dfrobot Gravity: Digital Vibration Sensor

AI context

Dense context for pasting into a model prompt — addresses, pinout, voltages, tolerances and the vendor's reference circuit.

BOARD: DFRobot Gravity: Digital Vibration Sensor, SKU DFR0027, 22 x 30 mm, two 3 mm mounting holes on a 5 mm grid, immersion-gold finish, one 3-pin Gravity connector (J1). The board is a single vibration switch plus three passives - no microcontroller, no I2C device, nothing to address. The product page describes it as detecting vibration, shaking, knocking and impact.
ADDRESSES: none. There is no bus device on the board, so a bus scan finds nothing and `addresses` is empty by measurement, not by omission.
PINOUT: three pads on the Gravity connector, numbered 1, 2, 3 as J1's own pads in DFRobot's schematic (footprint S3B_PH): pad 1 = +5 (supply in), pad 2 = D (digital signal, the switch node through the 470 ohm series resistor R2), pad 3 = GND. The pads are numbered by the schematic connector because the board prints no pad numbers and the wiki publishes no pin table.
SUPPLY: DFRobot rates the module 3.3 V to 5 V. The schematic labels the rail +5; the output is a pull-up to that rail, so at a 3.3 V supply the idle high level is 3.3 V and at 5 V it is 5 V - the pad swings with the supply rather than being regulated.
LOGIC LEVEL: R1 10 kOhm (marked 10K) pulls the switch node to the supply rail, so an open switch reads high; when the switch closes, the node is pulled to ground through the switch and the pad reads low through the 470 ohm series resistor. DFRobot publishes no absolute-maximum voltage for the signal pad and no input-current figure, so the safe rule is not to drive the pad above the supply rail and not to load it below about 10 kOhm.
REFERENCE CIRCUIT, as built: R1 10K from +5 to the switch node; K1 SW-18015P from that node to GND; R2 470 ohm in series from the node to J1 pad 2; C1 104P (0.1 uF) across +5 and GND as the only decoupling. The schematic's bill of materials is exactly five parts - C1, J1, K1, R1, R2 - so pull-up, series resistor and decoupling are all fitted and nothing needs adding.
TRAPS: the sensing element is a spring-contact vibration switch, and the board debounces nothing, so one mechanical event produces a burst of edges - debounce in firmware rather than trusting a single edge. Two of DFRobot's own figures are odd as printed and are quoted rather than corrected: "Switch life up to 10 million seconds" is a duration where a cycle count belongs, and "Open circuit resistance 10Mohm" is the open-contact resistance. The wiki blurb also says the sensor has interfaces labelled "A" for analog and "D" for digital, but both the schematic and the product description carry one digital output only - there is no analog pad on this board.