Power on Stage: Circuits, Not Speakers

Everything else in this track is about signal. This lesson is about what arrives before any signal does — through a hole in the wall, from a breaker you cannot see, on a circuit you are almost certainly sharing with something you have not thought about. It is the least glamorous subject in live sound and the only one that can end a set in the middle of a chorus.

The arithmetic is small and you do it once. What makes it worth doing is that the backstage rules of thumb are mostly wrong: it is not fine because it is only a few things, and you cannot just run another strip off that one. The gap between those sentences and the truth is invisible right up until the room goes dark.

A Socket Is a Budget, and the Wall Does Not Tell You Its Size

The number of outlets on a wall tells you nothing about how much power is behind them. Six sockets down one side of a stage are frequently six holes into a single circuit, protected by a single breaker, sharing a single total.

In North America a general-purpose circuit runs at 120 volts and is protected at 15 or 20 amperes. Volts times amps gives you the ceiling: 1,800 watts on a 15A circuit, 2,400 on a 20A one. But a breaker is not meant to be held at its rating indefinitely. Anything drawing for three hours or more counts as a continuous load, and the wiring rules require the protective device to be sized at 125% of that load — which, read from the other end, means you should plan on 80% of the breaker's rating. A gig is a continuous load by that definition, so the honest working numbers are 1,440 watts on a 15A circuit and 1,920 watts on a 20A one.

In the UK and most of Europe the shape is different and the total is larger. Mains is 230 volts, and every plug carries its own fuse rated from 1 to 13 amperes, so the most any single plug will pass is 13 × 230 = 2,990 watts, near enough 3 kilowatts. Behind that, sockets are usually wired as a ring final circuit protected at 32A, or a radial at 20 or 32A. The trap is assuming the 32 is yours: that figure protects the cable running round the building, not the socket in front of you, and the fuse in your plug still caps you at 13A no matter how generous the ring is. Bigger stages skip the domestic system entirely and provide 16A or 32A industrial connectors, which is a different conversation and a good sign.

Whichever side of the Atlantic you are on, the question to ask the venue is not "is there power over there" but "is that the same circuit as this" — and that question belongs in the advance, alongside everything else in the advancing lesson.

Your PA Draws Far Less Than the Carton Claims

Here is the number that surprises people. A powered speaker sold as a 2,000-watt box is not a 2,000-watt load on your circuit, and it is not close.

Take a current, widely used example: QSC's K12.2 is a class-D powered twelve, rated at 1,800 watts peak for the woofer and 225 for the horn. Its published mains draw at one-eighth power — the standard test condition, chosen because it approximates real programme material rather than a sine wave held at full tilt — is 1.9 amperes at 120 volts, or 1.1 amperes at 240. On standby it draws 0.2A at any voltage.

Two of those and a subwoofer are not going to trouble a 15A circuit. The reason is partly that class-D amplification is efficient, and partly the more fundamental one the headroom lesson sets out: music has a high crest factor, so peaks are brief and the average is far below them. The amplifier only draws what the music asks for, and the music mostly asks for very little.

Which leads to the useful inversion. The thing that trips the breaker at a small show is usually not the PA. Four 500-watt tungsten lighting fixtures are 2,000 watts of continuous, honest, heat-producing load, and they will exceed a 15A circuit on their own while the whole sound system sits comfortably inside it. Lighting, hazers, and anything with a heating element are the real consumers. Sound is the cheap tenant.

The Other Things on Your Circuit

The circuit feeding the stage in a small room very often continues into the rest of the building — the bar, the kitchen, the cellar. Two consequences follow.

The first is that a motor starting up is a much larger event than a motor running. A fridge compressor, a walk-in cooler, an air-conditioning unit: a conventional induction motor draws roughly six to ten times its running current in the instant before it comes up to speed, because a stationary rotor generates no back-EMF to oppose the supply. That surge lasts a fraction of a second, which is why the breaker often survives it, and why it can still be enough to push an already-loaded circuit over the edge at exactly the wrong moment.

The second is subtler and does not trip anything. That same surge drags the supply voltage down briefly, and everything on the circuit sees the dip. Your amplifiers' supply rails sag with it. On a rig running with plenty of margin you will never notice. On one running near its limit, that is the click, the thump, or the momentary loss of level that seems to happen at random and in fact happens every time the cooler cycles.

If you have a choice, put audio on a circuit of its own and let the lighting and the bar have theirs. If you do not have a choice, at least know which one you are on.

Daisy-Chaining: the Listing Says No, and the Listing Is the Rule

A power strip — a relocatable power tap, in the language of the safety standards — is listed for one arrangement: plugged directly into a permanently installed wall receptacle. Not into an extension cord, and not into another power strip. UL's product listing for these devices spells that out, along with a 20A total limit and the statement that they are not a substitute for fixed wiring.

That is not merely advisory. Occupational safety rules require listed equipment to be installed and used in accordance with the instructions in its listing, so using a strip outside those conditions is a violation of the rules themselves, not just of good practice.

The mechanism is worth understanding rather than memorising. Each strip is built and fused for the outlets on its own face. Chain three together and the first one in the line — the one actually connected to the wall — is now carrying everything downstream of it, through a cord and a set of contacts sized for a fraction of that. The failure is not dramatic and it does not announce itself. Connections warm up, the warmth increases resistance, the increased resistance produces more warmth, and the part of the chain that fails is the part nobody is looking at, under a riser, behind a curtain, in the dark.

Two strips into two different wall sockets is fine, and is the answer nearly every time. What you actually need on stage is a proper multi-outlet distribution unit with a cable rated for the whole load, plugged into one socket — which is what a rented distro is, and why they exist.

Long Cables Cost You Volts

A cable is a resistance, and current through a resistance produces a voltage drop, so the far end of a long run sits at a lower voltage than the socket. The wiring codes recommend keeping that drop to about 3% on a branch circuit, and it is easy to blow past that with the cords bands actually own.

Run 15 amperes down 30 metres (100 ft) of 12 AWG and you lose about 4.8 volts, which is 4% of 120 — already over the guideline. The same run in 14 AWG loses about 7.6 volts, or 6.3%. Thinner cord, longer run, or both, and it gets worse quickly. In the UK the higher supply voltage helps, because the same watts arrive at roughly half the current, and the drop scales with current.

The practical version: use the thickest cable you own for the longest run, uncoil it fully rather than leaving it on the drum, and treat a warm cord or a warm connector as information rather than a curiosity. Voltage drop is also cumulative, so a long cord feeding a strip feeding another cord is doing this twice.

Switch-On Order, and Why Amplifiers Are Rude at the Moment You Turn Them On

An amplifier with a large conventional transformer draws an enormous current in the first few milliseconds of being switched on — the transformer's core has not yet established its magnetic flux and the reservoir capacitors are empty, so for an instant the supply is looking at something very close to a short circuit. For a toroidal transformer above about 300VA, that inrush can reach the order of a hundred amperes for a few milliseconds. It is over before a breaker's thermal element can react, which is why it usually passes; large racks include soft-start circuitry precisely because it does not always.

Two habits follow from this, and they are the same two habits that protect the PA from a bang through the speakers:

On: everything upstream first — desk, sources, processing — then the amplifiers or powered speakers last, and one at a time rather than all at once from a single switch.

Off: exactly the reverse. Amplifiers and powered boxes off first, then work back up the chain. Anything that thumps as it powers down should not be connected to something capable of turning that thump into 126dB.

GFCI and RCD: the Device That Trips for a Reason

North American venues use ground-fault circuit interrupters; UK and European ones use residual current devices, typically rated at 30 milliamperes. Both do the same job by the same method — comparing the current going out with the current coming back and disconnecting when the two disagree, because a difference means current is returning by some path that is not the cable, and the most consequential such path is a person.

These trip on stages more than anywhere else, and usually not because anything is faulty. Every modern switch-mode power supply leaks a small, entirely legitimate current to earth through its interference-suppression filtering — commonly 0.5 to 3 milliamperes each. Individually trivial; add up a console, a lighting rig, several powered speakers, a couple of amplifier racks and the wireless system, and the standing leakage on one protective device climbs towards its threshold. The guidance is to keep the total under about a third of the device's rating — under 10mA on a 30mA unit — after which a perfectly healthy rig sits close enough to the edge that a damp night or one more thing plugged in will tip it over.

The fix is to spread the load across more than one protected circuit, which is another reason to ask which socket is on which breaker. The fix is never to defeat the protection. This is the same boundary the cables lesson draws around the ground-lift switch: a DI's lift disconnects a signal ground and is yours to use, while anything that disconnects a protective earth or bypasses a residual current device is not, no matter how effectively it silences a hum. The device that keeps tripping is doing the one job you would want it to do if the fault were real.

What This Means for You

• Find out which sockets share a breaker before you plug anything in. Six outlets frequently mean one circuit, and the wall gives you no way to tell by looking.
• Budget 1,440 watts on a North American 15A circuit and 1,920 on a 20A one. The full 1,800 and 2,400 assume a load that is not held for a whole show.
• In the UK, the fuse in your plug caps you at 13A regardless of the ring. The 32A on the breaker protects the building's cable, not your socket.
• Stop sizing your rig from the wattage on the box. A speaker advertised at 2,000 watts draws under 2 amperes in real use — the lighting rig is the load that matters.
• Get audio onto a different circuit from the bar and the lighting if you possibly can. A compressor motor starting draws six to ten times its running current, and everything on that circuit feels it.
• Never chain one power strip into another. They are listed for direct connection to a wall socket only, and the connection that overheats will be the one you cannot see.
• Use your heaviest cable for your longest run, and uncoil it. Thirty metres of 14 AWG at full load loses over 6% of your voltage before it arrives.
• Power up towards the amplifiers and power down away from them. It avoids the thump, and it stops several large inrush surges landing on one circuit at the same instant.
• When an RCD or GFCI keeps tripping, split the load across circuits. Every switch-mode supply leaks a little by design, they add up, and defeating the device is never the answer.