Home based EMC measurements

wide

Andy writes:

For frequencies that can be built with a patch antenna, the following procedure can be done:

1) Connect a signal generator to a patch antenna (for each frequency to be calibrated, which points UPWARD to the cosmos. 2) Suspend a dippole antenna 20 or 30 wavelengths above the patch.( Not a problem above , say, 500 mhz).... (minimal reflections this way) 3) The antenna should be a SMALL dipole, much shorter than a quarter wave at the lowest freq. 4) Develop a calibration curve for this antenna.

5) When making any measurements , then, at these frequencies, you must allow for reflections, depending on your measurement facility. This procedure is well established in literature for "open field antenna measurements".

To simplify this summary:

It is not possible to do this in a simple manner. Antenna field measurements are extremely complicated by reflections, absorption, and inaccuracy in the transmitted field. If you buy a CALIBRATED field strength meter, and attempt to make measurements, you may find up to

20-30 db field strength difference base on 2 or 3 feet of field strength meter movement...... this is a bear....

If you have a calibrated large room lined with 400 Mhz ECCOSORB, you would have not asked ( big grin).

At lower frequencies ... well, you just do the best you can. The near field curvature will kill you and reflections may bury you....

If you just want to build a ball park value, just put a signal generator into a dipole, and mount the pickup antenn 3-4 wavelengths away, and raise and lower it a half wavelength or so to see the range in values. This ain't bad for ball park figures.....

Sorry,.... but that is why EMC engineering is such a definite specialty.....

Andy , retired NARTE Licensed EMC engineer

The antennas need a calibration curve (sensitivity VS frequency) to be useful.

These guys have some neat looking stuff:

John

Hi All,

I have a spec analyser and I looking for suggestions on antennas and comb generators within an afordable budget..

I have seen some scanner antenna's that are log periodic that have a wide bandwidth.

The main goal is to get some idea if my product is radiatating somewhere near the EMC limits or not.

Thanks in advance.

Joseph

The simple answer is to listen to the output of the receiver and switch the EUT off and on - that'll determine where the emission is coming from.

Graham

what an utterly confusing website, I'm looking for something like this, whatever it is......

martin

After the first death, there is no other. (Dylan Thomas)

wide

It's not at all easy to calibrate an open area test site, and shielded enclosures are much more difficult to calibrate, since standing waves are inevitable unless you have the big bucks to make them semi-anechoic. However, relative measurements can be very useful, especially after you have some data from a calibrated measurement site. If you're located favorably (relatively quiet background spectrum), you might be able to get some useful relative measurements without really getting calibrated. Highly directional antennas can help (aim them away from strong background sources and toward the EUT), and you'll need a 20dB preamp with most spectrum analyzers. If you decide to build a Faraday enclosure, there are lots of ways to make them inexpensively. How you make the joints is more important that what metallic material you use. Some people use drywall coated with foil. We bought an old steel ocean cargo container and put some effort into making all its joints electrically connected. Paul Mathews

No argument there. Try this...

John

Comb generator approach is OK to the extent that you approximate an open area test site. Faraday enclosures can have quite strong resonances between comb peaks. Noise generators are the preferred signal source, since standing waves aren't such a problem. Paul Mathews

Some colleagues gave me a good tip.

They can do EMC testing with proper kit, but before you submit stuff for expensive testing you can detect most of the gross RF leaks just by waving a cheap AM radio around it.

Passing that test saves a wasted day of expensive EMC testing.

What if a get a Comb generator calibtated at a OATs test site.

Then I can do relative measurements regardless of antenna cable loss etc.

Meansure the comb freq / level and compare to the OATs test site results.

Provided the RF comb frequencies are up near the limits.

Then I can know weather to worry about passing the limits by a large safe margin etc.

Joseph

Expect to pay between $50 and $500 for a good broadband noise source on ebay. Quite often, the seller doesn't have specs, and it's buyer beware, but there are excellent values to be had. An ebay search on 'noise source' or 'noise generator' will turn up some possibilities:

Paul Mathews

I read in sci.electronics.design that Joe.G wrote (in ) about 'Home based EMC measurements', on Wed, 20 Apr 2005:

ARRL or RSGB Handbook (if possible, both). Look for 'noise diodes'.

--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
\'What is a Moebius strip?\'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

Paul Burke

Hi Paul, Are the suich things as wide band noise generators 10kHz to 1GHz for EMC testing.

I hadn't seen andy around.

Any pointers to wide band noise genators would be appreciated (kits / project and pro gear).

Regards Joseph

I find hot toddies warmer in winter, as the reefer doesn't really give out much heat.

A 20' container works nicely. Reefers are better, they are a lot warmer in winter.

Cheers Terry

Bruce Carsten's EMI sniffer probe is good there, too. Use with a tunable receiver, can be very cheap.

I have a mate who's company has an EMC test setup (alas no anechoic chamber) and thus have 4 hacks at it:

1) design for minimum EMI 2) sniffer probe tests. 3) use their test-setup, only $100/hr 4) then pay big bucks for the real deal.

Cheers Terry