docs: the sink's width as shipped — 32-bit words, the level applied in 32 bits
docs: 32-bit words from the decoder to the DAC
docs: the register of DAC boards; dac and output as they act
docs(audio-out): the GY-PCM5102 module proven; seven boards
Signed-off-by: foonerd <157698061+foonerd@users.noreply.github.com>
docs(audio-out): a PCM5102A module proven on the three lines; six boards
Signed-off-by: foonerd <157698061+foonerd@users.noreply.github.com>
docs(audio-out): PCM5122 boards with their own oscillators; the DAC Mini proven
Signed-off-by: foonerd <157698061+foonerd@users.noreply.github.com>
docs(audio-out): the WM8960 Audio HAT — the option, the bring-up, codecs written with the clocks stopped
docs(audio-out): the bench row names the three boards proven
Signed-off-by: foonerd <157698061+foonerd@users.noreply.github.com>
docs(audio-out): 3.3 V for header pins 1 and 17, with the LDO stage drawn
Signed-off-by: foonerd <157698061+foonerd@users.noreply.github.com>
docs(audio-out): the PCM512x driver, 64 fs, the bus survey; three boards proven on the bench
Signed-off-by: foonerd <157698061+foonerd@users.noreply.github.com>
docs: wiring a Pi-HAT DAC to the panel; Boards, the targets by role; pages follow the firmware
Audio out gains the bench form of the DAC connection: a figure and a
table of the three I2S wires, ground and the two I2C lines from the
HAT's 40-pin header to J8, J6 and J5, with the power rule (the HAT on
its own supply; J8 pin 2 is USB VBUS without a diode) and what the
firmware logs at boot. Every pin reference on the page now matches the
firmware that plays: BCLK GPIO 2, LRCLK GPIO 3, data GPIO 11, GPIO 12
spare, where the page had proposed GPIO 3, 12 and 11 before a wire was
connected. The standing table, the firmware section and the bench list
say what is proven (tone, FLAC at 44.1 and 96 kHz, MP3, the sink's
volume) and what is not (tier 1 on J5, 32-bit slots, I2S slave).
Boards is new: the ESP32 touch panels recorded as targets, read from
the makers' pages on 19 October 2026, one table per role, with what a
board must have to be a target and how to read a question mark.
Hardware and Home follow: the stereo route is built and plays, the
firmware crate exists, and the companion's crates come by tag where one
exists.
Signed-off-by: foonerd <157698061+foonerd@users.noreply.github.com>
docs(audio-out): powering a DAC board from the panel: the inline 5 V filter, with schematic
docs(audio): DAC2 HD parts and addresses from the kernel overlay; sources say which page gave what; pin 36's two jobs
docs(audio): corrections from review: HiFiBerry rows, tiers, expander lines, Pi header rails, TAD5242 headphones, WM8804 outputs, cable note, ES9020 status; J8 5 V settled on paper
docs(audio): Audio out as an options register against the panel's constraints; Hardware and Home follow
No part chosen. Pi DAC HATs, bare I2S modules, TI (PCM5102A, PCM5122,
TAD5242/5212), ESS (9023P, 9018K2M, 9028Q2M, 9038Q2M, 9039Q2M), a speaker
amplifier (MAX98357A), digital out (WM8804 class) and Hymnal, each with
clock, control, volume, output, tier and the maker's lifecycle word as
read today. Carrier board shaped as a Pi header plus a module header on
the same three lines; the PCM5102A kept as the worked example of a
soldered part. Bench list extended for tier-1 boards and the slave role.