# Radiated Emission EMC

#### uppi_777

Joined Feb 26, 2018
60
i had encountered few things While doing EMC RE Measurements.
i have seen emissions (Amplitude dBµV/m) emitting at particular frequency (lets say at 750 and 800MHz) therefore i calculated the single frequency which is usually referred to as the fundamental frequency or sometimes the first harmonic.
So for example my fundamental frequency of clock oscillator is 50MHz (measured 40dBµV/m) 15th harmonic is 750MHz (50dBµV/m) is and 16th Harmonic is 800MHz (dBµV/m)

So now i am pretty sure from which part the emission is coming from DUT but my question is Why the emissions are low at 50MHz and high at 750 and 800MHz if i assume its coming from a single component why the emissions are different at different frequencies

#### MisterBill2

Joined Jan 23, 2018
14,760
The initial frequency is only one part of the process of radiated emissions. There are assorted digital devices operating at the initial frequency, and they have very rapid transition times, which create harmonic frequency energy. And different portions of the conductors will resonate at the higher frequencies. So that is why there are harmonics that are emitted.

#### WBahn

Joined Mar 31, 2012
28,517
My first suspicion is also that your issue is coming from parts that have fast edge transitions. Try slowing them down with series resistors or use ferrite beads. Also, look for parasitic image loops in your traces that cause them to radiate and try to reduce their size.

#### uppi_777

Joined Feb 26, 2018
60
The initial frequency is only one part of the process of radiated emissions. There are assorted digital devices operating at the initial frequency, and they have very rapid transition times, which create harmonic frequency energy. And different portions of the conductors will resonate at the higher frequencies. So that is why there are harmonics that are emitted.
My question is why there is radiated field strength low at lower frequency 50MHz and radiated field strength at higher frequency 800MHz if a single component is emitting the radiation?

#### MisterBill2

Joined Jan 23, 2018
14,760
The very fast transitions generate harmonics of the repetition frequency. Those harmonics radiate while the initial frequency is a much longer wavelength that does not radiate so well. AND the active devices producing the harmonics deliver more power with the faster transitions.

#### KeithWalker

Joined Jul 10, 2017
2,804
The strength of radiation at different frequencies depends on the geometry of whatever is doing the radiating. It can act as a good radiating antenna at one frequency but not so good at others.

#### uppi_777

Joined Feb 26, 2018
60
The very fast transitions generate harmonics of the repetition frequency. Those harmonics radiate while the initial frequency is a much longer wavelength that does not radiate so well. AND the active devices producing the harmonics deliver more power with the faster transitions.
What do you mean by faster transitions?

#### WBahn

Joined Mar 31, 2012
28,517
What do you mean by faster transitions?
Are you familiar with Fourier series?

Any waveform can be expressed as the sum of harmonically-related sinusoids with different amplitudes and phases. It would require an infinite number of them to truly replicate the waveform, but if we use enough of them, we can get as close as we want.

Fast changes in a waveform require the higher-frequency harmonics to have larger amplitudes.

When you have logic that has an output change, if that change occurs over, say, 1 ns, it will have greater content at higher frequencies than if it changes over 100 ns. Higher frequencies are also generally radiated more easily from a circuit board because the parasitic antennas are on a scale that are better matched. This is why techniques called "edge-rate limiting" are often used to combat these EMI issues.

#### uppi_777

Joined Feb 26, 2018
60
Are you familiar with Fourier series?

Any waveform can be expressed as the sum of harmonically-related sinusoids with different amplitudes and phases. It would require an infinite number of them to truly replicate the waveform, but if we use enough of them, we can get as close as we want.

Fast changes in a waveform require the higher-frequency harmonics to have larger amplitudes.

When you have logic that has an output change, if that change occurs over, say, 1 ns, it will have greater content at higher frequencies than if it changes over 100 ns. Higher frequencies are also generally radiated more easily from a circuit board because the parasitic antennas are on a scale that are better matched. This is why tec
The very fast transitions generate harmonics of the repetition frequency. Those harmonics radiate while the initial frequency is a much longer wavelength that does not radiate so well. AND the active devices producing the harmonics deliver more power with the faster transitions.
hniques called "edge-rate limiting" are often used to combat these EMI issues.
The very fast transitions generate harmonics of the repetition frequency. Those harmonics radiate while the initial frequency is a much longer wavelength that does not radiate so well. AND the active devices producing the harmonics deliver more power with the faster transitions.
it’s needed an antenna of around 17 cm. If there is a cable of this length the above statement would be applicable

#### DickCappels

Joined Aug 21, 2008
9,877
Note that the radiation efficiency of a short wire is much better at high frequencies that at lower frequencies. I am sure your case is more complicated than that but radiation efficiency is part of the explanation.

What sort of device are you testing?

#### WBahn

Joined Mar 31, 2012
28,517
it’s needed an antenna of around 17 cm. If there is a cable of this length the above statement would be applicable
You don't need a half-wavelength dipole in order to radiate.

#### Janis59

Joined Aug 21, 2017
1,639

#### Janis59

Joined Aug 21, 2017
1,639
RE:""You don't need a half-wavelength dipole in order to radiate.""
Example N1: Antenna with some z=50 Ohm. Then N=U^2/R=i^2*R thus the 10 Volts give the 2 W and 0.2 Amperes.
Example N2: Non-antenna, just the 25 mm coil on air core some 20 mm long, just 4 turns with 3 mm wire. Circulating power Voltage 5000 Volts, current 20 Amperes. In the center air is becoming ionized and shines brightly. Whilst just the 5 cm off the coil there is E=500 V/m, 10 cm off is 50; half meter away 5 V/m; meter away 1 V/m and so on. Thus, if the antenna geometry is violated, the field intensity decreases very harsh, whilst if antenna geometry is observed, the field is going to radiate very far away.

#### MisterBill2

Joined Jan 23, 2018
14,760
ALL antennas radiate SOME. Resonant antennas are usually the best, but any element will radiate SOME. In addition, unseen resonance from coupling between different conductive elements may exist.
And any conductor carrying a current is an antenna to some extent.

#### Janis59

Joined Aug 21, 2017
1,639
ALL antennas radiate SOME. Resonant antennas are usually the best, but any element will radiate SOME. In addition, unseen resonance from coupling between different conductive elements may exist.
And any conductor carrying a current is an antenna to some extent.
That is why I posted some comparative figures to understand HOW MUCH they may differ.