5

I want to wirelessly communicate over low frequencies (<1kHz), and I noticed during some tests that 60Hz noise is showing up a bit at 180Hz (3rd harmonic), and also slightly at 300Hz (5th harmonic): (Ignore the stronger signal I was transmitting)

enter image description here

Why are only the odd harmonics of 60Hz noise visible?

Why are the harmonics of the signal I am transmitting not visible (e.g. 3rd harmonic of 125Hz signal is 375Hz)?

More generally, what frequencies should I assume to be occupied when transmitting in this low frequency range? I originally assumed 50/60Hz and all their harmonics (i.e. 100/120Hz, 150/180Hz, 200/240Hz, etc) should be avoided. Do I not have to worry about even harmonics? Are there other low frequencies commonly occupied in an in-home environment?

Edit Thanks for the answers. I just wanted to add that looking through a few papers on harmonic noise from appliances, it seems a foregone conclusion that only odd harmonics are significant. Here is a diagram showing some slight even harmonics (source):

enter image description here

abc
  • 619
  • 6
  • 21
  • I see a similar question here: https://van.physics.illinois.edu/qa/listing.php?id=572. Perhaps what I am experiencing is due to an active DC bench supply (from AC wall power) in the vicinity. – abc Jun 29 '17 at 16:06
  • 1
    What is your medium of communication - this makes a difference. What form do your carriers take i.e. are they sinusoidal? When data changes do you seemlessly move from one carrier frequency to another? – Andy aka Jun 29 '17 at 16:08
  • The even harmonics cancel each other out through de-constructive interference, this is why only the odd harmonics are present. – circuitpatrol Jun 29 '17 at 16:11
  • @circuitpatrol maybe you'd like to justify that? – Andy aka Jun 29 '17 at 16:12
  • @Andyaka I am communicating through air (wireless) using pure sinusoids via a magnetic transmitter and receiver (i.e. near-field). – abc Jun 29 '17 at 16:12
  • @circuitpatrol From what I have read recently, it sounds like harmonics are generated from loads depending on their AC response, and therefore even or odd harmonics could be present. See the link in the comment above. – abc Jun 29 '17 at 16:13
  • @Andyaka I don't believe the frequency shifting effects the harmonics, but after your comment I looked at the spectrogram more closely and realized that the 60Hz+harmonics only show up when the signal is being transmitted (you can see where it starts and stops). Therefore, I believe it is due the bench power supply (which powers the transmitter) and it's response to AC power. – abc Jun 29 '17 at 16:18
  • Is the power supply only switched on when transmitting or is the power supply always switched on and you disable transmissions some other (more passive) way? – Andy aka Jun 29 '17 at 16:30
  • @Andyaka The former: it is only switched on when transmitting – abc Jun 29 '17 at 16:31
  • 1
    Then that is your likely source of 60 Hz +plus odd harmonics - the internal transformer that steps down the voltage to the supply output circuit is slightly saturating (pretty normal) and producing odd harmonics. For my information, why are you using such low frequency carrier frequencies? – Andy aka Jun 29 '17 at 16:33
  • 1
    Even more significant than the transformer, the bridge rectifier is a strong source of (odd-)harmonic distortion -- it conducts only in narrow pulses at the peaks of the voltage waveform. A power supply with PFC (active power factor correction) would eliminate this. – Dave Tweed Jun 29 '17 at 16:46

2 Answers2

8

I am communicating through air (wireless) using pure sinusoids via a magnetic transmitter and receiver (i.e. near-field)

Chances are that the interference you receive is 60 Hz AND odd harmonics. This statement is based on the fact that the current taken by localized circuits (such as transformers) will be somewhat rich in odd harmonics due to core saturation.

Most common powerful loads/components such as transformers and induction motors tend to take a symmetrical (same shape for positive and negative half cycles) but distorted current. When the waveform is symmetrical there tends to be only odd harmonics. Taken to extremes a square wave only has odd harmonics: -

enter image description here

Why are the harmonics of the signal I am transmitting not visible (e.g. 3rd harmonic of 125Hz signal is 375Hz)?

Because you are transmitting pure sinusoids and they don't have harmonics. Of course if you don't change frequency seemlessly then you'll see some harmonics creeping in.

Andy aka
  • 434,556
  • 28
  • 351
  • 777
5

Why are only the odd harmonics of 60Hz noise visible?

Most loads that generate distortion on the power line are symmetrical with respect to voltage, so they produce only odd harmonics (e.g., full-wave rectifiers). The only way to get even harmonics is to have asymmetric distortion (e.g., half-wave rectifiers).

Why are the harmonics of the signal I am transmitting not visible (e.g. 3rd harmonic of 125Hz signal is 375Hz)?

If you are transmitting low-level sine waves, they probably don't get distorted by those loads in the same way, if at all.

More generally, what frequencies should I assume to be occupied when transmitting in this low frequency range? I originally assumed 50/60Hz and all their harmonics (i.e. 100/120Hz, 150/180Hz, 200/240Hz, etc) should be avoided. Do I not have to worry about even harmonics? Are there other low frequencies commonly occupied in an in-home environment?

All of them? In addition to harmonic distortion of the power wave itself, there are many loads (such as universal motors) that generate broad-spectrum noise that has no relationship to the power line frequency.

The power line in general is a nasty environment for communications. That's why most systems use ultrasonic frequencies (100 kHz and up), and sometimes only transmit close to the power wave's zero crossings, where the noise tends to be less.

Dave Tweed
  • 168,369
  • 17
  • 228
  • 393