0
\$\begingroup\$

I would like to test the EMI performance of our product in our lab. I have access to a small shield box, a spectrum analyzer.

  1. Should I test the device inside the shield box to isolate external noise?
  2. If I test the test outside the shield box, how should I factor out the ambient noise?
  3. What kind of antenna should I use? I would like to test the device EMI performance over a range of 30 MHz up to 1 GHz.
\$\endgroup\$
4
  • \$\begingroup\$ You typically can't test RF inside a box because of reflections and near field properties. Also it all depends on what kind of EMI you wish to measure: conducted, radiated or RF properties. Some of these you might be able to measure with near field probes - I have a set of expensive active probes for my spec but I never quite managed to use them in a meaningful way since they are so dang sensitive picking up literally everything in the building. \$\endgroup\$
    – Lundin
    Commented Jun 20, 2022 at 10:40
  • \$\begingroup\$ Anyway, I think you need to clarify more precisely what you wish to measure and if this is a product with or without RF parts ("intentional/unintentional radiator" as the standards like to call it). \$\endgroup\$
    – Lundin
    Commented Jun 20, 2022 at 10:41
  • \$\begingroup\$ I guess the most "meaningful" you can do is to use near field probes and check the general noisiness. If you have still obvious non-ambient noise more than 10 cm away, you have issues \$\endgroup\$
    – tobalt
    Commented Jun 20, 2022 at 10:42
  • \$\begingroup\$ If box is too small and background noise too high, consider going outside. \$\endgroup\$
    – winny
    Commented Jun 20, 2022 at 10:43

1 Answer 1

2
\$\begingroup\$

Forget about the box. What you need is near field probes and a optionally log-periodic type antenna.

You will always be able to tell if noise is from your EUT or from external just by switching your EUT off and compare measurements to EUT powered on.

The logper antenna can be used broadband from 100 MHz to 1GHz and it is very directional (~ 6-10 dBi) So if you aim at your EUT, it will not pick up too much noise from outside. Distance should be >50cm (far field) and <2m (because of reflections of the walls).

But you could also just measure the cable emissions, because almost always any RF emitted by EUT can be measured on the connected cables.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.