Hum or buzz in the sound system
very common
“There's a hum in the speakers that's there even with everything muted, and it gets worse when the house lights come up.”
Likely causes
-
Ground loop between chassis on different circuits, almost always via an unbalanced source
~35%Clean low hum at the mains fundamental with harmonics above it. Present with the input connected and the fader all the way down; gone the instant you unplug that input. A 1:1 isolation transformer on that one input kills it.
-
Unbalanced device into a balanced system, or a receiving input with a pin 1 problem
~20%Only the channels fed by that device hum. Changing the shield connection at the receive end changes the hum level: a properly built balanced input is indifferent to shield potential, so if it is not indifferent, the box is telling you it routes shield current through its audio ground.
-
Dimmer or VFD noise coupling into signal cable
~15%Harsh, buzzy and harmonic-rich rather than a smooth hum, and it tracks the dimmer position or a motor load. Ask someone to sweep the dimmer while you listen.
-
A second building service tying two grounds together: CATV coax, an OSP-fed device, a separate panel
~15%Pull the cable TV coax or the OSP-fed device and the hum stops dead. Where a legacy analogue video path shares the system (CATV baseband, composite, VGA) you will also see slow rolling hum bars in that path; a digital video path shows the same fault as sparkles or dropouts and never as bars.
-
Signal cable routed with or near mains, or a long unbalanced run
~10%Hum level changes as you physically move the cable away from the conduit or the transformer. Hands-on, and definitive.
-
Failing power supply in one device
~5%Hum persists with every input unplugged and all gain at minimum, and only on that one device, and the tone sits at twice mains frequency rather than at mains. Swap it out of the rack and the rack is quiet.
What to bring
- 1:1 line isolation transformer
- RTA or SPL app
- known-good balanced cables
- receptacle tester
- clamp meter
- headphones and listen bus
- screwdrivers
Safety
Never lift or defeat the equipment grounding conductor on a plug or a device to cure hum. It leaves the chassis unbonded, it is a shock hazard, it is a code violation, and it does not fix anything, it hides it. Mains wiring inside racks, receptacle changes and circuit moves are licensed electrical work in most jurisdictions.
Steps
-
Step 1: Unplug inputs one at a time, from the mixer or DSP downward
Whichever input's removal kills the hum owns the fault. With every input pulled and the amp still humming, the problem is downstream: amplifier, speaker line, or power. Two minutes, no tools, and it halves the system.
If that doesn’t do it
Hum survives with all inputs out: go to the amplifier leg.
-
Step 2: Unplug the amplifier input
Silent with the amp input open means everything upstream is suspect. Still humming with nothing on the input means the amplifier, its power, or the speaker line, including an induced loop in a long speaker run.
If that doesn’t do it
Amp hums bare: swap the amp to a different circuit, then swap the amp.
-
Step 3: Identify the frequency, and which harmonic dominates
By ear or with an RTA app. Note whether the dominant tone is the mains fundamental (50 or 60 Hz) or its second harmonic (100 or 120 Hz). Fundamental-dominant reads as an induced or conductive loop. Second-harmonic-dominant reads as power-supply ripple: a failing supply, or a loop closing through one, which sends you to the power-supply cause, not the shield. Buzz with strong high harmonics reads as SCR dimmer or switching noise. Three causes, three different fixes, so spend the thirty seconds.
If that doesn’t do it
Ambiguous: treat as a loop first, it is far more likely.
-
Step 4: Insert a 1:1 isolation transformer at the suspect input
Diagnostic and frequently the permanent fix. Quiet with the transformer in confirms a loop beyond argument. Carry one: it saves more afternoons than any meter in the bag.
If that doesn’t do it
No change with the transformer: the coupling is not conductive through that input. Go to routing and EMI.
-
Step 5: Get the offending device onto the same branch circuit as the rack
Same circuit, same phase, ideally the same receptacle strip. Devices on circuits fed from different panels are the classic setup for this fault, and a temporary extension lead is a legitimate test.
If that doesn’t do it
Still humming on one circuit: the loop is through a signal path, not the mains.
-
Step 6: Correct the interconnect and the shield
Balanced connections wherever the gear supports them. AES48 is a requirement on the equipment, not a wiring trick: a compliant unit bonds the cable shield to its shielding enclosure at the point of entry, at every connector, so shield current is carried by the enclosure and never enters the audio circuit. So bond both ends on a balanced interconnect, and if that hums, you have found a pin 1 problem in the receiving box, which you fix, replace or transformer-isolate. Lifting the shield at one end is an unbalanced-interconnect workaround, not AES48 practice, and it is not a design choice on a balanced line.
If that doesn’t do it
Improved but audible: re-route the signal cable away from mains and repeat the frequency check.
-
Step 7: Verify bonding of the rack and the AV ground
Rack bonded back to the TGB with a properly sized bonding conductor per ANSI/TIA-607 practice, not daisy-chained through equipment chassis. Confirm the receptacles feeding the rack are correctly wired and grounded before you blame the audio.
If that doesn’t do it
Bonding correct and hum remains: there is a second service or a building ground problem. Bring in the electrician with your measurements, not your opinion.
References
- AES48-2005 (r2019)
- ANSI/TIA-607
- NFPA 70 Art. 640
- NFPA 70 Art. 722 / Art. 725 (cabling requirements for power-limited circuits moved to Art. 722 in the 2023 edition; confirm the edition adopted in your jurisdiction)