Ok, so the resistors need to be of greater tolerance but can you hear the difference if lower tolerance resistors are used?
I recently bought FIIO K11 R2R DAC, as I was curious about R2R, which according to an Internet search uses 0.1% resistors (not sure that is accurate) and I cannot hear any difference between a modified Musical Fidelity M1 DAC (which itself was compared to the M6 model, unmodified M1 and Rega DACs).... and I'm struggling to hear any difference between any of them.
To test, I have a circuit that takes spdif input from a CD player (electrical or optical) with 2 electrical spdif output so they exact same signal goes to 2 DACs, the output of which goes to a preamp, a remote control unit selects either input.
I recently bought FIIO K11 R2R DAC, as I was curious about R2R, which according to an Internet search uses 0.1% resistors (not sure that is accurate) and I cannot hear any difference between a modified Musical Fidelity M1 DAC (which itself was compared to the M6 model, unmodified M1 and Rega DACs).... and I'm struggling to hear any difference between any of them.
To test, I have a circuit that takes spdif input from a CD player (electrical or optical) with 2 electrical spdif output so they exact same signal goes to 2 DACs, the output of which goes to a preamp, a remote control unit selects either input.
FIIO K11 R2R is not a full fledges r2r dac. It uses fpga to deal with non linearities. You cannot make a proper discrete 24bit r2r dac with resistors alone.
You can certainly measure higher distortion on 16-bit R-2R DACs with poor resistor matching. I suspect listening tests are going to be down to personal preference. Personally I prefer single-bit designs with a really good low jitter clocks I find they just sound smoother and more musical. I love the Pink Triangle DaCapo DAC, which, despite it's age and actually mediocre technical specs compared to more modern DACs, still sounds awesome to me!
Then again, I really like the sound of the Naim CDS-3 and that's an R-2R design using the PCM1704 DAC chips from Burr Brown...
Then again, I really like the sound of the Naim CDS-3 and that's an R-2R design using the PCM1704 DAC chips from Burr Brown...
Interesting! Do you know how they use the FPGA to correct? I can see how it would be possible to change timings with the FPGA, but how do they correct for resistor values? Higher but depth then only use the last few bits for corrections based on post assembly calibration? Or is there something else clever going on?