Failing an EMC test on the first attempt is one of the most common experiences for R&D and QC teams — don't panic, most issues have well-established fixes. This article goes deeper, covering the specific frequency bands behind common failures and realistic retest costs.
Radiated Emissions Exceeding Limits
This is the most common failure. The test frequency range typically covers 30MHz-1GHz (some products require testing up to 6GHz), with the applicable limit depending on product category and use environment (Class A industrial vs. Class B residential) — Class B limits are generally stricter than Class A. Exceeding the limit is usually related to PCB layout, clock circuits, or grounding design. Common fixes include:
- Optimizing PCB traces, shortening high-frequency signal line lengths
- Adding shielding enclosures or shielding materials
- Adjusting grounding to ensure sufficiently low ground impedance
- Adding ferrite beads on critical signal lines
Conducted Emissions Exceeding Limits
The test frequency range typically covers 150kHz-30MHz, and this most often shows up at the power input. Common fixes include:
- Adding an EMI filter
- Optimizing power module layout to reduce switching noise coupling
- Checking whether cable shielding is properly grounded
Immunity Test Failures
The product exhibits abnormal behavior (crashing, resetting) under external electromagnetic interference. Relevant test items include electrostatic discharge (ESD), radiated RF immunity, and electrical fast transient/burst (EFT/Burst) — the specific voltage/field strength levels for each item depend on which standard applies to the product. Common fixes include:
- Adding decoupling capacitors to critical circuits
- Optimizing the software watchdog mechanism for automatic recovery after abnormal events
- Adding common-mode chokes on signal lines
Retest Cost & Timeline Reference (For Budget Planning Only, Not a Quote)
A retest typically only covers the specific failed items, without needing to redo the full EMC test suite — so the cost is correspondingly lower than the initial full test. For timeline, if the corrective plan is straightforward (like a small hardware change such as adding a filter), a standard retest typically takes a few business days to about a week. If it involves respinning the PCB for validation, the timeline extends noticeably — that prototyping time needs to be factored into the overall plan.
Recommended Remediation Process
- Get the detailed failure data (not just "failed" — look at the specific frequency band and how much it exceeds the limit; the margin of exceedance directly affects which corrective approach makes sense — exceeding by 3dB versus 15dB calls for very different levels of intervention)
- Have your engineering team assess whether a hardware or software-level fix is more cost-effective
- Validate the fix on a small batch of samples
- Retest to confirm
Frequently Asked Questions
Q: Is a small margin of exceedance (like 1-2dB) easier to fix? A: Usually yes — small margins can often be resolved through software optimization or a simple filter component adjustment, without necessarily needing a major hardware redesign. That said, it still depends on which frequency band the exceedance occurs in and the product's structure, so it's not a universal rule.
Q: How much do Class A and Class B limits differ? A: The specific gap varies by frequency band, but generally Class B (residential/domestic environments) limits are stricter than Class A (industrial environments). If a product is sold into both residential and industrial settings, it's worth designing and testing to the stricter Class B standard from the start, to avoid needing to retest later if sales channels expand.
Throughout testing, we provide detailed raw data and engineering recommendations — not just a pass/fail conclusion — so your team can pinpoint issues more precisely and reduce the number of retest cycles.