yorkemc.co.uk » Consultancy & Research » EMC Research »

Feasibility study into the application of numerical modelling in EMC standardisation radio emission measurement methods

A report for OFCOM by York EMC Services Ltd, University of York

Authors: D W Welsh, D Bozec & M R Tyndall


Summary

YES (York EMC Services Ltd) have performed a feasibility study regarding the application of numerical modelling techniques to reduce measurement time for EMC tests. The most promising area for achieving a reduction in measurement time was the radiated interference measurement above 1GHz. Two modelling methods were investigated, a full wave method (TLM) and a point source model. Normally full wave methods require a significant amount of effort to give useful results so any reduction in test time tends to be negated by the time spent in devising the model. It was anticipated that the point source model might give sufficient accuracy with relatively little time spent in the modelling process.

At GHz frequencies the field plot of emissions from an EUT is highly complex and the only sure way to determine the peak emissions is to scan hemi-spherically with sub-wavelength measurement intervals. The hemi-spherical scan is impractical as it is hard to arrange and also very time consuming as thousands of points on the hemisphere must be measured and then analysed over a large frequency range. Exact modelling of this highly complex radiation structure is impossible, as even small deviations in the construction of the EUT will have a significant effect on the positioning of the lobes of emission. However, the statistics of the modelled emission are likely to be similar and these can be used to reduce the measurement effort.

The point source method has been used extensively in this project as it is quick and simple to use, unfortunately it did not accurately model the statistics of the emissions from the EUT under consideration. It is believed the method could be sufficiently accurate but a better knowledge of the input parameters is required to get useable results. It was demonstrated that the TLM method was a practical proposition up to at least 6GHz and inspection of the field plots suggests that the statistics are likely to be correct.

The main conclusion of this report is that a significant saving of test time could be made by applying one of the following approaches to radiated emissions testing above 1GHz:

  1. Fit to known probability distribution. With further development this approach could yield a very accurate way of establishing the maximum field strength radiated from an EUT.
  2. Under-sampled data set. The report has demonstrated that only a few measurements need to be taken to determine a field strength that is in the top 5% of field strengths to be radiated from the EUT. The actual maximum field strength is not measured or predicted but it is expected that with further research an estimate could be made of what it would not exceed.
  3. Reverberation chamber method. Using this approach the traditional EMC measure of field strength would be replaced by a total radiated power measurement. Limits would be defined in terms of total radiated power.

The first two methods would use anechoic chambers and turntables, these are currently readily available in UKAS accredited EMC laboratories. However, the anechoic performance at GHz frequencies for some chambers would be an issue that would need to be addressed. Reverberation chambers are not widely used so development of this approach would require significant investment by EMC laboratories.

Downloads:

You can download this report as a PDF file (6,519 kB).

You can also download the paper given at EMC York 2004 as a PDF file (1,136 kB).


Last updated: 2004-Sep-10

Top

Home