TL;DR -- EMI (electromagnetic interference) is the noise your device emits. EMC (electromagnetic compatibility) is the broader discipline of making sure your device neither causes interference nor gets disrupted by it. "EMI testing" means emissions testing. "EMC testing" means emissions plus immunity. The distinction determines which tests you run, what they cost, and whether your product is legal to sell in a given market.
The terms, defined
Three acronyms show up constantly in compliance conversations, and they get used interchangeably by people who should know better. Here is what each one actually means.
EMI -- Electromagnetic Interference. The unwanted electromagnetic energy a device generates. Your switching power supply dumps noise onto the AC line. Your clock oscillator radiates harmonics from PCB traces. EMI is the problem.
EMS -- Electromagnetic Susceptibility. How vulnerable a device is to external interference. An ESD discharge resets your microcontroller. A nearby radio transmitter corrupts your ADC readings. EMS is the vulnerability.
EMC -- Electromagnetic Compatibility. The discipline that covers both sides: controlling what your device emits (EMI) and hardening it against what it receives (EMS). A device that coexists with its electromagnetic environment without causing or suffering from interference has achieved EMC.
ANSI C63.14 puts it well: "We control electromagnetic interference in order to achieve the desired state of electromagnetic compatibility."
flowchart TD
EMC["EMC\nElectromagnetic Compatibility\n(the goal)"]
EMC --> EMI["EMI Control\n(Emissions Testing)"]
EMC --> EMS["EMS Control\n(Immunity Testing)"]
EMI --> RE["Radiated Emissions\nCISPR 32 / FCC Part 15B"]
EMI --> CE_test["Conducted Emissions\nCISPR 32 / FCC Part 15B"]
EMI --> HARM["Harmonics & Flicker\nIEC 61000-3-2 / -3"]
EMS --> ESD["ESD\nIEC 61000-4-2"]
EMS --> RI["Radiated Immunity\nIEC 61000-4-3"]
EMS --> EFT["EFT / Burst\nIEC 61000-4-4"]
EMS --> SURGE["Surge\nIEC 61000-4-5"]
EMS --> CI["Conducted Immunity\nIEC 61000-4-6"]
EMS --> VD["Voltage Dips\nIEC 61000-4-11"]
style EMC fill:#1e3a5f,color:#fff
style EMI fill:#2a4a6f,color:#fff
style EMS fill:#2a4a6f,color:#fff
When someone says "EMI testing," they mean emissions testing only, measuring the noise your device puts out. "EMC testing" means both emissions and immunity. People swap the terms all the time in conversation. On a purchase order, the difference is thousands of dollars and extra days of lab time.
Why people confuse them
Mostly it comes down to the US market. The FCC only requires emissions testing. There are no federal immunity requirements for consumer electronics. If you have only ever sold products domestically, you have only ever done "EMI testing," and you may have called it "EMC testing" your entire career without anyone correcting you.
The test labs do not help. A lab might advertise "EMC testing services" when what they mean is "we have a chamber and a LISN and we measure emissions." Smaller labs may not even own immunity test equipment.
And the acronyms are genuinely similar. Three letters, two of them shared. In a noisy meeting or a dense email thread, EMI and EMC look and sound almost identical.
None of this matters until you try to sell in a market that requires both halves. Then it hits your budget.
What EMI testing actually involves
EMI testing is the emissions half of EMC. The question is simple: does your device put out too much electromagnetic noise? The tests measure what your product radiates into the air and what it conducts back onto power lines.
Radiated emissions
Your device sits on a turntable inside a semi-anechoic chamber. An antenna, typically 3 or 10 meters away, scans across the frequency range (30 MHz up to several GHz) while the turntable rotates and the antenna moves up and down on a mast. The lab is looking for the worst-case orientation, the angle and height where your device radiates the most energy.
The measured field strengths get compared against regulatory limit lines. For FCC Part 15 Class B (consumer devices) at 3 meters:
| Frequency Range | Limit |
|---|---|
| 30 -- 88 MHz | 40.0 dBuV/m |
| 88 -- 216 MHz | 43.5 dBuV/m |
| 216 -- 960 MHz | 46.0 dBuV/m |
| Above 960 MHz | 54.0 dBuV/m |
Clock harmonics are the most common cause of failure. A 48 MHz USB clock generates harmonics at 96, 144, 192, and 240 MHz, and one of those will land near an FCC limit. Switching power supply noise and cable radiation are next on the list.
Conducted emissions
If your product plugs into AC mains, a Line Impedance Stabilization Network (LISN) measures the RF noise your device dumps back onto the power line. The measurement range is 150 kHz to 30 MHz, and the limits require your device to pass with both quasi-peak and average detectors.
This is mostly a power supply test. The switching fundamental and its harmonics dominate the data. Missing or undersized input filters (no common-mode choke, no X or Y capacitors) are the most common root cause of conducted emissions failures.
Harmonic current emissions and voltage flicker
Two additional emissions tests apply for products sold in the EU: IEC 61000-3-2 (harmonic current distortion on the mains) and IEC 61000-3-3 (voltage fluctuations caused by varying loads). These are not required by the FCC. They add roughly $500 -- $1,500 and half a day of lab time to a CE campaign.
What immunity testing adds
Immunity testing is the other half of EMC, sometimes called EMS testing. The question flips: can your device keep working when external interference hits it?
The cost jumps here. Instead of passively measuring what your device emits, the lab actively attacks it with calibrated disturbances while monitoring whether it keeps functioning.
| Test | Standard | What It Simulates | Typical Test Level |
|---|---|---|---|
| ESD | IEC 61000-4-2 | Human body or furniture discharge | ±4 kV contact, ±8 kV air |
| Radiated immunity | IEC 61000-4-3 | Nearby radio transmitters | 3 -- 10 V/m, 80 MHz -- 6 GHz |
| EFT / Burst | IEC 61000-4-4 | Switching transients on power lines | ±1 kV power, ±0.5 kV signal |
| Surge | IEC 61000-4-5 | Lightning indirect effects, grid switching | ±1 kV line-line, ±2 kV line-ground |
| Conducted immunity | IEC 61000-4-6 | RF coupled onto cables | 3 V, 150 kHz -- 80 MHz |
| Voltage dips | IEC 61000-4-11 | Brownouts, momentary power loss | 0%, 40%, 70% of nominal |
| Power frequency magnetic field | IEC 61000-4-8 | 50/60 Hz fields from nearby equipment | 3 A/m |
Each test has defined pass/fail criteria:
- Criterion A: Normal operation during and after the test. No degradation at all.
- Criterion B: Temporary performance loss during the test, but the device recovers on its own afterward.
- Criterion C: Temporary loss that requires user intervention (a reboot, a reset). Still a pass in some standards.
- Criterion D: Permanent damage or data loss. Always a fail.
ESD is the most common immunity failure. A discharge to an unprotected USB port resets the processor, corrupts data, or kills an IC. The fix is TVS diodes placed within millimeters of the connector pins, but that requires board space and a design that anticipated the problem.
Radiated immunity is the hardest to test outside of a lab. You need an anechoic chamber, a signal generator, an amplifier, and calibrated antennas. There is no cheap bench equivalent. Products that were never designed with immunity in mind tend to have analog circuits that pick up RF fields and digital buses that glitch under exposure.
Which markets require what
The EMC vs EMI distinction has direct financial consequences depending on where you sell.
| Market | Emissions (EMI) Required? | Immunity (EMS) Required? | What You Actually Test |
|---|---|---|---|
| US (FCC) | Yes -- Part 15 Subpart B | No | Emissions only |
| EU (CE marking) | Yes -- EN 55032 | Yes -- EN 55035 + IEC 61000-4-x | Full EMC |
| Canada (ISED) | Yes -- ICES-003 | No | Emissions only |
| UK (UKCA) | Yes -- same as CE | Yes -- same as CE | Full EMC |
| Australia (ACMA) | Yes -- AS/NZS CISPR 32 | Yes -- AS/NZS CISPR 35 | Full EMC |
| Military (MIL-STD-461) | Yes | Yes | Full EMC, single standard |
For medical devices, IEC 60601-1-2 requires full EMC testing (emissions and immunity) regardless of market, with stricter levels than commercial standards.
Bottom line: US-only means emissions testing. EU, Australia, or UK means emissions and immunity. Plan accordingly.
The cost gap between EMI and EMC testing
The terminology difference shows up on invoices.
| Scope | Typical Cost | Lab Time | What You Get |
|---|---|---|---|
| Emissions only (FCC) | $2,000 -- $5,000 | 1 -- 2 days | FCC SDoC or certification for unintentional radiators |
| Emissions + immunity (CE EMC) | $5,000 -- $12,000 | 3 -- 5 days | CE marking compliance for EMC Directive |
| FCC + CE combined | $8,000 -- $15,000 | 3 -- 5 days | Both markets, with test data reuse |
Immunity testing roughly doubles your EMC budget. A product that costs $3,000 to test for FCC will cost $8,000 -- $10,000 once you add the full CE immunity suite. The immunity tests themselves account for $3,000 -- $8,000 of that, plus 1 -- 4 additional days of lab time at $2,000 -- $3,000 per day.
Test data reuse helps when you are testing for both markets simultaneously. A product that passes FCC radiated emissions will almost always pass CE radiated emissions -- the limits are close, and the measurement methods overlap. Running FCC and CE at the same lab in the same campaign saves 20 -- 30% vs running them separately. The immunity tests are the incremental cost, not the emissions tests.
The expensive surprise is when a product designed for FCC-only later needs CE marking. If immunity was never part of the design, the product frequently fails EU immunity tests on the first attempt. ESD immunity is the most common failure: unprotected I/O ports, missing TVS diodes, poor ground bonding at connectors. Fixing these after the board is finalized means rework, a potential re-spin, and another round of lab time.
How EMI and EMS testing differ in practice
Beyond cost, the two halves of EMC testing differ in equipment, methodology, and what kind of problems they surface.
| Aspect | Emissions (EMI) Testing | Immunity (EMS) Testing |
|---|---|---|
| Goal | Measure what your device puts out | Stress your device with external disturbances |
| Equipment | Spectrum analyzer / EMI receiver, antennas, LISN | ESD gun, surge generator, EFT generator, RF amplifier, coupling networks |
| Environment | Semi-anechoic chamber or OATS | Shielded room or chamber (depending on test) |
| Pass/fail | Measured levels vs regulatory limit lines | Device functionality during and after disturbance |
| Pre-compliance | Feasible with $3K -- $8K in bench equipment | Difficult -- most tests need specialized generators costing $3K -- $20K each |
| Common fixes | Filtering, shielding, layout changes, spread-spectrum clocking | TVS diodes, decoupling, cable ferrites, improved grounding |
| Design stage impact | Can often be fixed with component-level changes | Harder to retrofit -- grounding and protection need to be designed in |
One practical consequence: pre-compliance testing works well for emissions but poorly for immunity. You can buy a spectrum analyzer, a LISN, and a set of near-field probes for $3,000 -- $8,000 and catch most emissions problems before they reach the lab. For immunity, an ESD gun alone costs $3,000 -- $8,000, a surge generator runs $8,000 -- $20,000, and radiated immunity requires a chamber you do not own. Most teams discover immunity problems at the formal lab for the first time.
This asymmetry is why FCC-designed products so often fail CE immunity tests. Engineers who have spent their careers on emissions develop strong intuition for clock harmonics, input filters, and cable radiation. Immunity is a different problem set -- ESD paths, RF susceptibility in analog circuits, voltage transient resilience -- and it requires design attention that FCC never forced anyone to develop.
Planning a test campaign: EMI-only vs full EMC
If you know your target markets upfront, you can plan your test campaign and design to match.
US-only product (EMI testing)
Test to FCC Part 15 Subpart B. You need radiated emissions and conducted emissions (if AC-powered). No immunity. Budget $2,000 -- $5,000 for lab testing, 1 -- 2 days of lab time.
Design priorities: input EMI filter on the power supply, clock edge rate control, cable filtering at connectors, shielded enclosure if the board is noisy. Focus your pre-compliance testing on radiated and conducted emissions.
US + EU product (full EMC testing)
Test to FCC Part 15B for emissions and EN 55035 / IEC 61000-4-x for immunity. Budget $8,000 -- $15,000 for a combined campaign, 3 -- 5 days of lab time.
Design priorities: everything from the emissions list, plus ESD protection on every user-accessible port (TVS diodes within millimeters of connector pins), decoupling on analog and sensor inputs for RF immunity, proper chassis grounding for surge and EFT withstand, and a power supply that rides through voltage dips. These need to be in the design from the start. Adding ESD protection to a finished board is doable but painful. Adding surge resilience to an AC input that was never designed for it usually means a board re-spin.
Expanding from US to EU after initial certification
This is the most expensive path. You have already committed to a design that may not account for immunity. Budget an extra $3,000 -- $8,000 for immunity testing, plus the real risk of $5,000 -- $15,000 in redesign if the product fails. Products in this situation fail EU immunity testing on their first attempt more often than not.
If there is any chance your product will sell outside the US, build ESD protection and basic immunity hardening into rev 1. The BOM cost is $0.50 -- $2.00 per unit. The cost of retrofitting it later is $5,000 -- $30,000 in engineering, re-spin, and retest fees.
The MIL-STD-461 exception
Military EMC testing under MIL-STD-461 does not split emissions and immunity into separate standards. One standard covers both, with requirements that are generally stricter than commercial limits in both directions. MIL-STD-461 also adds test categories that commercial standards skip, like power lead susceptibility and antenna port conducted susceptibility.
If your product targets defense applications, the EMC vs EMI distinction matters less day to day -- you test for everything under one standard. The budget and timeline are larger, though: $15,000 -- $50,000+ and 2 -- 6 weeks of lab time depending on the number of applicable test methods.
Quick reference: EMC vs EMI vs EMS
| EMI | EMS | EMC | |
|---|---|---|---|
| Full name | Electromagnetic interference | Electromagnetic susceptibility | Electromagnetic compatibility |
| What it means | The noise | The weakness | The goal |
| Testing scope | Emissions only | Immunity only | Both |
| FCC requires it? | Yes | No | Partially (emissions only) |
| CE requires it? | Yes | Yes | Yes (full) |
| Typical cost | $2,000 -- $5,000 | $3,000 -- $8,000 | $5,000 -- $12,000 |
| Common failures | Clock harmonics, SMPS noise, cable radiation | ESD resets, RF susceptibility, surge damage | All of the above |
Where this leaves you
Decide your target markets early and design for the most demanding one. Adding immunity hardening to a clean-sheet design costs almost nothing. Retrofitting it costs a lot.
When you book lab time, make sure you and the lab agree on scope. "EMC testing" should mean emissions plus immunity if you need CE marking. A lab quoting $3,000 for "EMC testing" probably means emissions only. Ask.
And when budgeting, account for the immunity delta. US-only products are cheaper to test, but expanding to the EU later without planning for immunity upfront is the most expensive path of all.
Our requirements tool maps your device to the right test list, or read the EMC testing hub for the full picture from pre-compliance through final certification.
Found an error or something out of date? Let us know.