapped due to the low needs of current for the dac chip and oap for smd coil : CLC. We could as well use carbon comp resistor to in serie after the diode bridge before the first C. as showed talentuous designers already.
It's also easy to split the board in two for ones needing + 5V for instance for the USB streamers, etc : use two bridged at the output plug and you have a - 0 + reg with two boards near each others. Simple but not simplier
Ig, you're saying that even before the discrete schematic dropped you did not love PSU1 and 2?IMHO no needs to use a PSU1 or 2 before... it mays kill all the goodness of such simple but good sounding and dynamic transcient of such shematic.
The MJE is quite famous but EUVL talked as well of some modern Toshibas that are fine to him IIRC for such regs.
The usual CRC could be swapped due to the low needs of current for the dac chip and oap for smd coil : CLC. We could as well use carbon comp resistor to in serie after the diode bridge before the first C. as showed talentuous designers already.
It's also easy to split the board in two for ones needing + 5V for instance for the USB streamers, etc : use two bridged at the output plug and you have a - 0 + reg with two boards near each others. Simple but not simplier
@diyiggy The PCB can be universal, it can follow more designs
- possibility for local feedback
- caps can be installed only if needed
- CRC can be disabled when R is wire jumper (or configured as CLC)
- transistors by choice can be installed
- splitting 2 outputs entirely
- rectification and filter caps
- option to separate the boards
Hi Miro, the reason for investigating an alternative way to power the EASYDAC is just that I do not want to have drained my phones battery in just a few hours, when listening to music.That is interesting
The DAC part needs 380mA, the pcm part takes 20 mA. So I let the USB pcm chip being powered by USB power (phone or computer) and power the DAC now either by a power bank or a second USB cable utilizing the socket I showed above.
Sound is good in that way - better than using a Y USB cable as you proposed. When I tried this, about 100mA came from the phone and 300 mA from the powerbank. Maybe the bringing together 2 differnt currents is good for an external hard drive, yet not for an audio device.
@asilker : no I didn't say that. I say I don't see the goodness by using PSU1/2 in serie right before a PSU3. Maybe I didn't understand what Wood-T-F say ?
I just share some feeling and subjective tweaking experience. Everyone should beleive their own ears, alas it asks you spend time and benchmark different supply, caps arrengement and so on. You may like to try many powersupply because it is a hobby so one likes to swap, tweak, try, shape the sound, etc. According the design and the load, there are place for many different typologies imo. Discrete are cool cause it can be good enough, simple and cheap and low noise enough. An illustration : many like the LT3042 integrated chip but I don't. despite it has low noise very good rejection, I don't like the sound in spite of a different integrated reg chip with not as good number or some discrete even worse but that gives subjective better sounding... power supplies is an important part in our audio hobby and btw it has a section in Diyaudio.
I just share some feeling and subjective tweaking experience. Everyone should beleive their own ears, alas it asks you spend time and benchmark different supply, caps arrengement and so on. You may like to try many powersupply because it is a hobby so one likes to swap, tweak, try, shape the sound, etc. According the design and the load, there are place for many different typologies imo. Discrete are cool cause it can be good enough, simple and cheap and low noise enough. An illustration : many like the LT3042 integrated chip but I don't. despite it has low noise very good rejection, I don't like the sound in spite of a different integrated reg chip with not as good number or some discrete even worse but that gives subjective better sounding... power supplies is an important part in our audio hobby and btw it has a section in Diyaudio.
Last edited: