Do two-way radios work inside lift cars, and what does it take to cover them?

Usually not, unless the lift was deliberately designed into the radio system. A lift car is a steel box inside a reinforced-concrete shaft, and the two together add roughly 40 to 60 dB of loss on top of whatever the signal has already lost getting through the building. A portable sitting at a comfortable -70 dBm in the lift lobby will typically read somewhere between -110 and -130 dBm once the doors close, which is at or below the radio's own noise floor. Covering the car means putting an antenna or a radiating cable inside the shaft — nothing you do at the repeater will fix it.

Why the lift car is worse than the rest of the building

Two barriers stack. The shaft itself is a reinforced-concrete tube, and in our surveys an RC shaft or core wall costs 25 to 35 dB at 400 to 470 MHz — considerably more than the 15 to 25 dB a 200 mm floor slab costs, because the wall is thicker and the rebar density around a lift core is higher. The car adds its own shell on top of that. A steel car with the doors closed behaves as a poor Faraday cage: not perfect, because the door gap, the sill and the ventilation openings above the car all leak, but good for another 15 to 25 dB.

The numbers worth holding on to:

• RC lift shaft or core wall: 25-35 dB at UHF

• Steel lift car shell, doors closed: 15-25 dB additional

• Reinforced-concrete floor slab, for comparison: 15-25 dB

• Design threshold for reliable DMR voice: -95 dBm at the portable

• Fade margin: 10 dB standard commercial, 15 dB for mission-critical areas

Put those together and a lift car needs 40 to 60 dB more delivered signal than the corridor outside it. That is not a margin you recover by fitting a longer whip or specifying a bigger repeater — 40 dB is a factor of ten thousand in power, and a repeater cannot be made ten thousand times larger.

The uplink fails first, and that is what makes it dangerous

A lift car is the clearest case of the asymmetry we see in most buildings. A 40 W repeater transmits at +46 dBm; a portable transmits 4 W at +36 dBm, less about 3 dB of body loss when it is on someone's hip. That is a 13 dB imbalance in the downlink's favour before anything else is considered.

The practical result is a lift car where a technician can still faintly hear the control room but cannot be heard back. Everyone reports it as "patchy," nobody reports it as a fault, and it only becomes a real problem when someone is stuck in a lift and needs to call out — which is precisely the scenario the coverage was bought for. Any acceptance test that measures only downlink signal will pass a lift that cannot actually be used. This is the same failure pattern described in our article on why the repeater can hear you outside but not inside.

What it actually takes to cover a lift

Three approaches, in ascending cost.

Cover the lobbies and exclude the cars. Most systems do this by default, whether or not anyone said so. An antenna in each lift lobby gives staff coverage while waiting and at the doors, and the car itself is a known gap. This is a legitimate design choice provided it is written down. It is not legitimate to leave it unstated and let the client discover it at handover.

A shaft antenna, top and bottom. A lift shaft behaves as an oversized waveguide: once energy is launched into it, it propagates along the shaft with far less loss than it would through the building. A directional or panel antenna (typically 7 dBi) mounted at the top of the shaft firing downward, and a second at the bottom firing upward, will usually serve a shaft of 15 to 20 storeys with a single feed at each end. Taller shafts want an intermediate antenna every 10 to 15 floors. The antenna covers the shaft; the last 15 to 25 dB into the car comes through the door gap and the vents, which is why the shaft-side level has to be generous rather than marginal.

A radiating (leaky) cable down the shaft. This gives the most uniform level along the full travel and is the right answer for a very tall shaft, a firefighting lift, or anywhere the coverage has to be even rather than adequate. It costs more, needs a proper cable route in the shaft, and requires coordination with the lift contractor over clearances and the trailing-cable route.

Whichever approach is chosen, shaft work has to be scheduled with the lift maintenance contractor and done under a lift-isolation permit. On a live building this is normally the longest-lead item in the whole installation, not the RF work.

Related questions

Does Singapore's fire code require radio coverage inside lift cars?

No Singapore code prescribes it the way the US International Fire Code Section 510 does. The IFC treats lift lobbies as critical areas requiring 99% coverage at -95 dBm and requires lifts to be measured at the primary recall floor — note that this is a measurement at a landing, not inside a moving car. In Singapore, coverage under COPIF and the Fire Code does not set a lift-car figure, so if you want the cars covered it has to be written into the tender specification.

Can one antenna cover a whole lift shaft?

For a shaft up to about 15 to 20 storeys, one antenna at each end is normally enough because of the waveguide effect. Beyond that, plan on an intermediate antenna roughly every 10 to 15 floors and verify with a measured drop test rather than a prediction.

Does the lift car need its own antenna inside it?

Rarely for two-way radio. The 15 to 25 dB through the doors and vents is recoverable with enough level in the shaft. An antenna inside the car requires running RF down the trailing cable and is normally only done where the lift is safety-critical and the shaft solution has been measured and found short.

Suneast designs lift coverage as an explicit line item rather than an assumption. We state in the proposal whether cars are in scope or excluded, we compute the uplink budget as well as the downlink so the car is tested in both directions, and we test with the doors closed and the car moving rather than parked at a landing. Where a shaft antenna is the right answer we size it from the measured shaft loss, not a rule of thumb — the same discipline described in our articles on antenna spacing per floor and signal loss through concrete slabs and basement walls.