roderickvd
100+ Head-Fier
I owned a Soekris dam1121 before with a self-built linear power supplies. Digital power from quad LT3045's, analog from a AMB σ22. Both from their own toroidal transformers. It was fed I2S from a Raspbery Pi 3B+, galvanically isolated with an IanCanada IsolatorPi II and reclocked by a Allo Kali 22/24 MHz. I have tried all sorts of filters both stock and custom filter packs.
It's good to remember that the Soekris units are mostly built around this same dam1121 module. What differs between the dacxxxx ranges is the number of modules (ladders), power supply topology, buffer stage and input/output connections. Of course the PSU and buffer stages as well as general hardware layout will impact the final result, but at its core they are all built around this same DAC module. I think that my PSU's and output buffer (none) are amongst the best that can be had.
I now own a Ares II.
Objectively I think both the Soekris units and the Ares II have measured very well on the likes of ASR and SBAF.
Technically - the level of implementation being equal - I think the Ares II leads by a few points.
Most importantly, it has dedicated oscillators whereas the Soekris has a single programmable clock. By definition a programmable clock will never reach the same level of performance as a dedicated clock (again, assuming both are of good build quality) and indeed it cannot match the jitter figures. Worse, the Soekris uses a PLL-like algorithm to track the input clock and program its oscillator accordingly. This algorithm has been shown on diyAudio to be unstable (and as control systems theory also dictates, a follower will never be as good as the incoming clock) causing the programmable clock to shift continuously. In short, the Ares clock is much better, offering lower jitter and perhaps more importantly: clock stability.
Second: the Ares PSU is linear from an oversized toroid. The latest Soekris incarnations are SMPS. In the end this may down boil down to preference (I for sure know that Hypex makes SMPS's with jaw-dropping measurements and the dac2541 is excellent too) but again, build quality being the same I think that linear supplies have the edge here.
Third: Coupling capacitors in the Ares are much higher grade Nichicon's (from the top of my head) where on the Soekris they are SMD-type.
Subjectively, I have never liked the stock Soekris filters. I don't find them engaging or dynamic. This became especially apparent for me when I switch to custom NOS filters (quasi-NOS, like in the Ares). These do make the Soekris come to life. To me, the Ares' stock filters are like the best custom Soekris filters. Out-of-the-box they are very good. I am still on the fence whether I like OS slow or NOS better.
Circling back, remember that all Soekris DAC's revolve around the same module. I think the Ares II will be better than the dac2541 because it has several technical advantages. I would say both are engineered equally well (not talking about the aesthetics) and so that technical advantage can and does shine. Certainly the clocking!
It's good to remember that the Soekris units are mostly built around this same dam1121 module. What differs between the dacxxxx ranges is the number of modules (ladders), power supply topology, buffer stage and input/output connections. Of course the PSU and buffer stages as well as general hardware layout will impact the final result, but at its core they are all built around this same DAC module. I think that my PSU's and output buffer (none) are amongst the best that can be had.
I now own a Ares II.
Objectively I think both the Soekris units and the Ares II have measured very well on the likes of ASR and SBAF.
Technically - the level of implementation being equal - I think the Ares II leads by a few points.
Most importantly, it has dedicated oscillators whereas the Soekris has a single programmable clock. By definition a programmable clock will never reach the same level of performance as a dedicated clock (again, assuming both are of good build quality) and indeed it cannot match the jitter figures. Worse, the Soekris uses a PLL-like algorithm to track the input clock and program its oscillator accordingly. This algorithm has been shown on diyAudio to be unstable (and as control systems theory also dictates, a follower will never be as good as the incoming clock) causing the programmable clock to shift continuously. In short, the Ares clock is much better, offering lower jitter and perhaps more importantly: clock stability.
Second: the Ares PSU is linear from an oversized toroid. The latest Soekris incarnations are SMPS. In the end this may down boil down to preference (I for sure know that Hypex makes SMPS's with jaw-dropping measurements and the dac2541 is excellent too) but again, build quality being the same I think that linear supplies have the edge here.
Third: Coupling capacitors in the Ares are much higher grade Nichicon's (from the top of my head) where on the Soekris they are SMD-type.
Subjectively, I have never liked the stock Soekris filters. I don't find them engaging or dynamic. This became especially apparent for me when I switch to custom NOS filters (quasi-NOS, like in the Ares). These do make the Soekris come to life. To me, the Ares' stock filters are like the best custom Soekris filters. Out-of-the-box they are very good. I am still on the fence whether I like OS slow or NOS better.
Circling back, remember that all Soekris DAC's revolve around the same module. I think the Ares II will be better than the dac2541 because it has several technical advantages. I would say both are engineered equally well (not talking about the aesthetics) and so that technical advantage can and does shine. Certainly the clocking!
Last edited: