Our In-House EMC Test Facilities — and What It Means for Your Build

Efficiency and power density are the headlines. But an amplifier also has to behave in the system it lands in, and EMC discipline is one of the less visible reasons a Pascal module drops into a demanding product predictably.

It is not a capability we added when the market started asking for it — it is how we have engineered for years: not just the part that makes the sound, but the part that has to live cleanly alongside everything else in your enclosure.

Why We Test EMC In-House — and What It Means for Your Build

Electromagnetic compatibility is where audio integration projects quietly succeed or expensively fail. We have kept it inside our own four walls for the better part of a decade — here is why, and what that changes for the product you build around our amplifier.

Why it matters

A Class D amplifier is efficient because it switches. The output stage chops the supply rail hundreds of thousands of times a second, with fast voltage edges. Those same fast edges are a broadband noise source, and the energy finds more than one way out: it couples back onto the mains and DC supply lines as conducted noise, and it radiates — not only from the board, but from the output cabling and from the enclosure itself.

That last part is what makes EMC a system problem rather than a component spec. An amplifier can sit comfortably inside the conducted limits and still fail radiated, because an enclosure has resonant dimensions: a chassis can radiate efficiently at the frequency whose wavelength matches its geometry, behaving like an antenna no one designed. EMC emerges from the interaction of the amplifier, its power supply, the harness, and the metalwork around them.

This is also why EMC is the trap in an integration project. Problems tend to surface at the certification lab — late, after the layout, cable routing, and mechanical design are already locked. A failure at that stage is not a tweak. It is a redesign loop, weeks of lost schedule, and another booking at an external test house.

We reached that conclusion years ago, which is why our own EMC facility is not a recent addition.

What we actually do

Pascal has run its own EMC facility for the better part of a decade. Every amplifier platform we have developed in that time has been measured in it — modules are characterised against CISPR 32, the multimedia-equipment emissions standard harmonised in Europe as EN 55032, to its Class B limits: the stricter residential-grade tier rather than the more relaxed Class A. Two disciplines, both in-house:

Conducted emissions EN IEC 61000-4-6 on the AC mains ports, 9 kHz to 30 MHz, measured through a LISN.

Radiated emissions EN IEC 61000-4-3 from 30 MHz to 1 GHz at a 3 m distance, in a semi-anechoic chamber, against the Class B quasi-peak limits.

The measurement chain is what you would expect of a controlled setup: a calibrated EMI receiver, a LISN on the conducted path, and a bilog antenna on a turntable for the radiated work.

The point is not the equipment list. It is that EMC has sat inside our development loop, not at the end of it, across every platform generation we have released since the chamber went in.

What that looks like in practice

Getting a switching amplifier quiet is rarely one clean measurement. It is iterative work — mains routing, output filtering, shielding, grounding and earthing — with each change re-measured to see what moved and what it cost elsewhere. Tightening conducted emissions can shift the radiated picture; solving a radiated resonance can re-open a conducted margin.

Owning the chamber is what makes that loop fast. We change one variable and see the result the same afternoon, rather than waiting on an external slot and shipping hardware back and forth. The outcome is a module that is quiet by construction rather than quiet by luck.

What accumulates over years of doing this is harder to see than the chamber itself, and it is the part that matters most to you. Every platform generation, every topology change, every filter revision has been measured in the same controlled environment, which means we are not reading a result cold. We have a reference library of known-good behaviour to compare against, and engineers who recognise a signature before the plot has finished drawing — this looks like a common-mode path through the output harness; that resonance is the chassis, not the module. Test capacity can be bought. The judgement that comes from years of measuring your own designs has to be built.

How you benefit

EMC becomes a design input, not a post-mortem. Because we measure as we develop, emissions behaviour is engineered into the module — switching scheme, output filter, layout, grounding — rather than discovered by you at certification.

You start from a characterised baseline. You integrate a module whose emissions profile is already known and documented, so your system-level test becomes a verification step rather than a gamble.

A faster, cheaper route to certification. Fewer redesign loops, fewer trips to an external lab, and less schedule risk on the part of the design that is hardest and most costly to fix late.

A partner who has seen it before. When a system-level issue appears, we can help isolate whether it originates in the amplifier or in the integration — because we know precisely how the module behaves in a controlled environment, and because years of measuring our own platforms tell us what tends to move when the enclosure changes.

Evidence for your technical file. Pre-characterised emissions data supports your own CE and compliance documentation, within the scope of your own certification requirements.