BFT ALPHA BOM → SCA output → Finder 40.52 relay → Shelly 1 Mini Gen4 SW input
Adds a gate is fully closed signal to Home Assistant without touching the BFT's safety/limit-switch circuit and without bringing 230 V into the controller's low-voltage (SELV) terminal strip.
| Fact | Source |
|---|---|
BFT SCA output (terminals 15‑16, DIP7 = OFF) is a gate-open warning light output, rated 24 V~, 3 W max. It is active while the gate is open/opening and inactive when the board sees the gate as fully closed. |
ALPHA BOM manual D811430 v06 §4, §6 (DIP7), §3 |
BFT closing limit switch is SW.C, an N.C. contact between terminals 7 (input common) and 12. LED DL6 goes off when it operates. Terminals 13/14 = 0 V / 24 V~ accessory supply (0.2 A max). |
ibid. §4, §5, Fig. 2 |
Shelly 1 Mini Gen4 SW is a mains-live-referenced input — "Connect the switch to the SW terminal of the Device and the Live wire." Supply is 110‑240 V~ only, no DC option. |
Shelly 1 Mini Gen4 printed user guide, step 3 |
Because SW senses mains potential and the BFT command circuit is 24 V SELV, the two
must never touch. The Finder relay provides the barrier: its coil sits in the BFT's
24 V world, its contacts switch the Shelly's Live. The 40.52 gives 6 kV / 8 mm
insulation between coil and contacts, and the 95.05 socket puts the coil terminals
(A1/A2) at the opposite end of the socket from the contact terminals — so the
physical separation is built in.
| Item | Part | Notes |
|---|---|---|
| Relay | Finder 40.52.8.024.0000 | 2 CO (DPDT), 24 V AC coil, 1.2 VA / 45 mA @ 50 Hz, coil range 19.2–26.4 V~, contacts 8 A / 250 V AC, coil↔contact insulation 6 kV / 8 mm |
| Socket | Finder 95.05 | Screw (box clamp) terminals, 35 mm DIN rail or panel mount, 250 V / 10 A. Fits 40.51 / 40.52 / 40.61 |
| Optional indicator | Finder 99.02.0.024.59 | LED module, 6–24 V DC/AC — plugs into the 95.05, gives a visual "SCA active" light. Very useful for commissioning (see §4 Test 1 first) |
| Enclosure | Small IP54/IP65 DIN box | Recommended — see §3 |
The relay coil must be AC (
.8.in the part code). A DC-coil 40.52 (.7.or.9.) will buzz and chatter on the BFT's 24 V~ supply.
Coil load is 1.2 VA, comfortably inside SCA's 3 W budget and the 0.2 A accessory limit.
Mount the relay socket next to the Shelly, ideally both in their own small enclosure, so that:
SW) stays inside that box;7‑8
(your START command) plus the new 24 V coil pair.This is what keeps you compliant with the BFT manual's §4 requirement that wires at different voltages be physically separated or given ≥1 mm extra insulation, and clamped near the terminals.
If you must mount the relay inside the BFT enclosure, keep the mains link in its own sleeving, route it well away from JP2/JP3 and the antenna wiring, and cable-tie everything down.
Tools: DMM with AC volts + continuity/resistance, insulated screwdriver, ferrule crimper, torch. Have the gate accessible so you can move it between positions.
Board powered, DMM on V~, probes on terminals 13 and 14.
A1/A2 after wiring (§7). If it still exceeds 26.4 V under load, do not leave it;
see §9.Watch LED DL6 (SWC) while moving the gate:
This is your ground truth for the rest of the tests.
Isolate 230 V at the breaker, set DIP7 = OFF, re-power.
This disables the 2nd radio channel. Confirm you are not using remote button 2 for anything (lights, a second gate, garage door) before committing.
Board powered, DMM on V~, probes on terminals 15 and 16.
| Gate position | Reading | |
|---|---|---|
| Fully closed (DL6 off) | record → | |
| Open / stopped mid-travel | record → |
A high-impedance DMM can show a few volts of phantom voltage across an open contact. Treat anything below ~5 V as "off". The 1.2 VA coil will settle it definitively once connected.
Isolate 230 V at the breaker and confirm dead with the meter before probing. DMM on continuity / Ω:
| Measure | Reading |
|---|---|
| 15 ↔ 13 | |
| 15 ↔ 14 | |
| 15 ↔ 7 | |
| 16 ↔ 13 | |
| 16 ↔ 14 | |
| 16 ↔ 7 |
You cannot check the switching of 15‑16 with the board off — the internal relay drops out when unpowered, so it will always sit in its rest state.
BFT ALPHA BOM (JP2) Finder 95.05 socket + 40.52.8.024.0000
┌───────────────┐ ┌────────────────────────────────────────┐
│ 15 SCA ●───┼──── 0.75 mm² ──────┤ A1 ┐ │
│ 16 SCA ●───┼──── 0.75 mm² ──────┤ A2 ┘ coil 24 V AC, 1.2 VA │
└───────────────┘ │ │
SELV only leaves the box │ 11 COM ◄──── L (brown, 1.0 mm²) ─────┼─── same Live
│ 12 NC ─────► Shelly SW ────────────┼─── that feeds
│ 14 NO unused │ Shelly "L"
│ 21 / 22 / 24 unused │
└────────────────────────────────────────┘
Coil polarity is irrelevant — it's an AC coil, 15→A1 / 16→A2 and
15→A2 / 16→A1 are equivalent.
Resulting logic: gate open → SCA on → coil energised → NC (11‑12) opens →
SW dead. Gate closed → SCA off → coil released → NC closed → SW live.
So Shelly input ON = gate fully closed.
Put the coil in series with the SCA contact, across the 24 V~ rail:
BFT ALPHA BOM (JP2) Finder 95.05 socket + 40.52.8.024.0000
┌───────────────┐ ┌────────────────────────────────────────┐
│ 14 24 V~ ●───┼──── 0.75 mm² ──────┤ A1 ┐ │
│ │ ┌─────┤ A2 ┘ coil 24 V AC, 1.2 VA │
│ 15 SCA ●───┼──── 0.75 mm² ┘ │ │
│ 16 SCA ●───┐ │ 11 COM ◄──── L (brown, 1.0 mm²) ─────┼─── same Live
│ 13 0 V ●───┘ link, 0.75 mm² │ 12 NC ─────► Shelly SW ────────────┼─── that feeds
└───────────────┘ │ 14 NO unused │ Shelly "L"
│ 21 / 22 / 24 unused │
└────────────────────────────────────────┘
Current path: 14 (24 V~) → A1 → coil → A2 → 15 → [SCA contact] → 16 → 13 (0 V)
The 16 → 13 link is a short jumper inside the BFT enclosure. Which of 15/16 you
feed does not matter — 14 → coil → 16 with 15 → 13 is equally valid.
Same resulting logic as Variant A: Shelly input ON = gate fully closed.
A1/A2) 16 → 13 jumper: same 0.75 mm² stranded with ferrules.11, socket 12 → Shelly SW) SW input draws well under 1 mA, so 1.0 mm² is chosen for mechanical strength and
terminal fit, not current. Use 1.5 mm² if you want it to match the rest of the
mains wiring in the box or if local practice requires it.| Terminal | Accepts |
|---|---|
| Finder 95.05 | solid 1×6 mm² / 2×2.5 mm²; flexible 1×4 mm² / 2×2.5 mm² |
| Shelly 1 Mini Gen4 | 0.5–4 mm² solid, stranded or bootlace ferrules; strip 6–7 mm; torque 0.4 Nm |
| BFT ALPHA JP2 | not stated in the manual — assume 1.5 mm² max. Verify visually; 0.5–0.75 mm² fits comfortably |
Cable-tie all new wires to an existing bundle or a strain-relief point near the terminals, as BFT §4 requires.
Isolate 230 V, wire per §5 or §6, plug the relay into the socket, fit the retaining clip, then:
11, 12, the L wire) and the SELV side (socket A1, A2, and
BFT terminals 7, 12, 13, 14, 15, 16). This should read fully open — it is your check
that nothing is mis-terminated.SW live).A1/A2 with the gate open: expect 20–26 V~. This confirms
the coil is properly driven and lets you check the rail under load (Test 1).If the input flickers during the closing phase, some BFT firmware flashes SCA while
closing — add a for: 00:00:02 delay on the HA side rather than chasing it in hardware.
In the Shelly web UI (or app):
With type = Switch + Detached, the Home Assistant Shelly integration exposes the input
as a binary_sensor (a Button-type input would give you an event entity instead,
which is not what you want here).
If the 24 V rail measured above 26.4 V under load in step 3, options are: add a series resistor sized against the 45 mA coil current, use a relay with a wider coil range, or drop the SCA approach and drive the sensing from a 230 V-side arrangement. Do not simply leave the coil overvoltaged.
Rename the entity to something readable, e.g. binary_sensor.gate_closed — where
on = fully closed.
A minimal template cover, given that a single START input can only ever be "pulsed" (4-step logic cycles open → stop → close → stop):
# legacy template cover syntax — check the docs for your HA version
cover:
- platform: template
covers:
driveway_gate:
friendly_name: "Driveway gate"
value_template: "{{ is_state('binary_sensor.gate_closed', 'off') }}"
open_cover:
- action: switch.turn_on
target: { entity_id: switch.gate_start }
close_cover:
- action: switch.turn_on
target: { entity_id: switch.gate_start }
stop_cover:
- action: switch.turn_on
target: { entity_id: switch.gate_start }
value_template returning true means open, so it is the inverse of the closed
sensor. All three commands issue the same START pulse — that is inherent to driving the
gate from one input, not a limitation of this wiring.
automation:
- alias: "Gate closed sensor looks stuck"
triggers:
- trigger: state
entity_id: switch.gate_start
to: "on"
conditions: []
actions:
- delay: "00:00:15"
- condition: state
entity_id: binary_sensor.gate_closed
state: "on"
- action: notify.persistent_notification
data:
message: >
Gate START was pulsed 15 s ago but the closed sensor never cleared.
Check the 24 V coil pair to the Finder relay.
This arrangement is fail-safe against faults in the mains sensing path
(L → 11 → 12 → SW): a broken wire or failed contact leaves SW dead, which reads as
not closed — conservative, and you notice it because the sensor never reports closed.
It is not fail-safe against faults in the 24 V coil pair. Because SCA is inactive when the gate is closed, "closed" is the relay's quiescent state — so a broken coil wire, a dead coil, or DIP7 being flipped back to ON all leave the relay released and the sensor stuck reporting closed even with the gate wide open. Do not build "arm the alarm / lock up / it's safe to leave" logic on this signal without the watchdog in §10.
If you need positive-energy confirmation of the closed state, the alternative is to drive the coil from a spare N.O. contact on the closing limit microswitch instead of from SCA — there, closed means coil energised, so any fault in the coil circuit reads as not closed. That requires the limit microswitch to have an unused contact set and shares its common with BFT terminal 7, so it needs its own verification pass.
7‑12 limit switch loop, and never connect
the Shelly SW input directly to any JP2/JP3 terminal.