Beacon block board: power-rails

Boards / beacon nets cut at the block boundary by tools/blockgen.py; bench header pins are the only addition

The beacon's power input and rails on their own board: the two ideal diodes that join main 5 V and the lighting battery's 5 V, the two 3.3 V regulators and the Q43's load switch, for bench proof of the hand-over before the beacon is built.

KiCad 3D render of the Beacon block board: power-rails, top side. — Top

KiCad 3D render of the Beacon block board: power-rails, bottom side. — Bottom

Revision: A
Sheets: 1
Parts: 27
Board: 42 × 45 mm
Copper layers: 2
Footprints: 31
Location: bench
Bus: none: the beacon's 5V_IN and LIGHTS connectors on one side, an 11-pin header for the rails and the ESP32 signals
Status: Rev A draft: routed, DRC clean (2026-10-10), not ordered

What it does

A bench board carrying the beacon’s power and rails blocks: the two LM66100 ideal diodes that OR the main 5 V and the lights node’s +5V_LIGHT into the always-on rail, the AP2112K that makes 3.3 V for the ESP32 and the radios, the second AP2112K that makes the flight-controller-side 3.3 V from the FC port’s 5 V, and the TPS22917 load switch through which the ESP32 can drop the Q43 to save the lighting battery in lost mode. The thing to see on the bench is the hand-over: pull main 5 V and the always-on rail must not dip; restore it and nothing back-feeds the lights node.

Like the other block boards it is cut from the beacon’s design data by a tool and checked net by net against the beacon, so it is the beacon’s own power path on a board of its own.

Where it sits

On the bench. The beacon’s 5V_IN (PH-2) and LIGHTS (GH-4) connectors are on the board, so a bench supply and a lights node (or a second supply standing in for it) plug in as they would on the aircraft; the rails and the ESP32’s control lines come to an 11-pin 2.54 mm header. 42 x 45 mm, two layers, every part on the top face.

Main parts

Part	Function
2 x LM66100	Ideal-diode OR of main 5 V and +5V_LIGHT into +5V_AO, the always-on rail
AP2112K-3.3 (x2)	3.3 V for the always-on side; 3.3 V for the FC side from the FC port’s 5 V
TPS22917	Load switch for the Q43’s 5 V, on by default, dropped by the ESP32 in lost mode
J1, J2	The beacon’s 5V_IN and LIGHTS connectors, same parts and pin order

Connections

Pin	Net	Stands in for, on the beacon
JH1-1	GND	ground
JH1-2	+3V3	the always-on 3.3 V rail, out to the bench (the ESP32 and radios’ supply)
JH1-3	+3V3_FC	the FC-side 3.3 V, out (the isolators’ FC side)
JH1-4	+5V_AO	the always-on 5 V rail, out (the piezo)
JH1-5	+5V_FC	the FC port’s 5 V, in from the bench (what GPS1 or TELEM2 would bring)
JH1-6	+5V_MAIN	the main 5 V as sensed by the opto
JH1-7	Q43_5V	the switched 5 V for the Q43, out
JH1-8	VBUS	USB 5 V, in, for the USB-powered bench case
JH1-9, 10	ESP_LIGHTS_RX, ESP_LIGHTS_TX	the lights-node UART, through to the header
JH1-11	Q43_EN	the ESP32’s load-switch control, in

Status

Rev A draft, 2026-10-10 (EDA 8a289ae): generated by tools/blockgen.py beacon power+rails, 17 nets, ERC clean apart from the same regulator-output tie the beacon’s own ERC waives, routed to 0 open, DRC 0 errors and 0 warnings, every net a subset of the beacon’s. Not ordered. Bench test: the hand-over between main 5 V and +5V_LIGHT with a scope on +5V_AO, the Q43 switch under the ESP32’s control, and the load each regulator carries against its rating.

Schematic and layout

Files

From the EDA repository at revision b201257 (2026-10-12). Open the files in KiCad 10 or later.