Good chance that they survived 
Some people claim that higher ccs currents sound better but I'm not able to hear a difference between 50mA to 100mA spare current. So i reduced the current to 80mA (50mA spare + 30mA draw) for the analog rails in my first build - that way I only have to use a really tiny heatsink on the IRF9530/530.
Some people claim that higher ccs currents sound better but I'm not able to hear a difference between 50mA to 100mA spare current. So i reduced the current to 80mA (50mA spare + 30mA draw) for the analog rails in my first build - that way I only have to use a really tiny heatsink on the IRF9530/530.
If you knew the number of cd players I broke in 25 years of diy ... lucky enough I bought them very cheap second hand when CD players were often seen at 10 to 30 euros... Most of all Marantz & Philips, but not only... 
There is no Chief Omelet without breaking many eggs ... that's why tubes are not for me
... lucky enough I bought them very cheap second hand when CD players were often seen at 10 to 30 euros... Most of all Marantz & Philips, but not only... There is no Chief Omelet without breaking many eggs
... that's why tubes are not for me 
Some observations along the way - the 100n caps C13-16 and C21-24 are a tight fit using the MKP2D031001F00JO00 - they barely fit in together - but they will fit. Also a tight fit here are the C10-12 which nestle in between the film caps. Again, they'll fit, but they are tight. I note the mouser heatsink at the bottom has a support leg which I just snipped off. Because of the way that support leg is attached to the heatsink, the heatsink fins can't be faced away from the regs, ie as per the silkscreen. Again, no biggie, just something to note is all.
I had 2k7 instead of the amended R15 2k5, but that worked perfectly fine. Also I forgot to order 50k trimpots but adjusted down a 100k before installing, and of course that worked fine too. I haven't loaded it, or looked at noise etc, but all looks OK.
I had 2k7 instead of the amended R15 2k5, but that worked perfectly fine. Also I forgot to order 50k trimpots but adjusted down a 100k before installing, and of course that worked fine too. I haven't loaded it, or looked at noise etc, but all looks OK.
Hi,
congrats. Acording your regs before you perhaps don't need all the local decouplingcaps but the 100 nF. If so you can use the areas that were for the lythics for more headroom legs spacing.
It's advised to leave free the pots circuitry if you don't have a scope to trim correctly. It is said somehere in the thread. So don't populate it
cheers.
congrats. Acording your regs before you perhaps don't need all the local decouplingcaps but the 100 nF. If so you can use the areas that were for the lythics for more headroom legs spacing.
It's advised to leave free the pots circuitry if you don't have a scope to trim correctly. It is said somehere in the thread. So don't populate it
cheers.
Reclocking:
I think I saw somewhere in the thread that there was no benefit in reclocking the lines before this DAC. This is somewhat at odds with personal experience with other DACs (more conventional ones).
So, Miro/previous builders - have you used reclocking for this DAC and what is the experience you had?
Fran
I think I saw somewhere in the thread that there was no benefit in reclocking the lines before this DAC. This is somewhat at odds with personal experience with other DACs (more conventional ones).
So, Miro/previous builders - have you used reclocking for this DAC and what is the experience you had?
Fran
Thanks Fran for your observations
I tested the reclocking ... it has zero benefit while a good asynchronous xmos board is used (like jlsounds). For a simple PCM2706 it may have a potential benefit, but I personally didn't notice the difference there.
The benefit is certain for Delta-Sigma converters (sourced from bad crystals and synchronous USB transfer).
I tested the reclocking ... it has zero benefit while a good asynchronous xmos board is used (like jlsounds). For a simple PCM2706 it may have a potential benefit, but I personally didn't notice the difference there.
The benefit is certain for Delta-Sigma converters (sourced from bad crystals and synchronous USB transfer).
dixit: "...maybe more people would notice something "
Yes... indeed (+1 for Xmos, and I stream from Linux/Unix'OS).... frankly : digital front-end is a very important part of a DAC... too much underated*, the habits are to talk just about the dac chip... just a small, though still important, part of a DAC device... I'm maybe biased in this 32 bits world !
You can totally change the sound of a DAC by just playing on the power supply on a frontend and of course by the layout and parts of the frontend itself... sometimes just with a cap**, a little passive part at iso layout ! Go figure why ?
*: that's a complex thing and simple worlds like : FiFo, powersupply, etc, can not give you a result, seen as single solution. It's a sum of many things : layout, passive parts, active parts, the sensibility of the sum on the i/o side, powersupply typologie and passive parts that play with it, and the happy marriage with your whole hifi chain to find the individual listening tasting you like and adapt adds happy trade-offs... so a complex whole thing but you can give up and save monney (time) with...
**I can see before they come the attacks of the fanatics of the theory book coming... but try!
***:***( Xmos chips hardwares and soft are a little still above for now and most of the time)
"Yes
... indeed (+1 for Xmos, and I stream from Linux/Unix'OS).... frankly : digital front-end is a very important part of a DAC... too much underated*, the habits are to talk just about the dac chip... just a small, though still important, part of a DAC device... I'm maybe biased in this 32 bits world !You can totally change the sound of a DAC by just playing on the power supply on a frontend and of course by the layout and parts of the frontend itself... sometimes just with a cap**, a little passive part at iso layout ! Go figure why ?
*: that's a complex thing and simple worlds like : FiFo, powersupply, etc, can not give you a result, seen as single solution. It's a sum of many things : layout, passive parts, active parts, the sensibility of the sum on the i/o side, powersupply typologie and passive parts that play with it, and the happy marriage with your whole hifi chain to find the individual listening tasting you like and adapt adds happy trade-offs... so a complex whole thing but you can give up and save monney (time) with...
**I can see before they come the attacks of the fanatics of the theory book coming... but try!
***:***( Xmos chips hardwares and soft are a little still above for now and most of the time)
Last edited: