Better EMI filtering/suppression for high power electronics? 2

[79BirdofPrey]

I work for a company that designs and manufactures high power electronics for power conversion/regulation in aircraft. My question was regarding EMI; I wanted to know if there are new, more efficient methods for EMI filtering/suppression. Typically our filtering methods involve adding inductance and capacitance as needed, as well as adding ferrite components as shielding. Weight is a critical point in aviation, so if there are lighter and smaller components/methods, then I'd like to implement them.


Thank you,

Joseph

 

[Noway2]

1) If you are filtering on the input to the circuit, instead of using a passive filter you could use a power factor correcting front end (rectifier).
2) I assume you are shielding the entire box. If this is the case you may be able to lighten the shield by virtue of the fact that it doesn't need to be solid, but this will require knowledge of the frequency your trying to filter. Think of the shield on the door of your microwave as an example.
3) On the output, you will probably still require a transformer. An Al transformer would be lighter than a Cu one.
4) If increase the switching frequency of your system you can make the magnetic components smaller, but this also increases the losses.

 

[VEBill]

Spread spectrum clocks are sometimes used to 'smear out' the resultant noise in the frequency domain. It's arguably an evil trick.

 

[79BirdofPrey]

Hi all,

Would a spread spectrum clock pertain to power electronics for aircraft? I thought that was primarily for telecommunications.

Also, are there any innovative, new ways to handle EMI filtering? We utilize all of the aforementioned methods already, and I'm looking for somthing better.

 

[Noway2]

I am not aware of any new ways to handle EMI. The fundamentals still strongly apply.
[ul]
[li] Curtail it at the source and keep it from leaving your system[/li]
[li] a wide, flat surface will be a lower impedance (inductance) channel than a wire[/li]
[li] don't inadvertently create an antenna.
[li] proper grounding is very important[/li]
[li] high speed signals (as determined by rise/fall time will follow the path of least inductance[/li]
[li] EMI susceptibility is the flip side of the same coin as emissivity. Prevent the latter and you avoid the former[/li]
[/ul]

Two great book resources about EMI and noise would be Henry Ott and Howard Johnson's books. Both of these have helped me solve real world EMI and noise issues in an analytical fashion and avoid the guessing games that constitute the normal approach. One thing to remember is that EMI control almost always has to be designed in and it is often times extremely difficult and expensive, if not impossible to correct an EMI issue on a bad design.

http://www.amazon.com/Reduction-Techniques-Electronic-Systems-Edition/dp/0471850683

and

http://www.amazon.com/High-Speed-Di...peed+digital+design+a+handbook+of+black+magic

 

[VEBill]

Google terms like: spead spectrum clock EMI

It's not directly related to the communications technique that shares the name, but there's some common ground in the initial concept of getting away from using a pure carrier/clock.

 

[VEBill]

By the way, much of EMI control is the exact opposite of a good antenna.

 

[MagicSmoker]

79BirdofPrey: one very effective means of reducing EMI without reducing efficiency is to use quasi-resonant or zero voltage/zero current switching. Another effective means is interleaving. Both of these require substantial changes to the design and, as usual, TANSTAAFL. Interleaving - as of a boost converter for PFC or a buck converter for output regulation - can be extremely effective both at reducing EMI as well as reducing overall weight (particularly in the size of the filtering capacitance).

Otherwise, I agree with the above sentiment that it is best to use the physical design of the converter to preclude the emission, if not the formation, of EMI in the first place.

 

[Sparweb]

I remember reading military handbooks about EM shielding and filtering (two sides of the same coin).
Detailed design guides with example analysis and data for various shielding materials, apertures, at various frequencies....
You can search and download MIL-HDBK's from the US army Assist webpage.


STF

 

[Skogsgurra]

For drive systems, and they are a lot like other power electronic systems, there are a few soft-switching techniques that can be used to reduce EMI down to very low levels. Google Switch Circuit and the likes.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

..or read here:

http://gke.org/pub/files/The Switch Circuit.pdf

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.