C-Bus network limits and bridges: how big can one network be?

This question comes up on nearly every medium-to-large job we quote: “How many switches and dimmers can we hang off one C-Bus network before it falls over?” The short answer is that a single C-Bus network has three hard limits working together — cable length, unit loads, and supply current — and the moment any one of them runs out, you split the system and join it with a Network Bridge. Get this wrong at design time and you end up chasing flaky comms, dropped messages and reset units six months after handover. Get it right and the network just works for the next twenty years.

Here’s how the numbers stack up, and how we plan them on real installs around Melbourne.

The three limits of a single C-Bus network

Every standalone C-Bus network is bound by the same set of figures. The diagram above lays them out, but in plain terms:

• Maximum total cable length: 1000 m. That’s the total length of pink Cat5 C-Bus cable across the whole network, not the longest single run. Daisy-chain it, star it, branch it — add it all up and you must stay under 1000 m.
• Maximum 100 unit loads. Each device you connect (a relay, a dimmer, a Saturn switch, a network interface) presents a “unit load” to the network. The published rule of thumb is 100, but read the next section — the real ceiling is current, not a head-count.
• Maximum 2 A of C-Bus current. This is the one that actually matters. The 100-unit-load figure is just shorthand for the current those units draw.

Why 2 A is the real limit

C-Bus is a low-voltage SELV system running at a nominal 36 V DC across the pink cable, carrying both data and the power that runs the input units (switches, sensors, PIRs). Output units — relays and dimmers like the L5504D2U — are mains-powered for their switching side, so they draw very little from the bus itself, but they still present a unit load for their comms electronics.

Every device sips a small amount of current from the network. A “unit load” is roughly defined so that 100 of them add up to the 2 A the network can sustain. Input units with backlit displays, like eDLT or Saturn Zen switches, draw more than a simple relay channel, so a network heavy on fancy switch plates hits the current wall well before it hits 100 physical devices. That’s why we always tally the actual unit-load value from each device’s datasheet rather than just counting boxes.

When you exceed 2 A, you don’t get a tidy error — you get voltage drop along the cable, units browning out at the far end, and intermittent comms that are miserable to diagnose. Schneider Electric publishes the unit-load figures for each catalogue item; it’s worth pulling them when you spec a large board. There’s good background reading over on the Clipsal site if you want the manufacturer’s own figures.

System power supplies: feeding the 2 A

The current the network can draw has to come from somewhere, and that’s the job of the system power supply. A single 5500PS supplies around 350 mA to the bus. Because one network can use up to 2 A, you can run multiple supplies in parallel on the same network to share the load — but the total of all supplies must not exceed 2000 mA.

So in practice:

• One 5500PS (350 mA) is fine for small networks.
• Larger networks need several supplies distributed along the cable run, not all bunched at one end.
• The sum of every supply on the network stays under 2000 mA — overshoot that and you risk overstressing the bus.

Where you place the supply matters

Don’t park all your power supply capacity at one end of a long run. C-Bus current flows out along the cable, and resistance in the cable causes voltage to drop the further you get from the supply. Spread your supplies through the network — roughly toward the centre of the load — so no device sits at the end of a long, lightly-fed run. On big jobs we map the cable topology first, then drop supplies in at the points that keep the whole network’s voltage comfortably in range. This is the single biggest reliability factor on large C-Bus systems, and it’s pure design-time planning. More on cabling practice in our C-Bus network articles.

When you’ve maxed out: the Network Bridge

Hit any of the three limits — 1000 m, 100 unit loads, or 2 A — and the answer is the same: split the install into two networks and join them with a Network Bridge (5500NB). The bridge sits between two physically separate C-Bus networks, each with its own power supply and its own 2 A / 1000 m / 100-unit budget.

Two bridged networks gives you up to 2000 m of total cable and double the unit-load capacity. You can chain further networks for very large sites — think multi-storey homes, big estates, or commercial fit-outs — with bridges defining the boundaries between them.

1. Decide the split logically. We usually divide by floor, by wing, or by switchboard. Keep devices that talk to each other constantly (a switch and the dimmer it controls) on the same network where possible to minimise cross-bridge traffic.
2. Give each network its own power supply budget. Each side is a complete network — its own 5500PS (or multiples up to 2000 mA), its own cable-length tally, its own unit-load count.
3. Install the 5500NB between them. The bridge has a connection to each network and is configured in C-Bus Toolkit to pass only the messages that need to cross.
4. Configure which groups pass across. This is the part people get wrong — see below.

Group addresses are local — mind the clash

This is the trap with bridged systems. Group Addresses on the Lighting application (56) run 0–255 and are local to each network. Network A can have Group 10 and Network B can have a completely different Group 10, and they won’t interfere — until you ask the bridge to pass Group 10 between them, at which point they collide.

The Network Bridge controls exactly which group addresses are passed from one network to the other. So your job at programming time is to:

• Allocate group numbers carefully so that any group which needs to travel across the bridge is unique on both sides.
• Only pass the groups that genuinely need to cross — a hallway scene that spans two floors, an “all off” trigger, a master switch. Don’t flood the bridge with traffic that stays local.
• Keep a clear addressing plan documented, because the next person to touch the system (possibly us, in five years) needs to know what crosses where.

Plan it on paper before you pull cable

The recurring theme here is design-time planning. Nine times out of ten when a customer rings us about an unreliable C-Bus system that “worked fine when it was smaller”, we find it’s grown past one of these limits without anyone re-checking the budget — too much cable, too many backlit switches, or all the power supply capacity dumped at one end of the board.

Before the first metre of pink cable goes in, we tally the expected unit loads, total the cable length, work out how many supplies we need and where they sit, and decide whether the job is one network or several bridged together. It’s far cheaper to draw a second network on the plan than to retrofit a bridge into a finished, plastered house.

That’s the lot — the 1000 m / 100-unit / 2 A limits, the 2000 mA supply ceiling, and the Network Bridge as your way past all of it. Plan the power and the addressing properly and a big C-Bus system is every bit as solid as a small one.

If you’re scoping a large C-Bus install around Melbourne — or you’ve inherited one that’s grown unreliable and you suspect it’s overloaded — get in touch with us via our contact page and we’ll help you map the networks out properly. — Adam and the DUKE team.