Every installer has met the cable that hums only when it runs beside a dimmer rack, or the 20 m unbalanced run that picks up a radio station. The fix is rarely a better shield. It is a balanced line, where two conductors carry the same signal with opposite polarity and the receiving circuit subtracts them, cancelling whatever noise both legs picked up in common. This article explains what a balanced XLR cable does electrically, which construction details actually protect the balance, and where balancing stops helping.
Quick answer: A balanced XLR cable uses three conductors: pin 2 hot, pin 3 cold and pin 1 shield, with the signal pair twisted at a consistent lay and driven by equal impedances. The receiver subtracts the two legs, so common-mode noise cancels. Use balanced XLR for any run over about 5 m, in high-interference rooms, and wherever ground potential differs between racks. Shielding alone does not make a line balanced.
Balanced XLR Cable: Why Three Conductors Reject Noise
The word balanced describes a circuit, not a connector. A three-pin XLR shell can be wired unbalanced and will still fit every socket in the building. What makes a line balanced is that two conductors carry equal signal voltages of opposite polarity into a differential input with matched impedances on both legs. Everything the cable does either supports that condition or quietly breaks it.
What balanced means at the receiver
Assume the source sends +1 V on pin 2 and -1 V on pin 3, referenced to pin 1. The receiver amplifies the difference between the two legs, which is 2 V of wanted signal. Now assume the cable passes through an electromagnetic field that induces 0.4 V on both legs equally. The difference between the legs is still 2 V, because 0.4 V appears on both and cancels in the subtraction. The wanted signal is doubled and the noise is rejected, which is why a balanced input can be 6 dB hotter for the same source level and dramatically quieter at the same time.
The rejection is never perfect. Common-mode rejection ratio, or CMRR, is the measure of how much of that shared noise survives. A well-built cable and a good differential receiver reach 60-90 dB of rejection in the audio band. A cable with mismatched leg resistance, uneven capacitance or a poor shield bond can lose 20-30 dB of that in a single run, which is the difference between silence and an audible hum.
The three conductors and what each one does
Pin 2 is hot, the in-phase signal. Pin 3 is cold, the same signal inverted. Pin 1 is the shield and reference. The modern US convention is pin 2 hot; legacy European equipment built to an older convention exists with pin 3 hot, and mixing the two conventions inverts polarity on one channel without causing any obvious noise. Label every assembly and keep the convention consistent across an inventory.
The shield has one job in a balanced line: intercept interference before it reaches the conductors, and provide a low-impedance reference path. It is not a signal return and it must not carry audio current. That distinction matters at the connector, where the shield should bond to the metal shell with a short, wide connection rather than a thin twisted pigtail.
Why twisting matters more than most buyers expect
Two parallel conductors form a loop, and any changing magnetic field through that loop induces a voltage. Twisting the pair at a consistent lay shrinks the effective loop area to almost nothing because the field alternates direction along the cable, so the induced voltages in successive half-twists cancel. A tight, regular twist is therefore the primary defense against low-frequency magnetic hum from transformers and dimmers.
Consistent lay also keeps the two legs electrically similar along the whole length. If the twist opens at a bend or the pair is untwisted for 100 mm behind the connector, capacitance and inductance become unequal at that point, and noise couples into one leg more than the other. What was common-mode becomes differential, and no receiver can subtract it out. This is why the last 100 mm at the termination is often where a hum problem lives.
The electrical numbers that define a good balanced cable
| Parameter | Good balanced cable | Typical failure mode | Effect on the line |
|---|---|---|---|
| Leg resistance match | Within 1-2 percent | One strand broken at a connector | Loses CMRR, hum appears under load |
| Capacitance core to core | 45-80 pF/m | High-capacitance audio cable on long runs | High-frequency roll-off, edge rounding |
| Capacitance leg to shield | Matched within 5 percent | Crushed jacket, uneven insulation | Converts common-mode noise to differential |
| Shield coverage | 85 percent braid or better | Spiral shield opened at a bend | Radio frequency ingress, buzz near transmitters |
| Twist lay | Regular, 20-40 mm | Untwisted pigtail at termination | Magnetic hum pickup near transformers |
| Shield resistance | Low, with drain wire in foil builds | Foil with thin drain only | Ground potential differences appear as noise |
Where balancing stops helping
Three conditions break the model, and buyers should know them before specifying cable as the fix for a noise problem.
Ground loops. If two devices sit on different ground potentials, current flows through the shield and the resulting voltage appears in series with the signal. A balanced input rejects some of it, but a large potential difference will overwhelm the rejection. The answer is a single grounded reference, an isolation transformer, or an optical link, not a heavier shield.
Unbalanced gear at one end. Driving a balanced input from an unbalanced output is workable; the reverse is not. Passive adapters that short pin 3 to pin 1 convert the line to unbalanced and throw away the rejection. Beyond about 5 m, an unbalanced TS or RCA run is exposed to hum and radio frequency pickup in a way no cable construction can fully correct.
Phantom power. Condenser microphones receive +48 V on pins 2 and 3 relative to pin 1. Connecting unbalanced equipment to a phantom-fed input can put that voltage across an output stage and damage it. Use an isolation transformer or a blocking adapter, and never assume a passive cable will protect the device.
What buyers ask about balanced cable
These questions come up in nearly every specification review, and the answers belong in the quotation rather than in a follow-up email.
- Does a balanced cable need a shield if the pair is twisted? Yes. Twisting handles magnetic coupling; the shield handles electric field and radio frequency interference, and it provides the reference. They solve different problems.
- Is star-quad worth it? It raises common-mode rejection further by averaging noise across two pairs, at the cost of capacitance and stiffness. Specify it for long mic runs in hostile rooms, not for dense rack wiring.
- Can I use a balanced cable for unbalanced signals? The cable will carry the signal, but the line loses its noise rejection. It is acceptable as a wiring convenience, not as a design decision.
- Should the shield be connected at both ends? For a cable, yes. Lifting one end is a diagnostic step for ground loops in installed systems, not a default build.
- What is the maximum balanced run? Line level to about 100 m is routine; mic level is best kept under about 50 m. Both are classes, not guarantees.
Three application scenarios
- Live stage, 40 m multicore to front of house. Balanced XLR with 95 percent braid and a tight twist handles the distance, but keep mic tie lines away from dimmer feeds and motor cables. Where a run must cross a power bundle, cross at right angles rather than running parallel.
- Studio patch bay, short runs, many connections. Balance still matters because the room contains computers and switch-mode supplies. Use low-capacitance braided cable, keep shield bonds short at the punch block, and avoid lifting shields on a patch bay unless a specific ground loop demands it.
- Corporate AV, laptop and media player into a house system. The source is unbalanced, so use a direct box or an active interface at the source rather than a passive adapter at the rack. The balanced run then starts at the box output, which is the only place where the conversion is clean.
Bench verification sequence
- Measure resistance from pin 2 to pin 2 and pin 3 to pin 3 end to end. The two readings should match within a fraction of an ohm.
- Measure capacitance from pin 2 to pin 1 and from pin 3 to pin 1. A large difference points to a damaged conductor or crushed insulation.
- Check continuity from pin 1 to the metal shell at both ends. A high reading means the shield bond is relying on a pigtail.
- Confirm no continuity between pin 1 and either signal pin; a short here will unbalance the line and often mute it.
- Flex the cable while watching the meter to catch a broken strand that only opens under movement.
- Test in the actual system with phantom power engaged if condenser mics are in use, then check for direct current on the signal pins.
Specification checklist
Write the following into the purchase order: two twisted conductors of equal gauge and stranding, 45-80 pF/m core-to-core capacitance, matched leg-to-shield capacitance, 85 percent or better braid coverage with a short shell bond, a jacket that survives coiling, and a documented pin 2 hot convention. Add a 100 percent continuity and short test per assembly. Those requirements describe a cable that supports the balance rather than merely fitting the connector, and they are the cheapest insurance against a hum that only appears after installation.
Related guides: shield construction for balanced lines, XLR cable wiring conventions, XLR mic cable construction.
FAQ
1. What is the difference between a balanced and an unbalanced XLR cable?
A balanced XLR cable carries the signal on two conductors with opposite polarity plus a shield, so a differential receiver can subtract common-mode noise. An unbalanced cable uses one conductor plus shield as the return, so any noise on the line is amplified along with the signal. Both fit the same connector, which is why the distinction is often missed at installation.
2. How much noise does a balanced line actually reject?
A good cable with a good differential input reaches 60-90 dB of common-mode rejection in the audio band. Real-world figures depend on how well the two legs are matched in resistance and capacitance. A single broken strand or an untwisted pigtail at the connector can cost 20-30 dB, which is enough to turn a quiet line into an audible hum.
3. At what length does balanced cable become necessary?
Practically, above about 5 m. Below that, an unbalanced run in a quiet room is usually fine. Above it, hum and radio frequency pickup grow quickly, especially near dimmers, transformers or computer equipment. Line-level balanced runs to about 100 m are routine, and mic-level runs are best kept under about 50 m.
4. Does a balanced cable fix a ground loop?
No. A ground loop is caused by a potential difference between two safety grounds, which drives current through the shield. Balancing rejects some of the resulting voltage but cannot remove a large difference. The correct fixes are a single grounded reference point, an isolation transformer, or an optical link between the two systems.
5. Can I connect unbalanced equipment to a balanced input with a passive adapter?
You can, and it usually works, but the line behaves as unbalanced from that point onward. A passive adapter typically shorts pin 3 to pin 1, which discards the rejection the input could provide. Keep such runs short, and never connect unbalanced output gear directly to a phantom-powered input, because the 48 V can damage the output stage.