[IDEA] Upgrading my DAC Hat

[hifinet]

Assembled DAC with CRC 10uF film caps (which, along with the module, physically dominates the available space) and Texas Components TX2575 naked bulk metal foil I/V resistors. The CRC (Component Research Corp) caps are metallized polycarbonate (similar audio properties to polypropylene), and are used pulls from military equipment. In my opinion, the audio qualities of PC and PP tend to parallel their optical properties. Polycarbonate is optically clearer than polypropylene, and is used in eyeglasses. I do not know of a current manufacturer of polycarbonate capacitors. WIMA made the MKC series polycarbonate caps. The 10uF at 40VAC is very compact 11x21x31.5mm (1.240" L x 0.433" W). No personal experience. PC was replaced by PPS. This dielectric apparently sounds excellent, and the capacitors are small in size. Both PC and PPS are hard to find. Polypropylene will be the easiest film caps to find.

An Allo Kali reclocker will reduce jitter and improve sound quality. Power applied to the Kali is providing power to the Raspberry Pi and DAC HAT.

Let the DAC "burn-in" for at least 60 hours before doing critical listening tests. The Nichicon UKA caps need a good 60 hours of burn-in time. 

A couple of notes. The output polarity of this DAC is inverted. Use polarity inversion in Moode. Configure...Audio...MPD Options...DSP options...Polarity inversion ON. This inverts polarity for both channels.

This DAC has no output filter. This creates a sinc droop or 2.5dB roll off at 20kHz. Also, ultrasonic images (for 44.1kHz sample rate the images are >24kHz) pass unimpeded. This could potentially cause IM distortion in the audio band with some audio equipment. On my system, I hear no problems in this regard. Analog filters on the output can cause significant problems with music. From this Burr-Brown data sheet "...the phase response of an analog filter with these amplitude characteristics [i.e. higher order (9-13 pole) analog filter] is nonlinear and can disturb the pulse-shaped characteristic transients contained in music." Your ears (or tweeters) will filter images > 24kHz.

If you think that you are getting IM distortion due to unfiltered ultrasonic images, you can try upsampling with SoX. Upsampling to 384 kHz will push ultrasonic images past 192 kHz. Images at this frequency would be unlikely to cause IM distortion in the audio band.  It will also correct sinc droop. In my experience, SoX upsampling sounds worse than the native bitrate with this DAC. Upsampling with PGGB and millions of taps does a much better job (request a free trial).