hi Magnus, no i have not tried different power source(s) yet. Did try 2 different modes (though one is explicitly suggested not to be used - but i don’t know how it really works):
- power by SMPS adapter that comes with S2 to MicroUSB and computer connected via USB cable …
- powered by USB only by the same computer (nothing connected to Micro-USB) …
1st config -you are supposed NOT to use but not sure where it draws power from in that case - but it worked just fine and i heard no differences between the two scenarios.
Possibly with better power supply it might sound better but sounds silly to spend 100EUR+ to power a device that costs 350EUR… and though i have Intona USB isolating device - based on everything i’ve read it wouldnt really make any difference as S2 has excellent USB isolation built-in so didn’t even bother (left Intnona in front of Benchmark when used as a DAC). But if anyone confirms that iFi power which is 60-70 EUR i think makes a noticable difference ill go for it !
In the meantime got squeezelite to upsample (client for LMS) to 700khz+ for all content … funny what little sbc can do (Roon does it on a server so endpoints really need no computing power) … not that i can hear any difference compared to standard squeezelite @ 384/352khz) but it was fun tweak to try that too.
Funny part with Roon when upsampling to DSD512 is that it uses tons of CPU power - i do have xeon server (ranging from 2.1GHz to 3.1GHz depending on the load/cores turbo) … and for 44.1khz content it does 2x speed processing but with some DSD64 or HD (96/192khz) content it goes only 1.5-2x with multicore options selected - otherwise its even too slow)… either that technology needs serious horsepower or algorithms could still be improved a lot … i see only 2-3 cores sweating and all others not used).