Audio Tick Output

A per-pulse audible click through an I2S amplifier (e.g. MAX98357A) on supported boards. A short boot chime plays once when the device starts so you know the audio path is alive.

ESP32 only. AudioTick uses the ESP-IDF I2S driver, which is not available on the ESP8266. See “Audible feedback on ESP8266” below for the alternative.

Shipped builds

Env family	Hardware	I2S pins
xh_s3e_*	MINI ESP32-S3-N16R8 (“XH-S3E-AI”) with on-board speaker	Fixed at BCLK=15, WS=16, DOUT=7
esp32_audio_*	Generic ESP32 + MAX98357A wired by you	Configurable in Config > Audio Tick (defaults BCLK=26, WS=25, DOUT=22)
esp32s3_audio_*	Generic ESP32-S3 + MAX98357A wired by you	Configurable in Config > Audio Tick (defaults BCLK=15, WS=16, DOUT=7)

On the generic families, verify your wiring matches the pins in Config before enabling. The XH-S3E envs have the pins locked to match the on-board speaker.

Audible feedback on ESP8266

For per-pulse audio on an ESP8266 use Pulse Out: single GPIO into a piezo, or a small speaker via a MOSFET, or an attenuator into a powered speaker. No I2S amplifier required.

Settings (tick pref group)

Pref	Range	Default	Notes
enable	0 / 1	0	Reboot to apply after enabling.
volume	0-100	60	0 mutes; click amplitude scales linearly.
freq	300-6000 Hz	900	Click centre frequency.
decay	2-100 ms	4	Decay time constant of the click body.
engine	0 / 1	0	0 = Pool, 1 = Chirp.
chime	0-8	1	0 = silent boot, 1 = random pick, 2-8 = specific.

Throttle

Audio is rate-limited to 20 clicks/second. Above that the audio thins out while the counter still tracks every pulse.

Voice announcer

Audio-capable builds can speak the current reading aloud at a configurable interval, using a 48-word ADPCM vocabulary (Cori, en_GB) baked into flash. Each announcement reads CPM as a natural English number (“two hundred and forty”), optionally followed by the dose (“twelve point three micro sievert per hour”, switching to milli sievert above 1000 µSv/h). When CPM is above the warning or alert threshold the relevant state word prefixes the announcement (“alert. two hundred and forty CPM.”). When RadMon uploading transitions state, the announcer can also say “RadMon online” or “RadMon offline” on the next interval boundary.

The announcer obeys the audio quiet-hours window: nothing is spoken between tick.aq_from and tick.aq_to.

Settings (voice pref group)

Pref	Range	Default	Notes
vol	0-100	80	Voice volume. Independent of the click volume.
announce_enable	0 / 1	0	Master switch for periodic announcements.
announce_interval	10-3600 s	60	Seconds between announcements.
announce_usv	0 / 1	0	Append dose in µSv/h or mSv/h.
announce_radmon	0 / 1	0	Speak RadMon online/offline transitions.

Test routes:

• /say?w=alert plays a single vocabulary word
• /say?w=two,hundred,and,forty plays a comma-separated sequence
• /announce triggers the current announcement immediately

Alert klaxon

Five klaxon patterns are synthesised on-device and fire automatically when CPM crosses the alert threshold (rising edge only). A configurable cooldown prevents repeated triggering during a sustained alert, and the audio quiet-hours window suppresses the klaxon at night.

Pattern	Name	Sound
0	Alarm	Deep alternating high/low tones, reactor-style
1	Two-Tone	Police two-tone wail
2	Wail	Slow rising-and-falling air raid siren
3	Horn	Naval AOOGA horn
4	Voice	Cori says “alert” three times

Settings (klaxon pref group)

Pref	Range	Default	Notes
klax_enable	0 / 1	0	Master switch.
klax_type	0-4	0	Pattern (see table above).
klax_vol	0-100	55	Klaxon volume.
klax_cd	1-60 min	5	Cooldown after a klaxon fires.
aq_from, aq_to	HH:MM	(off)	Audio quiet-hours window.

Test routes: /klaxon?t=0 through /klaxon?t=4 play the chosen pattern immediately, bypassing the cooldown but still respecting quiet hours.

See also

• NeoPixel
• Outputs index