Can I compare the Electromagnetic interference (EMI) of different DC-DC converters using software?

Thread Starter

timothyarora

Joined Sep 28, 2010
5
Please tell me whether can I compare the Electromagnetic interference (EMI) of different DC-DC converters using the software?
I know EMI depends on the PCB design. But if I leave this aspect, there are other factors that should be considered for EMI.
 

DickCappels

Joined Aug 21, 2008
6,534
Do you mean conducted radio frequency interference?

This is usually done by powering the power supply through a Line Impedance Stabilization Network, which gives a nearly constant impedance and provides a standarized output for your receiver. Usually the receiver is a spectrum analyzer, often connected to a computer that operates the analyzer and plots the result.

You might be able to get away with looking at the output of the LISN using an RF voltmeter and then noting the peak readings.
 

Thread Starter

timothyarora

Joined Sep 28, 2010
5
I'm looking for the EMI comparison of basic converters such as buck, boost, buck-boost etc. DC-DC converters have both common mode and differential mode noise. Yes, it is not related to radiated emission.
I'm looking to compare using software such as Matlab(simulink), PSIM.
 

WBahn

Joined Mar 31, 2012
26,068
You are going to have to have sufficiently good models of everything in the system to be able to compute the signals with sufficient fidelity to be useful. That is far, far from a trivial task. There are specialized modeling and analysis software packages out there, but they are usually very expensive with long learning curves.
 

Papabravo

Joined Feb 24, 2006
14,244
The theoretical answer is yes. The practical answer is to ask how much time and money you have at your disposal. As has been pointed out the trick is making the models sufficiently accurate to be useful. The answer to the modeling question depends on who designed and built these supplies. If it is an off the shelf supply that has been certified then the manufacturer probably has the information, but is unlikely to share it. If you designed and built the supply and it has not been certified then you probably don't have enough information to construct a model. In either case time and expense is yours to shoulder.
 

andrewmm

Joined Feb 25, 2011
532
Well. A few problems to be over come.

You are looking purely for a software comparison, so you are going to idealy be using a simulation package, such as LTspice.

basic idea is you model your converter in the package, then you model a LISN.

https://electronics.stackexchange.com/questions/413317/how-to-simulate-the-impedance-for-this-lisn-circuit-in-ltspice

https://forum.allaboutcircuits.com/threads/emi-filter-simulation-in-ltspice.142156/

https://www.analog.com/en/analog-dialogue/articles/speed-up-the-design-of-emi-filters-for-switch-mode-power-supplies.html

Now the BUT.

Like all simulations , they are just that, basically a model of the real thing.

The more realistic your model is, the more like the real thing , so you end up adding capacitors and inductors to model the parasitic in the system, such as the coils being above a ground plane, leakage to ground, et all.

Now how you define what these parasitics are, and their size, is the art side.
Knowing what to simulate as important for the question your asking.
 
Top