0
\$\begingroup\$

I was wondering if low frequency EMI can occur with PWM or other sub MHz signals. I am aware that harmonics of these signals can generate EMI but I was wondering if the fundamental frequency can radiate as well ?

The reason I am unsure is because at sub MHz frequencies the wavelengths are very long (over 300m), so PCB traces and wires of a few cm at most should be able to transmit them.

In most cases, am I right in assuming that radiated emissions are more common in the higher frequencies ? I have seen online that even radiated emissions testing begins at around 30MHz.

\$\endgroup\$
5
  • 2
    \$\begingroup\$ It can. It's just whether or not it causes a problem due to transmission and reception efficiency. It's not really interference until it interferes... \$\endgroup\$
    – DKNguyen
    Commented Nov 26, 2019 at 21:05
  • \$\begingroup\$ I understand that it depends, but you're saying that very long wavelength signals can radiate from "small" PCBs ? Do you think that higher harmonics can radiate with more power than lower harmonics due to their wavelength being closer to the traces / wires ? \$\endgroup\$
    – Ryan
    Commented Nov 26, 2019 at 21:15
  • \$\begingroup\$ That's a good question. I've never given that much thought and I'm not an RF guy. \$\endgroup\$
    – DKNguyen
    Commented Nov 26, 2019 at 21:16
  • 1
    \$\begingroup\$ @Ryan, they will radiate. The question is how much. If the antenna dimensions are much smaller than the wavelength, then the answer is "not much at all". Yes, it's likely that higher harmonics will radiate with more power than the fundamental in your scenario. \$\endgroup\$
    – The Photon
    Commented Nov 26, 2019 at 21:19
  • 1
    \$\begingroup\$ LF EMI due to magnetic emissions <300kHz and high current \$\endgroup\$ Commented Nov 26, 2019 at 21:23

1 Answer 1

1
\$\begingroup\$

Yes low-frequency EMI emissions are a thing and very common in switching power supplies designs. Most of the energy is carried in the edges of the switching supply and will couple and propagate all over your board/assembly if not implemented right.

From working in the automotive industry, I've encountered many of these designs creating problems in EMC chambers. It can be both conducted, most of often through the power or external harness, or emitted radiation. The fundamental can show up along with harmonics, sometimes creating broad-spectrum emissions (frequency "humps"). When in the 100's of kHz range, these are bad for AM radio. We prevented headaches picking a part out of the AM frequencies range but even then, if the switching edges are super sharp, you'll get very harmonics. The shape of the switching waveform can be controlled by the low and high-side driver drive strength among other things. Yes you will make you switcher design slightly less efficient but you will get yourself a pass at the EMC chamber. Be sure not to bring cross-conduction in the picture else this makes everything worse ;)

\$\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.