Skip to content

Crado Pilot Program · Now welcoming hardware teams.Book a 30-minute call

Crado
Get started / Introduction
ON THIS PAGE

GET STARTED

Introduction

Crado connects a product revision, its applicable requirements and the evidence used to review an engineering finding. Its initial workflow supports radiated-emissions investigation, from report confirmation to a reviewed next-test plan and revision-specific results.

A test report states what was measured on one device, in one configuration, under recorded conditions. Crado keeps those facts attached to each other, so that a later reader can see which revision a result concerns, which requirement it was evaluated against, and what the evaluation relied on.

What a radiated-emissions investigation is

Connected electronics radiate unintended electromagnetic energy, often from clocks, switching regulators and cables. A radiated-emissions test measures that energy at set distances and frequencies and compares it with the limits in the applicable rule. For unintentional radiators sold in the United States, these limits are set out in 47 CFR 15.109.

An investigation begins when a test reports an emission above or close to a limit. The engineering questions that follow are practical: is the reported value right, which part of the design is a plausible source, what should be tested next, and does a later measurement on a changed revision actually show improvement?

How the work is divided

Crado uses three layers, each with a defined responsibility.

Model-assisted interpretationModels help interpret source material and propose candidate explanations during an investigation. Their output is treated as a proposal until confirmed. Report values are proposed from the report text and confirmed by an engineer before use.
Deterministic evaluationVersioned rules evaluate confirmed inputs. The same inputs and rule version produce the same result. A rule runs only when its required inputs are present.
Engineering reviewEngineers confirm extracted values, challenge interpretations and decide the physical next step.

Models do not produce pass or fail results. Evaluation results come from rules applied to confirmed values, and the decisions that follow belong to engineers.

First investigation

An investigation starts from a reported radiated-emissions finding. The sequence below describes the workflow. Screen names in the product may differ.

  1. 01

    Identify the product and revision

    Name the product and the revision that was tested. Evidence is recorded against this revision and does not transfer automatically to others.

  2. 02

    Provide an authorized report

    Supply a test report you are permitted to share. During a pilot, the reports and data in scope are agreed in advance.

  3. 03

    Confirm extracted values

    Check the frequency, level, detector, distance, polarization and reported limit and margin against the source. Correct or reject anything that does not match. See Report confirmation.

  4. 04

    Inspect the investigation

    Review what is observed, what is known about the product, what is inferred and what is missing. Candidate explanations are labelled as inferred.

  5. 05

    Review the next test

    A next-test plan is a proposal. An engineer accepts, changes or rejects it and decides what is run on the bench or in the chamber.

  6. 06

    Record results when available

    Record each result against the revision that was actually tested, with its conditions. A completed test is not a confirmed cause; see Observations and hypotheses.

Scope and limitations

In scope today

  • Radiated-emissions investigation for connected electronics.
  • Bounded checks under 47 CFR 15.109 for supported inputs. The exact clause and conditions are listed in Regulatory coverage.
  • Revision-specific recording of findings, next-test plans and results.

Not in scope

  • Certification, test-lab services or lab booking.
  • Replacing an accredited test report or a regulatory determination.
  • All of FCC Part 15 Subpart B. Only the checks listed under coverage are supported.
  • CISPR 32 / EN 55032, immunity standards and military standards.

LIMITATION

Extraction quality depends on the source. Scanned or incomplete reports produce more values that need correction, and some checks cannot run until missing conditions are supplied.