Site Attenuation Prediction for Refurbishing an Older EMC Chamber

The IBM RTP/065 EMC Chamber required a refurbishment in order to accommodate larger computer systems. One of the planned enhancements was to replace the existing carbon-loaded foam RF absorbers with a combination of ferrite tile and RF absorber. This report analyzes the site attenuation characteristics of the proposed chamber with supplier provided RF absorber loss data.

Information directly from vendor proposals showed simulations had indicated that the site attenuation values were right on the +/- 4 dB limit required by the EMC regulatory bodies. This work was intended to increase/decrease confidence in the likelihood that the new chamber would meet the necessary site attenuation limits. Due to business concerns, senior IBM management decided that there was no room for error and that the newly refurbished chamber must be operational on schedule.

While proposals and data from more than one supplier were included in the original analysis, only the chosen supplier results will be included here (since they ultimately won the contract with IBM). (The chosen supplier was ETS-Lindgren in Cedar Park, Texas. For additional information, please contact the supplier or the author at bruce.arch@ieee.org.)

Simulation Technique

The basic simulation technique used for this analysis is ray-tracing. It is assumed that the first reflection is the major reflection and there is one reflection from each wall. The amplitude and phase of the path includes the free-space loss/phase as well as the loss/phase from the absorber material. Supplier provided absorber information is used in all cases (with reflection being dependant on approach angle and frequency). Free-space loss is based on 1/r distance.

Site Attenuation Simulation

The requirements for site attenuation include the following:

1. All combinations of direct ray and reflected rays must have a maximum of no more than +/- 4 dB from theoretical (direct with one reflection from the metal floor) as the receive antenna is scanned in height from 1 meter to 4 meters above the metal floor.
2. The source antenna must be placed in five different locations on the turntable (center, along turntable edge nearest to receive antenna, along turntable edge furthest from receive antenna, and at remaining 90 degree offsets along turntable edge. The source antenna heights will be one meter and two meters for horizontal polarization and one meter and 1.5 meters for vertical polarization. The receive antenna must be maintained at 10 meter distance from the source antenna.

Site Configuration

Figure 1 shows the general site configuration for the enclosure. The non-rectangular shape potentially changes the site attenuation prediction from the suppliers, since their tools only operate with rectangular enclosures. The basic internal dimensions of the metal walls are 44 feet wide by 80 feet long. Also shown is the turntable size change from the previous 4.5 meter to 6 meter diameter. The ceiling height is 25 feet from the metal floor.

Figure 2 shows the chamber configuration with the larger turntable moved towards the left end by one foot (as proposed by the chosen supplier).

Antenna Locations for Site Attenuation

Figure 3 shows the chamber configuration with the five transmit antenna locations shown on the turntable for the case where the receive antenna is located along the center line of the chamber. The proposals from the suppliers only allowed for the center line position of the receive antenna. However, IBM typically uses the receive antennas offset, so there can be two antennas being used simultaneously (for 30-200 MHz and for 200-1000 MHz). This desired configuration for antenna placement is shown in Figure 4.

Site Attenuation Prediction with the Chosen Supplier’s Proposal

The proposal from the chosen supplier included ferrite tiles and cone absorber material. The three walls behind the receive antennas had 1 meter thick cones, and all other walls and the ceiling had 1.5 meter thick cones.

The refection loss from the supplier is shown in Figures 5 and 6 for the one meter cones and Figures 7 and 8 for the 1.5 meter cones.

The site attenuation requirement is to have the actual received signal be within +/- 4 dB of the theoretical value from a direct ray and one reflection from a perfect ‘ground’ plane. Figures 9 through 16 show the predicted site attenuation using the absorber reflection data from the chosen supplier, and assuming one reflection from each of the eight walls. Both centered and offset receive antenna positions are included. All cases meet the +/- 4 dB requirement with a comfortable margin.

Conclusion

All the simulation results showed a reasonable margin for the predicted site attenuation from both supplier proposals. While the supplier proposals were at the limit of +/- 4dB, it is believed that they were being overly conservative and adding extra margin in their predictions (since their simulation tools could not directly handle the non-rectangular shape of the room and the non center line placement of the receive antennas).

These predictions were based on the supplier provided data for their absorber materials. Both suppliers that bid the IBM project had experienced and well known engineers responsible for their data, and confidence was high that the data from both companies was accurate.

The final conclusion from this analysis was that there was a comfortable margin with the data from either supplier. The final decision on supplier selection was not dependant on the technical performance.

Once the project was completed, the actual testing showed the site attenuation test results met the required +/- 4 dB requirement with a comfortable margin.