Will a higher-power repeater or a higher-gain antenna fix our coverage problem?

Usually neither, and for different reasons. Doubling repeater transmit power buys you 3 dB in one direction only - the direction that is almost never the problem - while a single 200 mm reinforced-concrete wall costs 15 to 25 dB in both. A higher-gain omni antenna is at least reciprocal, so it helps transmit and receive equally, but it produces that gain by flattening the vertical radiation pattern, which is the opposite of what a multi-storey building needs. In most buildings where staff can hear the repeater but cannot be heard, the constraint is the uplink and the receive path, and adding downlink power makes the symptom worse rather than better.

Why more transmit power only fixes half the link

A repeater link has two independent budgets. Downlink is the repeater talking to the portable; uplink is the portable talking back. They are not symmetrical to begin with. A 40 W repeater sits at +46 dBm; a 4 W portable sits at +36 dBm. That is 10 dB in the repeater's favour before anything else, and once you add roughly 3 dB of body loss for a radio worn at the hip and the fact that a stubby helical antenna typically presents somewhere between -3 and 0 dBi against a 2 dBi reference omni, the uplink is 13 to 16 dB behind the downlink on the same physical path.

Raising repeater power widens that gap. Going from 25 W to 100 W is +6 dB - the user now hears the repeater in places where the repeater still cannot hear them. That is not extra coverage; it is a larger zone of one-way coverage, which is more dangerous operationally than a plain dead zone because staff believe they are in contact. This is the failure mode covered in detail in our article on why staff can hear the repeater but the repeater cannot hear them.

The numbers that matter when someone proposes more power:

• Doubling transmit power is +3 dB. Quadrupling it is +6 dB.

• At an indoor path-loss exponent of 3.0 to 3.5, +6 dB buys roughly 1.5 times the distance, not double.

• One 200 mm reinforced-concrete wall costs 15 to 25 dB. A reinforced-concrete lift or stair shaft costs 25 to 35 dB.

• Design threshold for reliable DMR voice is -95 dBm at the portable, with 10 dB fade margin for commercial systems and 15 dB for mission-critical.

• So quadrupling repeater power does not buy you one internal concrete wall, in one direction, at four times the amplifier cost.

What a higher-gain antenna actually does to a high-rise

Antenna gain is passive and reciprocal. Unlike transmit power, it applies to both paths, so a 9 dBi collinear does give you 9 dB on downlink and 9 dB on uplink. That makes it a better instinct than a bigger amplifier. The catch is where the gain comes from: an omni antenna has no power source, so gain is produced entirely by compressing the vertical beamwidth and pushing energy toward the horizon. A 2 dBi omni radiates a broad pattern; a 9 dBi collinear may have a vertical beamwidth of under 10 degrees.

In a 30-storey building with a rooftop antenna, that narrow pattern points at the horizon and away from the floors directly below. Raising the gain measurably reduces signal on the floors you were trying to cover. High-gain omnis earn their place on flat, wide sites - a large single-storey warehouse, a port yard, a spread-out campus. They are the wrong instrument for vertical coverage, and swapping a 2 dBi antenna for a 9 dBi one on a tower block regularly makes the complaint list longer.

There is a second constraint specific to Singapore. IMDA licences are written in terms of effective radiated power, which is transmit power plus antenna gain minus feeder loss. Adding antenna gain raises ERP just as surely as adding amplifier watts. Licence-free and pre-set localised allocations sit around 1 W ERP, and larger private land mobile network licences run up to about 25 W ERP - figures to confirm against IMDA's current licence conditions at application, not to design against from memory. A higher-gain antenna can put an otherwise-compliant system over its licensed ERP without anyone touching the repeater.

How to tell in ten minutes whether you are uplink-limited

Walk to the worst location with two people and two radios. Have the person at the repeater or console transmit. If the field user hears them cleanly but their reply is broken or does not open the repeater at all, the system is uplink-limited, and no amount of downlink power or repeater output will help. If both directions are equally poor, the path itself is the problem and the fix is distance and obstruction, not gain at either end.

Two things are worth ruling out before blaming the building: a CTCSS tone or DMR colour code mismatch produces exactly the same symptom as an uplink failure, and an elevated receive noise floor from LED drivers or variable-frequency drives will desense the repeater receiver by 10 to 20 dB while every transmitter in the system tests perfectly.

What actually fixes it

Three approaches, in ascending order of cost. First, fix the receive side: a low-noise preamplifier and a properly tuned duplexer recover uplink sensitivity, and a better feeder run - 1/2 inch plenum coax at roughly 4.5 dB per 100 m, or 7/8 inch at 2.8 dB per 100 m - reduces loss on both paths at once. Second, use passive reciprocal gain where the geometry suits it. Third, and the only approach that reliably works through concrete, move the antennas to the users. A distributed antenna system removes path loss from both budgets simultaneously, because a portable 15 m from a ceiling antenna on its own floor never has to punch through the slab at all. We compare the receive-side fix against a full DAS in receiver multicoupler and preamplifier versus a distributed antenna system.

Related questions

Does a higher-gain antenna on the portable radio help? Yes, and more than people expect, because it improves the weak end of the link. A longer whip in place of a stubby helical can recover 3 to 5 dB on both paths, though it is often rejected on ergonomic grounds by staff who carry the radio all shift.

Will a bi-directional amplifier solve an uplink problem? A BDA amplifies both directions, so it addresses the uplink where a repeater power increase does not. But it also amplifies the noise floor, requires 15 to 20 dB more donor-to-service antenna isolation than its gain setting, and needs an IMDA licence and type approval.

If power does not fix coverage, why specify a 40 W repeater at all? Because the repeater has to drive a distribution network with real losses - splitters, couplers, and long riser runs can consume 20 to 30 dB before the signal reaches an antenna port. The power is spent on the DAS, not on punching through walls.

How Suneast approaches this

We measure both budgets before quoting anything. A site visit records downlink level and uplink talk-back at the same test points, so the limiting path is identified from data rather than assumed, and the receive noise floor is measured rather than inferred. Where the uplink is the constraint, we quote the receive-side work and the antenna distribution that fixes it, and we state the ERP the design will present for the IMDA licence application. Where a client has already bought a higher-power repeater that did not help, we can usually reuse it - the equipment is rarely wrong, but the end of the link it was aimed at usually is.