0
\$\begingroup\$

I'm building a radiated noise generator. My intention is to expose my developed equipment to electromagnetic noise. I want to check the behavior of the equipment in noisy situations.

So I want to build something that emits large amounts of electromagnetic noise.

Maybe I should:

-> Make an inductor with a ferrite core, wound by myself;

-> Use an electrolytic capacitor;

-> Create a high frequency DC-DC converter;

-> Use a very high load, for example 5 Ohm.

The intention is that the high current/frequency of the inductor will produce high amounts of EMI.

Is my idea a good one? Do you have any suggestions?

Thank you very much


I didn't find any manual for irradiated EMC testing on the internet.

Is there any manual that, in a reasonably simple and inexpensive way, teaches how to test the reliability of an equipment for external noise?

\$\endgroup\$
6
  • 3
    \$\begingroup\$ There are standardised tests for EMC testing. The methods are carefully described so they can be repeated. Don’t try to re-invent the wheel with a random attempt. \$\endgroup\$
    – Kartman
    Jun 21, 2021 at 12:36
  • \$\begingroup\$ Could place your device near (but not within arcing range) of a small Tesla coil. That emits a wide range of frequencies, so could influence nearby equipment. Doesn't tell you much about how or why though - that's where a controlled EMC test is needed. \$\endgroup\$
    – rdtsc
    Jun 21, 2021 at 13:51
  • \$\begingroup\$ Search term: "radiated immunity". This will help you find more information. I have never gone through radiated immunity testing but I have seen the setup. They generally have a lab-type RF source and a transmit antenna. The antenna is held at a fixed distance from the device under test (DUT) inside an anechoic chamber. The anechoic chamber prevents the RF from escaping into the world at large where it might cause all kinds of trouble. \$\endgroup\$
    – user57037
    Jun 21, 2021 at 19:02
  • \$\begingroup\$ There are typically 4 formal terms used in EMC testing: radiated emissions, radiated susceptibility, conducted emissions and conducted susceptibility. Sometimes "susceptibility" may be referred to as immunity. Radiated susceptibility is what you wish to test in this case. \$\endgroup\$
    – Lundin
    Jun 22, 2021 at 9:24
  • \$\begingroup\$ Anyway, while coming up with all manner of fishy home-made pre-compliance tests may give you something useful, I would strongly recommend a design review focusing on EMC. Such questions would be on-topic here and there's lots of expertise available. Far from an EMC guru myself, but I've been through lots of formal & informal tests both. \$\endgroup\$
    – Lundin
    Jun 22, 2021 at 9:31

4 Answers 4

3
\$\begingroup\$

A chattering electromagnetic relay, with a 120 V ~ or a 240V ~ coil, would make a good EMI source.

enter image description here

The relay coil is to be wired in series with its NC contact.

\$\endgroup\$
2
  • \$\begingroup\$ Or use a big contactor used for machinery etc. They give very evil EMI, although it is mostly low frequency noise. \$\endgroup\$
    – Lundin
    Jun 22, 2021 at 9:09
  • \$\begingroup\$ Yes, I've used contactors too! \$\endgroup\$
    – vu2nan
    Jun 22, 2021 at 10:15
0
\$\begingroup\$

When testing equipment for susceptibility you normally would want to be able to precisely control the frequency and power level of the noise source. A DC DC converter typically has square wave voltage waveforms and ramping current waveforms. This is going to generate noise not only at the switching frequency, but also at the harmonics.

While you can use a DC DC converter for testing its not going to be a very precise test due to the multitude of frequencies. Also your inductor is not really optimized for radiating energy in a specific direction or power level. So short of measuring it, you probably have no idea what you are actually exposing your device to.

A much better option is to create an oscillator that runs at some specified frequency and connect it to an antenna (either bought or DIY). This would give you much better control over the power level, direction, and frequency of the signal.

If you are doing this test so that you can have some confidence that your equipment will pass radiated susceptibility testing, then the best route is to recreate the test environment as closely as possible. Normally you would do this by renting the proper antennas and signal generators from a test equipment supplier.

If you just want a cheap way to satisfy your curiosity regarding the behavior of your equipment when exposed to a DIY noise source then go ahead and build the oscillator.

Also be sure you are not doing this in an area where you could interfere with someone else's equipment.

\$\endgroup\$
1
  • \$\begingroup\$ Although it is useful to know the frequency + harmonics that caused failure, it's not always something one should stare blindly at. I've had EMC tests where they say something like "it failed at 1.3GHz". And then there is absolutely no relation between that frequency + harmonics to anything in my DUT. All I know is that something somewhere absorbs that energy and then that causes the failure. Good immunity designs blocks a broad frequency range and bad designs will be vulnerable to all manner of frequencies. \$\endgroup\$
    – Lundin
    Jun 22, 2021 at 9:18
0
\$\begingroup\$

It depends on the frequency range:

  1. radiated ESD can be done with a piezo lighter attaching on top of it a home-made marconi antenna (a small square cut from PCB plus a short monopole in the center of it; the two isolated one from the other)

  2. radiated high-frequency in the GHz range: take an old mobile with external antenna connector (my old Ericson and Siemens were like that) and connect an antenna of your choice (logperiodic, Yagi, ...). This is modulated field at 1 single frequency (ok, more than one in a narrow band) at about 900 MHz.

  3. radiated high-frequency in the hundreds MHz range: do the same as 2) but with a VHF or short-range radio (433 MHz). If you find a 868 MHz short-range radio, you can replace test 2, in case it is harder to find an old telephone like that. These radio should have detachable antenna for better flexibility (you can increase the EIRP by a factor of 3-4 easily), otherwise you just use them as they are in close proximity to your equipment. (intensity can be calculated with far-field assumption, antenna gain approx, tx power down to about 15-20 cm.

  4. radiated medium-frequency: better magnetic field so a sort of coil, but a few turns, otherwise field will change dramatically because of resonances and stationary waves along the coil wire; some turns may give directivity, but should be spaced to reduce stray cap and increase resonant frequency (e.g. spacing of 5 mm, wire diameter 1 mm). How to generate the feeding signal: the chatter relay above @vu2nan or a square wave generator. The generator can be just a switching relay between two voltages of a battery with resistor and the coil in the middle, or the same but replacing the relay with a MOS full bridge (4 MOS) driven by a 555 or so (plus bit of gate driving).

And you cover almost everything from a hundred kHz to GHz. Of course I have omitted the details otherwise it would have been much longer!

\$\endgroup\$
0
\$\begingroup\$

If you want a brutal broad spectrum source of EMI buy a Tesla coil.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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