Do we remove the 8n2's on the FIRDAC with this filter ?
Are you still using this ? I've been trying it with wima polypropylene filter caps ,10uf BG NX coupling . I left the 8n2 on the dac though so presume they need removing
Guys it's important the LPF keep the characteristic impedance of source to avoid reflections and have more acourate sound. Isn't a joke, I learned first doing S/PDIF & balanced digital cables so I used for my noDac also with great success. If source have 376 ohms you must have to terminate the LPF with a resistor of the same value....
For what it's worth, I have tried to design my filters to be as good a virtual ground as possible to the DAC cores while still providing some passive filtering before the first active circuitry, so that most of the high-frequency current doesn't flow through active circuitry.
The lower the filter input impedance, the better the cancellation between the signal-dependence of the currents drawn by the positive and negative DAC sides should work. The SMD filter capacitor right at the DAC board ensures that the impedance is low at very high frequencies, the shunt feedback makes it low at very low frequencies, in between there is an impedance bump because I don't know how to make it filter otherwise.
I don't know how much performance degradation you get with higher load impedances, so please feel free to experiment with other approaches.
The lower the filter input impedance, the better the cancellation between the signal-dependence of the currents drawn by the positive and negative DAC sides should work. The SMD filter capacitor right at the DAC board ensures that the impedance is low at very high frequencies, the shunt feedback makes it low at very low frequencies, in between there is an impedance bump because I don't know how to make it filter otherwise.
I don't know how much performance degradation you get with higher load impedances, so please feel free to experiment with other approaches.
I'm just curious about the filter you designed could you post the general schematic please with red circle for filters, TIA.
See DAC3filter_6.pdf attached to post #1931, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7483396
I later started working on a filter with more passive filtering, but never really finished it. That is, I made a prototype with one working channel on perfboard, but never made a PCB for it. See posts #3028, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7672339 , #3138, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7681475 , and #3304, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7692816 The document attached to post #3028 also relates to the original filter (sections 1...5 apply to the original filter, section 6 to the filter with more passive filtering).
By the way, bohrok2610 did build and measure the DAC with a filter based on post #3028 but implemented with an OPA1632, see post #3100, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7680453 He didn't see much (or even any?) improvement.
I later started working on a filter with more passive filtering, but never really finished it. That is, I made a prototype with one working channel on perfboard, but never made a PCB for it. See posts #3028, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7672339 , #3138, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7681475 , and #3304, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7692816 The document attached to post #3028 also relates to the original filter (sections 1...5 apply to the original filter, section 6 to the filter with more passive filtering).
By the way, bohrok2610 did build and measure the DAC with a filter based on post #3028 but implemented with an OPA1632, see post #3100, https://www.diyaudio.com/community/threads/return-to-zero-shift-register-firdac.379406/post-7680453 He didn't see much (or even any?) improvement.