Ferrite Cores on the output cable of a VFD

[Dinoduck94]

TL;DR Summary

How to size a Ferrite core on the outgoing cable of a Variable Frequency Drive

Some Variable Frequency Drive manufacturers (such as Allen Bradley) supply a toroidal Ferrite core to wrap the outgoing cables around to reduce EMF.

My question is, if a company doesn't supply a ferrite core, how do I work out if one is needed, and how do I correctly size it?

I understand that the switching frequency can be set to either 2, 4, 8 or 16kHz - so would this be the frequency that I'd use? If so, how come a lot of ferrites seem to operate in the MHz range?

Thanks

 

[Baluncore]

... to reduce EMF. ...

I think you mean EMI.

... how do I work out if one is needed, and how do I correctly size it?

There may be one inside the VFD. You do not need one if there is not an EMI problem, are you looking for more problems?
The manufacturer may be required by the FCC in USA, to take some action to reduce EMI. Providing a core with the device satisfies that legal requirement. An external core is usually not needed.

If so, how come a lot of ferrites seem to operate in the MHz range?

An 8 kHz square wave has harmonics that go all the way up to MHz.
Since all wires pass through the ferrite toroid in the same direction, it is being used to remove common mode noise, the power currents are cancelled. The RF EMI is at MHz frequencies so it requires a high frequency ferrite rather than an iron powder core.

 

[berkeman]

My question is, if a company doesn't supply a ferrite core, how do I work out if one is needed, and how do I correctly size it?

I agree with @Baluncore -- as long as the VFD has a CE Mark on it and an EMI test certification (most likely to "Level A" for industrial equipment), and as long as you currently do not have radiated EMI issues with the other equipment in your installation, I wouldn't worry about it.

If you *are* having EMI issues that are related to the VFD (they stop when you turn off that drive), then check with the manufacturer to see what they recommend. If this is an off-brand VFD with no obvious customer support available, then start with standard EMI ferrite cores (maybe Material 43 or lower frequency). You can order design kits from the ferrite core manufacturers for doing such investigative work. Take a look at the ferrite kits from these manufacturers, for example:

https://www.heilind.com/products/steward/brochure_ferrite_emi_cores.pdf

https://www.fair-rite.com/

 

[Dinoduck94]

Thanks all for your responses.

as long as the VFD has a CE Mark on it and an EMI test certification (most likely to "Level A" for industrial equipment), and as long as you currently do not have radiated EMI issues with the other equipment in your installation, I wouldn't worry about it.

Agreed that it generally isn't something to worry about, but I don't want to be an engineer that doesn't understand EMI and only looks to solve an issue once it has been experienced. I want to be proactive and to know when I am likely to have an EMI issue, and know how best to resolve it - hence me looking into Ferrite cores.

An 8 kHz square wave has harmonics that go all the way up to MHz.

When you say that the harmonics go all the way up to MHz, how do I find out what the harmonic frequency of an VFD is? Is the a general rule of thumb?
If I'm to size a ferrite core then I assume it's this harmonic frequency that I would need to use in any calculation?

You can order design kits from the ferrite core manufacturers for doing such investigative work. Take a look at the ferrite kits from these manufacturers, for example:

https://www.heilind.com/products/steward/brochure_ferrite_emi_cores.pdf

https://www.fair-rite.com/

I will have a look at these ferrite design kits - thank you.

 

[Baluncore]

When you say that the harmonics go all the way up to MHz, how do I find out what the harmonic frequency of an VFD is? Is the a general rule of thumb?

A square wave has odd harmonics. An 8 kHz fundamental sinewave has harmonics of
H3 = 3*8 kHz = 24 kHz;
H5 = 5*8 kHz = 40 kHz;
H7 = 7*8 kHz = 56 kHz; … all the way to;
H125 = 125*8 kHz = 1 MHz; … and above that.
https://en.wikipedia.org/wiki/Square_wave#Fourier_analysis
The relative amplitude of the n'th harmonic will be 1/n; so the 3'rd harmonic will have an amplitude of 1/3; and the 125'th harmonic will have a relative amplitude of 1/125 = 0.008 = 0.8% = 42 dB down.

If I'm to size a ferrite core then I assume it's this harmonic frequency that I would need to use in any calculation?

The high frequency edges of the internal switching may escape along the output cable. Most will be significantly attenuated by filters in the VFD. You will need to identify if and where in the spectrum EMI is actually a problem.

For a VFD switching at 8 kHz, and using an AM radio, you should be able to hear the odd harmonics every 16 kHz across the band. There will also be other harmonic noise from other nearby switching supplies. Turn the VFD off and on again to confirm the VFD is the source of the EMI.