Long XLR Cable Runs: Capacitance, Level and Noise

A 300 ft run from a delay tower back to the desk looks like a job for the same cable used at 20 ft, and often it is. Then the console reports a dull top end on the drum overheads, or the channel clicks every time a dimmer pack changes scene. The balanced pair is doing what it was designed to do; the length has moved the problem from level to capacitance and ground. Knowing where those limits sit is what separates a working long run from a hopeful one.

The Electrical Cost of Distance

A microphone cable is a capacitor with a signal path through it. Each foot of twisted pair adds 20 to 40 pF between the two conductors, and that value is set by insulation thickness and the dielectric material, not by price. On a 100 ft run at 30 pF per foot the pair presents about 3 nF, which together with the source impedance forms a single pole low-pass filter. Nothing about that filter is digital or mysterious: it is the same RC network you would draw on paper. Both conductors are identical, so the loss is symmetrical on each leg and largely cancels at a differential input. What surprises people is how high the corner sits, which is why most long runs fail for reasons other than bandwidth.

Capacitance per Foot and the RC Corner

With a 200 ohm source and 3 nF of cable, the corner lands near 265 kHz. Halve the length and it doubles; double the length and it halves. That arithmetic is worth doing before a long install, because it tells you whether the cable is a factor at all. At 500 ft and 15 nF the corner is still around 53 kHz, well above any microphone’s usable range. A cable with unusually high capacitance per foot, or a source closer to 1 kohm, moves the number into the audible band quickly.

Why the Corner Sits Above the Audio Band

The reason a 150 to 200 ohm microphone survives 300 ft is that its output impedance is low compared with the reactance of the cable capacitance at 20 kHz. Reactance falls as frequency rises, so at 20 kHz a 15 nF pair is only about 530 ohm, still above the source. Push the source to 1 kohm, as some vintage mics and passive DI boxes do, and the divider starts to matter: 100 ft gives a corner near 53 kHz, and 500 ft brings it down to roughly 10 kHz.

Noise Rejection Has Conditions

A balanced input rejects whatever appears equally on both legs. It does that by subtracting one leg from the other, so any voltage common to both cancels while the difference survives. The rejection is never infinite and it degrades as frequency rises, because the two legs no longer look identical to the interfering source. A perfect pair with a perfect input might manage 60 dB or more of rejection in the midband; a real cable with a corroded pin 3 contact and a shield terminated at one end will do far less. That is why a long run that hums is usually a termination fault rather than a length problem. Distance only multiplies whatever the termination and the connector already do to the pair.

What Common-Mode Rejection Actually Cancels

The input amplifier sees two voltages and amplifies the difference. If a noise current induces 1 V on both legs, the difference is zero and nothing reaches the converter. But the cancellation depends on the source impedances matching. Add 2 ohm of contact resistance to one leg and the induced voltages no longer match, so a fraction of that 1 V becomes differential signal. At 60 Hz the error is small; at 10 kHz, where cable capacitance unbalances the legs, it grows. Clean connectors are part of the noise budget.

RF on an Unshielded Pair

Common-mode rejection does nothing for a signal that arrives on only one conductor. Radio frequency from a nearby transmitter, a lighting dimmer or a switching supply couples into an untwisted pair or through a shield that is not bonded at the connector shell. Once inside, it rectifies at an input stage and appears as a buzz or a burst of noise. A shield with high coverage and a low resistance drain path is the defense, plus keeping the pair twisted right up to the pins rather than splayed open inside the connector.

Ground Potential Between Buildings

Two buildings rarely share a ground reference. The difference can be a fraction of a volt or several volts, depending on load balance and the distance to the service entrance, and it appears across the shield of any cable that connects equipment in both places. That voltage drives current through the shield, and the shield is not a perfect conductor, so part of the voltage develops between the two ends of the cable and adds to the signal. The audible result is hum at the mains frequency and its harmonics, often modulated by whatever else is drawing power. Measure it before blaming the cable: with equipment disconnected, read the AC voltage between the two chassis, then repeat with the cable connected, and the difference tells you how much current the shield is carrying.

Snake, DI or Fiber: Picking the Right Tool

A snake is the right answer when the problem is cable count, not distance. Multicore bundles many pairs into one jacket with a shared overall shield, so one pull handles sixteen channels, but each pair still has the capacitance of its length. A DI box solves a different problem: it converts a high impedance unbalanced source, such as a guitar or a keyboard, into a low impedance balanced line that can survive the run. A transformer inside the DI can also break a ground loop. Fiber is for the cases copper cannot solve, meaning runs beyond roughly 1000 ft, complete galvanic isolation between two services, or a path through a building where grounding is unknown. Fiber carries digital audio, so it needs a converter at each end, and the analog signal is converted before the run rather than after it.

Estimated response and risk by run length, assuming 30 pF per ft and a 200 ohm source

Real-World Application Scenarios

At an outdoor festival, a delay tower sits 400 ft from the monitor world, and the crew runs a digital snake rather than analog copper because the two ends sit on different generator grounds. In a church that feeds an overflow room through a gymnasium, a 250 ft run picks up hum from the shared service, so an isolation transformer goes in at the receiving end. In a studio with a live room 60 ft from the control room, a star quad cable handles the distance next to dimmer packs without measurable noise, and the run is short enough that capacitance never enters the conversation.

Related Guides on This Site

For the wider design context, read the long XLR cable guide first, then the balanced audio cable guide for how the differential pair behaves, and the general XLR cable guide for jacket, gauge and connector choices that hold up in a venue. If the run crosses buildings, plan isolation before anyone pulls cable.

Frequently Asked Questions (FAQ)