Skeptischism
Active Member
- Joined
- Sep 6, 2019
- Messages
- 229
- Likes
- 124
But thats what you did... you compared to desktop R2R dacs and now you complain when I bring out the reality of such a comparison? the MSB is ridiculously bad; yet another with horrible spurious tones making ingress into the dac analogue output from the power supply. They may actually be the worst performing hiend dacs i've ever seen. Even stereophile had a difficult time defending the performance.
I find the whole idea of building discrete R2R puzzling ... and having made a decision to design another one, why bother doing it if you are only going to use such small SMD resistors? given these are your main source of noise and distortion and 0805 and 1206 size are significantly better than 0603 or 0402 in this regard.. the problems you face doing it discrete are related to the matching of the discrete resistor elements and their TEMPCO. have a guess what the best past, present and future way of combating this is? laser trimming on chip resistors in an IC.
One could posit that taking the difficult proposition of making a discrete R2R portable and getting about the performance one might expect from that decision (terrible) wasnt the best use of resources ...
It looks like and measures as one would expect such a design would. Audio Dacs were designed decades ago, to decode the digital output from a CD. for this amount of engineering to fall short of being able to resolve CD quality 40 years later at a cost of 3000USD, isnt what I would call a great achievement. that it is similar to other failed attempts at this design problem is unsurprising.
I find the whole idea of building discrete R2R puzzling ... and having made a decision to design another one, why bother doing it if you are only going to use such small SMD resistors? given these are your main source of noise and distortion and 0805 and 1206 size are significantly better than 0603 or 0402 in this regard.. the problems you face doing it discrete are related to the matching of the discrete resistor elements and their TEMPCO. have a guess what the best past, present and future way of combating this is? laser trimming on chip resistors in an IC.
One could posit that taking the difficult proposition of making a discrete R2R portable and getting about the performance one might expect from that decision (terrible) wasnt the best use of resources ...
It looks like and measures as one would expect such a design would. Audio Dacs were designed decades ago, to decode the digital output from a CD. for this amount of engineering to fall short of being able to resolve CD quality 40 years later at a cost of 3000USD, isnt what I would call a great achievement. that it is similar to other failed attempts at this design problem is unsurprising.
Last edited: