Testing products’ reliability under harsh conditions
T&MW Staff -- Test & Measurement World, 7/1/2004
| For more articles about Electrical & Physical Test topics, see: EMC Test Environmental Test |
| To find manufacturers and distributors, visit these sections of the Buyer's Guide: Plus, also see: |
Perhaps the most daunting challenge in the area of physical and electrical environmental test is EMI compliance for high-frequency products. Many products use clock frequencies above 1 GHz, and many more emit harmonics exceeding that frequency. Current EMI standards, however, specify neither procedures nor acceptance criteria for evaluating a test site at frequencies above 1 GHz.
In the US, FCC CFR 47 Part 15 requires measurements up to 40 GHz. The standard relies on ANSI C63-4.2001 for a site-evaluation procedure. Currently, a site that meets acceptance for 30 MHz to 1 GHz may also be used above 1 GHz. But test sites behave differently above 1 GHz than they do at lower frequencies, which renders the evaluation method inadequate. Outside of the US, few standards require measurements above 1 GHz, but that must change to keep up with technology.
“We need a single, worldwide evaluation procedure and set of test criteria,” says Werner Schaefer, senior compliance engineer at Cisco Systems (San Jose, CA). Schaefer participates in subcommittees from both the International Special Committee on Radio Interference (CISPR) and American National Standards Institute (ANSI) that aim to develop evaluation procedures and criteria for test sites.
Currently, engineers evaluate test sites by measuring their response at five locations to a swept-sine signal from 30 MHz to 1 GHz. Because radiation patterns change as frequencies rise above 1 GHz, five locations aren’t enough. Adding more locations increases site-evaluation time to the point where the swept-sine method becomes impractical. Mike Windler, business manager for EMC at Underwriters Laboratories (Northbrook, IL) and chair of the ANSI subcommittee working group that’s drafting the requirements, reports that a new method may make test manageable. This method replaces the spectrum-analyzer/signal-generator frequency sweep with a time-domain method using a vector network analyzer, which generates a sufficient bandwidth in less time.
It would help the industry if the CISPR and ANSI subcommittees come to agree on one worldwide standard. In the meantime, as Windler notes, “technology will continue to outpace consensus.” T&MW
