Why your flagship headphones might need a mains upgrade before an amp

I was at a dealer in Surry Hills a few months back, sitting with a six-figure headphone stack — a flagship planar, a Class A amp the size of a toaster oven, a DAC that costs more than most people's cars — and the owner pulled out an extension lead from a hardware store to power the whole thing while he found "the good cable." The difference when he swapped it for the dedicated run was not subtle. Noise floor dropped, bass tightened, the soundstage stopped sounding slightly smeared at the edges. Same amp. Same headphones. Different electricity.
That's the uncomfortable truth about statement headphone systems that nobody wants to put on a spec sheet: at the point where your DAC, amp and headphones are all genuinely transparent, the thing holding the system back is very often the wall. Not the headphones. Not the amp. The mains power feeding all of it, and whether the circuit it sits on is quiet, correctly earthed and not sharing a board with a fridge compressor or a dimmer switch.
I want to be careful here, because this is exactly the kind of topic where high-end audio earns its reputation for snake oil. Magic mains cables that claim to "realign the electrons" are nonsense, and I'll say that as someone who covers six-figure gear for a living. But the underlying physics of mains quality, earthing and isolation is real, well understood, and worth your attention before you spend another cent on an amplifier upgrade.
Why mains quality matters more as the system improves
A cheap desktop DAC/amp with mediocre power supply regulation is, in a strange way, forgiving. Its own internal noise floor is high enough that a bit of mains-borne hash gets buried under circuit noise you were never going to hear anyway. Push upmarket — into the territory of gear like the RME ADI-2 DAC FS (check price) or genuinely reference-grade headphone amps — and the internal noise floor drops to the point where external noise sources start to matter. You've removed the masking.
This is the same principle that applies to loudspeaker systems at the statement end, and it's one reason the gap between a good system and a great one often isn't about buying a better box. It's about removing everything a better box lets you hear that you didn't want to hear. Marcus will tell you the same thing about amplifiers and gain structure — noise floor is noise floor, whether it arrives acoustically, electrically or digitally.
Switch-mode power supplies are the usual suspect. Most modern homes run dozens of them — LED drivers, phone chargers, induction cooktops, some NBN equipment — and all of them inject high-frequency switching noise back onto the mains. It's broadband, it's everywhere, and a headphone amp with a sensitive, low-noise front end can pick some of it up either through the power supply itself or through ground-loop coupling with other connected gear.
Ground loops are the real villain, not "dirty power"
If I had to pick the single most common and most fixable problem in high-end desktop systems, it's a ground loop between the computer, the DAC and the amp. You'll hear it as a faint hum or buzz that rises and falls with what's plugged into USB, or that disappears when you unplug the Ethernet cable from your streamer. That's not "dirty power" in the generic sense people throw around — it's a specific, diagnosable issue where two pieces of equipment are earthed at slightly different potentials and the difference finds its way into the signal path.
The fix is almost never a fancier power cable. It's usually one of: an isolated USB link, a DAC with proper galvanic isolation on its digital inputs, or simply running the computer and the audio chain off the same circuit so everything shares one earth reference. I've seen a four-figure amp "fixed" by moving a power board six feet to a different wall socket. No new gear purchased. Just a different, quieter path to the switchboard.
Australian homes vary a lot in how their circuits are run, and older properties in particular can have shared neutral or earth arrangements across multiple power points that create exactly this kind of problem. If you're serious about a desktop system and you're in a position to do it, a dedicated circuit back to the switchboard — ideally on its own breaker, not shared with anything that switches a heavy load — removes an entire category of noise before you've spent a dollar on hi-fi.
What genuinely helps, and what's theatre
I'll be blunt about where I land on this, because it's a mildly contrarian position in a hobby that loves to spend on cables: a well-made mains lead to Australian safety standards does the job a mains lead needs to do. I have not personally measured a repeatable, blind-tested difference between a quality standard power cable and an exotic one costing ten times as much, on the same circuit, with the same equipment. Where I have measured differences — real, repeatable ones on a scope — is with power conditioning that actually does something electrically specific.
A genuine isolation transformer breaks the direct electrical connection between your wall circuit and your gear, which can meaningfully reduce both noise coupling and ground-loop current. A well-designed regenerator rebuilds the AC waveform from scratch rather than filtering it, which is a different (and more expensive) approach again. Simple passive filtering with ferrite chokes and capacitors can knock down high-frequency switching noise but does very little about the ground-loop problem, which is the one that actually produces audible hum in most desktop rigs.
What doesn't help, in my experience: directional mains cables, "audiophile" power points that cost as much as a decent DAC, and anything marketed with the word "quantum" on the box. If a claim can't be explained in terms that AES, the IEC or a basic circuit diagram would recognise, treat it as marketing until someone shows you a measurement.
A cheap way to diagnose your own desk before spending anything
Before buying a conditioner or booking an electrician, there's a free diagnostic worth doing. Plug your entire headphone chain — amp, DAC, source — into a single power board on one wall socket, rather than spreading it across multiple outlets or circuits. Then listen with volume up and no music playing. If hum or hiss drops noticeably compared with how things were wired before, you've found a ground loop, and the fix is architectural rather than something you buy.
If the noise floor is still elevated with everything on one circuit, the next cheap test is to unplug anything non-essential on that same circuit — printers, monitors with their own power brick, USB hubs — one at a time. Switch-mode supplies on cheap accessories are a disproportionately common noise source, and I've traced more desktop hum to a budget USB hub than to anything resembling "dirty mains" in the way cable marketing describes it.
Only once you've ruled out ground loops and nearby switching noise does it make sense to think about conditioning or a dedicated circuit. Spending on either before doing this diagnostic is spending to fix a problem you haven't actually identified, which is a pattern I see constantly at the top end of this hobby — people buying their way past a problem that a $0 cable rearrangement would have solved.
When spending on power genuinely pays off
There's a real tier of system where the maths changes. If you're running a statement DAC and amp alongside other gear on a shared circuit in an older house, or if you're in a share house or apartment with electrically noisy neighbours (shared switchboards in older apartment blocks are a known source of this), a dedicated circuit run by a licensed electrician under the relevant AS/NZS wiring rules is a legitimate, measurable upgrade. It's also, frankly, one of the least glamorous purchases in high-end audio, which is probably why it gets skipped.
I'd put this ahead of a second amplifier upgrade for most people chasing the last bit of black background in a desktop system built around gear like the Chord Electronics Mojo 2 (check price) or a full desktop stack from our best desktop DAC/amps roundup. You can have the most linear amplifier in the world and still hear a 50Hz hum sitting under the music if the earthing on your circuit is doing something it shouldn't.
None of this applies evenly across every listener. If you're running a mid-tier desktop setup and you're happy with it, I wouldn't go looking for problems that aren't bothering you — this is a top-of-the-chain consideration, relevant once the rest of the system is already resolving detail you didn't know was missing. For context on where that ceiling actually sits, it's worth reading how far a genuinely reference digital front end extends this logic in our piece on the best DACs and network streamers.
The headphone-specific wrinkle: cable and connector earthing
One thing that catches people out specifically in headphone systems, rather than loudspeaker ones: balanced headphone connections (4.4mm, XLR4) don't automatically solve ground-loop hum the way people assume. A true balanced output does reject some common-mode noise, which is genuinely useful, but if the noise is being injected before the balanced stage — in the DAC's power supply, say — balanced wiring downstream doesn't touch it. I've had readers email convinced their new balanced cable "should have" fixed a hum that was actually sitting upstream in a shared USB ground with their laptop.
Open-back headphones also make this more audible than closed-back pairs, simply because they tend to be more revealing at low level and because people listening on open-backs are usually doing so in quiet rooms specifically to hear detail — the exact scenario where a faint 50Hz buzz becomes impossible to ignore once you've noticed it. If you've ever wondered why the same amp sounds dead silent at a hi-fi show and faintly noisy on your own desk, the show's power infrastructure, usually run through dedicated, well-earthed distribution boards for exactly this reason, is a big part of the answer.
Where I'd actually spend the money
If you've got a four-figure-plus headphone amp and you're still hearing any hum, hiss or grain that wasn't there when you auditioned it at a dealer, don't reach for the next upgrade in the catalogue. Reach for a multimeter, a different wall socket, and twenty minutes of patient troubleshooting first. The dealer demo almost certainly ran off a clean, dedicated circuit with nothing else plugged into it — your desk at home rarely does. Electricians worth their licence will tell you the same thing I tell readers: fix the circuit before you fix the component. It's not as exciting as a new DAC, and it won't get written up in a gear review. But it's very often the actual difference between the system you heard at the show and the one sitting in front of you now.
— Sofia Laurent, High-End & Statement Systems Editor, Sound Technology
Common questions
- Do expensive audiophile mains cables actually make a difference?
- For a cable meeting standard safety ratings, I haven't seen repeatable, measurable evidence that exotic mains cables outperform a well-made standard one on the same circuit. Where real differences show up is in isolation transformers, regenerators and circuit wiring, not the cable itself.
- How do I know if my headphone amp hum is a ground loop?
- Plug every component in your chain into a single power board on one wall outlet rather than spreading them across multiple sockets or circuits. If the hum drops or disappears, you've got a ground loop, and the fix is in how things are earthed and connected, not a new cable or amp.
- Is a dedicated power circuit worth it for a desktop headphone system?
- It's worth it once the rest of your chain is resolving enough detail that electrical noise is audible at normal listening levels, which usually means a genuinely high-end DAC and amp. For a mid-tier desktop setup it's rarely the limiting factor and I wouldn't prioritise it over the headphones or source.
- Does a balanced headphone cable fix mains hum?
- Only if the noise is being introduced after the balanced stage. Balanced wiring rejects some common-mode noise picked up along the cable run, but noise injected earlier, in a DAC's power supply or over a shared USB ground, passes straight through regardless of connector type.
I'm Sofia, and I get to play with the silly stuff — the statement amplifiers, the reference loudspeakers, the cost-no-object systems that most of us will only ever hear at a show. Someone has to, and I take it seriously: at this level the price stops mapping to performance and starts mapping to engineering, craft and ego, and part of my job is telling you which is which. I love the extreme end of this hobby, but I'm not dazzled by a big number on a price tag.
Covers flagship and cost-no-object reference systems
More from Sofia Laurent
Negative feedback in amplifier design: the argument that never endsNegative feedback cuts amplifier distortion but divides designers. Here's what it does, why some statement amps avoid it, and what it means for you.
The synergy tax: why mismatched gear costs more than cheap gearThe synergy tax explained: why pairing mismatched hi-fi components quietly costs more than buying modest, well-matched gear.
Rear-port floorstanders near a glass wall: a risky pairingRear-ported floorstanders against glass walls cause boundary reinforcement and reflections. Here's how to place, measure and tune them properly.
Infrasonic bass below 20Hz: chasing the octave you feel, not hearBelow 20Hz is infrasonic bass: felt more than heard. Here's what it takes to reproduce it properly, and whether a statement system should bother.
Why statement systems still get beaten by the zone two speakerA $150,000 reference system is only as good as the worst-sounding zone it's networked to. Here's why multiroom integration is the statement system's real weak point.
Why a statement floorstander can be the wrong speaker in the roomSix-figure floorstanders fail in rooms that were never built for them. Here's how to tell if your space can actually support one before you buy.