Do satellite or voting receivers make sense as a cheaper alternative to a full in-building DAS?

Sometimes, but far less often than the cost comparison suggests. A voting system adds receivers, so it only improves the uplink — the path from the portable back to the repeater. It does nothing at all for the downlink. In a building whose problem is genuinely one-directional, that is a legitimate and cheaper fix. In a reinforced-concrete basement, lift lobby or service core, where attenuation is symmetric and both directions fail together, adding receivers buys you a system in which the control room hears the guard perfectly and the guard still hears nothing back.

What a voting receiver actually does

A satellite or voting receiver is a second (third, fourth) receiver tuned to the repeater's input frequency, sited away from the main repeater, usually inside or near the weak area. When a portable transmits, both the main repeater receiver and the satellite receiver hear it. A comparator — the voter — compares the signals and passes whichever is cleanest into the repeater, which then retransmits on the output frequency from the original site. The link from satellite receiver back to the voter is either an RF link on a separate frequency pair or, on modern DMR systems, an IP connection over the building's Ethernet.

The architecture is genuinely useful, and it is well documented. The problem is where that documentation comes from. Almost everything written about receiver voting is wide-area amateur, GMRS and public-safety material — hilltop repeaters, campus-sized sites, multi-square-mile coverage. In-building coverage gets a passing mention at best, and the asymmetry caveat below is almost never stated.

Why voting looks like the right answer for a building

Because the complaint that triggers it is real and correctly diagnosed. Uplink is normally the limiting path in a two-way radio system, and the arithmetic is not subtle:

• Repeater transmitting at 40 W: +46 dBm

• Portable transmitting at 4 W: +36 dBm, less 3 dB body loss at the hip

• Net asymmetry in the repeater's favour: about 13 dB

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

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

• Reinforced-concrete floor slab: 15–25 dB per slab; RC core or shaft: 25–35 dB

So there is a band roughly 13 dB wide in which the portable hears the repeater fine and the repeater cannot hear the portable at all. That is the classic "they can hear us, we can't hear them" fault, and I have written separately about why the repeater cannot hear portables inside a building. If a survey shows the downlink comfortably above −95 dBm everywhere and the uplink failing, a satellite receiver placed in the problem zone removes that 13 dB of asymmetry plus whatever structure sits between the zone and the rooftop. It is the correct fix.

Where voting stops working

That 13 dB window is the entire scope of the remedy. Concrete does not care which direction a signal is travelling. A basement carpark two slabs and a transfer plate below the lobby is down 40 dB or more in both directions. If the portable reads −105 dBm on the downlink there, no receiver you add anywhere improves that number by a single dB — the guard's radio still cannot decode the repeater, and on DMR it will not degrade gracefully either, it will simply stop.

This is the specific thing the available material does not say. Ask which of a receiver-side fix or a DAS is right for a building and the answer you get back describes the receiver-side option accurately and in detail, and frequently does not mention distributed antenna systems at all. The same pattern shows up in the multicoupler and preamplifier question: a confident, well-sourced description of one option, and silence about the other.

What it actually costs against a DAS

The cost case is weaker than it looks once the full parts count is in. Each satellite receiver needs its own receive antenna sited inside the problem zone, a receiver, front-end filtering, a link path back to the voter, and a share of a comparator. The link is either a second IMDA frequency assignment and a link radio at each end, or an IP drop. Note what has already happened: to place that receive antenna you have run cable into the problem zone. Having done so, a passive DAS branch to the same location costs a directional coupler and a length of 1/2" plenum coax — 4.5 dB per 100 m at 450 MHz — and solves both directions. Two or three problem zones is usually the crossover. Past that, the voting system is both more expensive and half a solution.

Does a satellite receiver need its own IMDA licence in Singapore?

Each fixed receive station is radio apparatus and belongs on the licence. An RF linkback also needs its own assigned frequency pair, which means a further assignment and station licence entry on top of the repeater's — verify the current position with IMDA before budgeting it. An IP linkback avoids the extra spectrum entirely and is the sensible default in a building that already has structured cabling.

The decision rule

Survey both directions before choosing. Record signal level at the worst point on each floor, uplink and downlink, against a −95 dBm threshold with 10 dB of margin. If downlink passes everywhere with margin and only uplink fails, voting is a real candidate and may save money. If downlink fails anywhere you need coverage, the building needs distributed antennas, and receivers are the wrong end of the problem. A downlink-only survey cannot answer this question, and neither can a datasheet.

Related questions

Does receiver voting help a lift car? No. A steel car in an RC shaft is 25–35 dB down in both directions. Lift coverage needs a radiating cable or shaft antenna on the distribution system.

Can you combine voting with a DAS? Yes, and on large mixed sites it is the right answer — the DAS handles the shielded interior, voting handles detached outbuildings or yard areas too far for a cable run.

Is voting worth it for a construction site? Often, because the geography is horizontal and outdoor rather than vertical and shielded. That is the case voting was designed for.

How Suneast approaches this

We survey uplink and downlink separately before recommending an architecture, because the two answers frequently differ and only one of them is visible on a standard coverage plot. Where a building is genuinely uplink-limited in a small number of zones, we will propose voting receivers and say so. Where it is not, we design a distributed antenna system to a −95 dBm threshold with 10 dB of fade margin and 95% area coverage, and we handle the IMDA station and network licensing for whichever architecture is specified.