Finished both channels of my PCM63 DAC and the digital front end power supply board.
The PCM63 is the Holy Grail of vintage Burr-Brown R2R chips. Each board uses two of them in parallel, the two boards will stack on top of each other, similar to my PCM56 DAC. The output of each board are the current output pins of the parallel chips. I will again be using a Sowter step up transformer for the I/V stage.
I am again planning a tube output stage, but a different flavor this time around. I have a healthy stock of NOS E182CC tubes, a very good sounding audio dual triode. Each section has a relatively high gm of 15mA/V. I'd like to use a pair of E182CC on the output with the sections wired in parallel in a transformer coupled cathode follower output buffer with a 1:1 output transformer, something like this.
Being completely transformer coupled and with a triode well suited for buffer service, I am expecting very good results with this setup. However, since a cathode follower has no voltage gain, the Sowter step up transformers will be responsible for converting the current output of the DAC to a line level output voltage.
Without getting too heavy into the technical details, these current output DACs perform best if presented with a low impedance load. Generally, less than 100ohms is best for PCM63, otherwise it will start generating audible harmonic distortion. To reach 2Vrms output with a 1:10 step up and parallel PCM63, roughly an 8.2K resistor will be needed on the secondary of the step up transformer, presenting the DAC chips with an ~82ohm primary impedance.
This should work well in theory, but I plan to measure the distortion on the secondary of the step up transformers with a dummy load before committing to this I/V stage. If the measurements look good, then I'll order parts for the output buffer and design the chassis. This time around, I plan to include some tube aesthetics with the E182CC buffer tubes sticking partway out of the top plate, should look pretty cool.
I have next week off before starting my new job, going to test this and also devote some time to the Aegis project.
The PCM63 is the Holy Grail of vintage Burr-Brown R2R chips. Each board uses two of them in parallel, the two boards will stack on top of each other, similar to my PCM56 DAC. The output of each board are the current output pins of the parallel chips. I will again be using a Sowter step up transformer for the I/V stage.
I am again planning a tube output stage, but a different flavor this time around. I have a healthy stock of NOS E182CC tubes, a very good sounding audio dual triode. Each section has a relatively high gm of 15mA/V. I'd like to use a pair of E182CC on the output with the sections wired in parallel in a transformer coupled cathode follower output buffer with a 1:1 output transformer, something like this.
Being completely transformer coupled and with a triode well suited for buffer service, I am expecting very good results with this setup. However, since a cathode follower has no voltage gain, the Sowter step up transformers will be responsible for converting the current output of the DAC to a line level output voltage.
Without getting too heavy into the technical details, these current output DACs perform best if presented with a low impedance load. Generally, less than 100ohms is best for PCM63, otherwise it will start generating audible harmonic distortion. To reach 2Vrms output with a 1:10 step up and parallel PCM63, roughly an 8.2K resistor will be needed on the secondary of the step up transformer, presenting the DAC chips with an ~82ohm primary impedance.
This should work well in theory, but I plan to measure the distortion on the secondary of the step up transformers with a dummy load before committing to this I/V stage. If the measurements look good, then I'll order parts for the output buffer and design the chassis. This time around, I plan to include some tube aesthetics with the E182CC buffer tubes sticking partway out of the top plate, should look pretty cool.
I have next week off before starting my new job, going to test this and also devote some time to the Aegis project.
Last edited: