ABSTRACT
The EUROTEM® is a patented compact symmetrical TEM device with 4 striplines, which are surrounded by a fully absorber-lined ferrite enclosure. Due to the broad band nature of the device, it is well suited for radiated emissions and immunity testing, compliant with the standards. The EUROTEM® devices are scaleable and can be built in different sizes. The range of test objects starts at printed circuit boards all the way up to two metre high racks. For large test objects, the arrangement can be fitted into an existing fully absorber-lined room. This arrangement is called EUROTEM®-Antenna. By this retro-fit, the conventional antenna mast and the typical EMC antennas are substituted with the EUROTEM® antenna leading to savings in driving power and consequently in considerable cost savings for the amplifier. In contrast to the 10 year old GTEM technology, the EUROTEM® is equipped with polarisation switching. The test object will be put on a turntable in the same manner as in any open area test site or absorber chamber. There is no rotation through the three spatial axes which is required using a GTEM cell and the corresponding manipulator. The EUROTEM® is superior to the conventional technology by showing better field quality, better suppression of the non-TEM field components and in particular, the better cost efficiency factor.
1.INTRODUCTION
The implementation of the EMC Act, 1996 has induced a world-wide increase in demand for measurement cells and test facilities of various kinds [1-15]. The corresponding manufacturers claim these facilities to be fully compliant or at least suitable for measurements during the development phase of the product. A more detailed analysis however leads to deficiencies in clearly describing the field quality. It is a questionable argument to claim repeatability and relative measurement results to be sufficient for development testing. The danger is to operate with too high tolerances with respect to the final compliance test of the various devices under test. There is also another question to be answered: Is it really necessary to test fairly small objects like an ordinary telephone set in huge anechoic chambers with a test distance of 10 metres?
IEC and CISPR standardisation committees increasingly realise the existence of historically-grown deficiencies and contradictions regarding the measurement procedures. This in particular true for the sector of Information Technology.
The main interest for radiated emission testing today is the frequency range from 26 MHz to about 2 GHz. This was one of the main development goals of the EUROTEM®. Additionally the aim was to simplify the operation, to improve the field quality and to reduce the investment cost, not only of the EUROTEM®, but rather including the additionally required test equipment. In contrast to the OATS, there is less measurement time required because there is no change in the antennas and no antenna height scan.
Top
2. The EUROTEM® Family
Presently the EUROTEM® 2 (figure 1) is the smallest member of the family with a typical test volume of 0.35 cm x 0.35 cm. In addition to the basic set-up with AC mains and DC filters, as well as fibre-optic feed-throughs and several connectors such as BNC, N and RS232, an electromagnetically transparent turntable with a typical load of 20 kg can be retro-fitted. The shielding effectiveness of the device reaches 100 dB above 150 kHz. Switching of the polarisation of the electromagnetic field is performed by rotating the balun enclosure against the 4 inner striplines. The biggest member of the EUROTEM® family is the EUROTEM® antenna. This arrangement is shown in figure 2, demonstrating the way to fit it into an existing fully absorber-lined ferrite chamber with the dimensions 7 x 4 x 3 metres.
Figure 1: EUROTEM® 2
Figure 2: EUROTEM®Antenna with 2m Test Object on the Turntable
|
Figure 2 shows our antenna device in an early design stage. Experience taught us to fit some foam absorber lining in front of the termination of the device. This greatly improved the field quality. To ease the access to the test object, the EUROTEM® antenna can be fitted into the absorber room with its rear end facing the door. In this area in particular, the quality of the absorbers has to be high because there is a stronger interaction with the back wall in comparison to the side walls. The electromagnetic fields are mainly concentrated in the space defined by the four outer striplines. The field strength decreases rapidly moving aside from these lines. This relatively small stray field could be used to further reduce the absorber lining of the side walls and consequently saving absorber cost.
Top
In front of the back wall, which connects the four resistor termination modules, there are additional foam absorbers fitted.
One of the main advantages of the four stripline arrangement is the freedom to route the supply cables of the EUT. In contrast to the GTEM cell for example, there is no need to penetrate metallic walls, which are not ferrite lined. Normally, the lines are fed through the centre opening of the turntable. By definition, the maximum coupling length of the connecting cables which are subjected to field illumination are naturally dependant on the size of the TEM device. Looking at the EUROTEM® antenna, this is not a problem because the cables are exactly configured in the same way as on any OATS by using an 80 cm high electromagnetically transparent Styrofoam support on the turntable. Another nice feature, due to the perfect symmetry of the device, is the ability to route cables along the longitudinal symmetry axis through the centre of the back wall. Ideally this should not lead to any cable coupling, in reality however, there is a small curvature of the E & H field due to a spherical wave propagation within the TEM device. If this poses a problem, for example in very precise calibration measurements, compensation by a di-electric lins could be introduced.
The iso-field lines of a cross-section of the EUROTEM®-Antenna with ± 1 dB and ± 3 dB respectively test area are shown in figure 6. The ± 1 dB area of the EUROTEM®-Antenna is 72 x 72 cm. To compare this to the GTEM 1750, we refer to figure 7.
Figure 6: Iso-field Strength Lines in EUROTEM®-Antenna (strip line length projection 4.8m)
Figure 7: Iso-field Strength Lines in GTEM 1750 (total length 8m) |
Top
4. Balun
A broad band transformer will introduce symmetry and then transform the 50 Ohm cable impedance to the 200 Ohm impedance of the striplines. This 200 Ohm stripline impedance is created connecting 1 pair of lines in parallel as demonstrated in figure 3. In addition to a high return loss, which means good impedance matching (figure 8), a low insertion transmission loss (figure 9) is required.
 Figure 8: Good Return Loss of the Balun (20dB equals 1% reflected power, VSWR 1.2)
Figure 9: Insertion Loss of the Balun for Both Output Ports (1dB equals 20% performance loss by heat) |
In an endurance test of four hours, the balun was subjected to 100 Watts RF power at 1 GHz from a 50 Ohm amplifier. The resulting rise in temperature did not exceed human body temperature. In order to perform this test, two baluns were connected in the back-to-back mode (200 Ohm/200 Ohm) so that the measurement impedance at each end was 50 Ohms.
Other balun designs have been created pushing the upper and lower frequency. A major success was achieved by improving the common mode rejection (figure 10). Further baluns have been optimised for the low frequency end resulting in 150 kHz.
Figure 10: Common Mode Rejection of a Balun |
Top
5. Power Consumption of the EUROTEM® to Generate EM Test Fields.
For immunity testing between 27 MHz and 1 GHz (lately 80 MHz - 2 GHz), the standard usually requires a three metre test distance between the tip of the antenna and the face of the test object. To generate 10 V/m modulated with 1 kHz 80% modulation, amplifiers of up to 500 Watts are needed for the low frequency range, for example, 60 MHz. These very high powers are required because of the relatively bad matching between the 50 Ohm output impedance and the transmitting antenna which is physically too small compared to the wavelengths. An already optimised version of such a broad band antenna is the X-Wing BiLog manufactured by Chase, model number CBL 6140. Figure 11 compares the drive power of this antenna to our EUROTEM® antenna. It is worth mentioning the distance from the feed point of our antenna to the placement of the EUT is about 3 metres. The high efficiency of the EUROTEM® antenna however, is directly related to the mode of well-controlled field propagation within the stripline.
The BiLog antenna does not have this propagation mode but rather, uses the conventional type of radiation by an antenna. In order to improve such a conventional antenna system, the size of the antenna should electrically reach half a wavelength. This makes it rather impractical at 30 MHz because five metres space is simply not available in small anechoic chambers. This requirement will hold true for vertical as well as horizontal polarisation of the antenna.
Figure 11: Driving Power to Generate 10 V/m Modulated in 3m Distance
|
The difference in power consumption according to EN 61000-4-3 modulated and unmodulated signal is the factor 3.24 which equals 5.1 dB. For calibration, the unmodulated signal is being used. For the EUROTEM® 2, this results in a drive power for the unmodulated signal of about 1 Watt to generate 10 V/m in the test volume (figure 12). 1 Watt equals +30 dBm in a 50 Ohm system. Figure 13 shows the according drive power for the EUROTEM®-Antenna. It is important to consider the larger test volume. Trippeling the stripline distance results in 10 dB higher power requirements for the same field strength of 10 V/m.
Figure 12: Drive Power of the EUROTEM® 2
to Generate 10V/m Unmodulated in the Test Volume
Figure 13: Drive Power of the EUROTEM®-
Antenna (Prototype) to Generate 10V/m Unmodu- lated
in the Test Volume (conducting frame of the termination not fully covered with foam absorbers, therefore reflections appear above 400 MHz) |
Top
6. Comparison of the Field Quality with the GTEM 1750
Using a design model comparable to EUROTEM®-Antenna, we utilised a Lithium-Niobate E-field sensor to prove the good non-TEM component suppression with respect to the working or mainfield component (figure 14). Aside from the 20 - 30 dB non-TEM component suppression, a relatively smooth frequency characteristic of the main component becomes evident. This is one of the main differences to the GTEM 1750 as demonstrated in figure 15.
Figure 14: Decoupling Characteristics of the Main and Side Components in a EUROTEM® (field sensor only specified up to 1 GHz)
Figure 15: Decoupling Characteristics of the Main and Side Components in a GTEM 1750 |
The main component in the GTEM 1750 exhibits a variation of approximately 5 dB. This hints at a coupling into higher order modes which results from imperfect absorber terminal coverage. In the EUROTEM®, these high order modes experience additional damping by the absorber-lined side walls.
Additionally, there is a strong longitudinal field component at abut 120 MHz, which is mainly caused by the imperfections of the end termination of the GTEM cell.
In the case of the symmetrical EUROTEM®, we find the test volume being distributed symmetrically around the longitudinal symmetry axis. This automatically means a better suppression of unwanted non-TEM components and consequently, better field quality.
Top
7. Using the EUROTEM® for Emission Measurements
It is well-known since the early time of research and from the Crawford cells, that due to the law of reciprocity, TEM cells may be used both ways for radiated emissions and immunity. Reciprocity, in this context, means the attenuation between the transmitted and the receiving signal is the same when transmitter and receiver are inter-changed at the antenna. To compare different emission facilites, a comparison noise source (VSQ) is used. One way to realise such a radiation source is to use a 10 MHz comb generator. Another way of doing this is using a frequency synthesiser with an appropriate frequency multiplier or simply a source generating white noise. Our VSQ generates a needle spectrum which rolls off pretty strongly towards higher frequencies. This spectrum is additionally shaped by the use of the small 200 MHz resonant broad band dipole. Figure 16 shows the measurements using such a test radiator being placed in the centre of the test volume of the EUROTEM® 2. By spatially rotating this dipole through the three-room coordinates, the suppression of the non-TEM components can be demonstrated. Experience teaches however to cover the metallic surface of the self-contained, battery-driven comb generator by ferrite lining. This results in an improved symmetrical behaviour of the device and the suppression of unwanted unsymmetrical surface currents which would normally appear by forming a corner dipole of 90° angle to the rotational symmetry axis of the broad band dipole arrangement. This asymmetry could reach values as high as 20 dB.
Figure 16: Direct VSQ Emission Data from the EUROTEM® 2 (example of non-TEM component suppression) |
This decoupling data must be considered as being excellent considering the fact that the 40 cm long braod band dipole antenna almost short-circuits the available space between the striplines of 50 cm. For the EUROTEM®-Antenna the data is simular however 10 dB lower (1,5 to 0,5 m gap). In the lab the non-TEM wave suppression sometimes exceeded 40 dB.
In order to gain additonal confidence in our measurements, we tested the VSQ on four different 10m OATS which were all accredited by the German EMC Accreditation System and belonged to various test organiations scattered through the Federal Republic of Germany. Doing so resulted in a very good agreement of these facilities with respect to our EUROTEM® measurements (figure 17). As a side remark, it is important to bear in mind that the VSQ has only 10dB noise margin left at 1 GHz taking EN 55022 Class B as a limit.
Figure 17: Emission Measurements by VSQ and Various Test Facilities (curve shape determined by antenna of test radiator)
|
In the real world, not only battery-driven test radiators will have to be measured but rather complete systems including cables. Even in this case, we have been able to demonstrate reasonable good correlation. Judging the overall accuracy in emission testing, one has to keep in mind that only comparing almost perfect OATS among each other with such systems including cables may easily lead to differences in excess of ±10 dB.
Top
8. Procurement of the EUROTEM® in Comparison to Other Facilities
Table 1 shows the comparison of the necessary drive power, the investment cost as well as the maximum dimensions of the test object for various facilities. The EUROTEM® 2 cannot necessarily be directly compared to the other facilities because of it's limited test volume. One major advantage of all EUROTEM® devices is the simple way of changing the polarity of the electromagnetic field, in contrast to the GTEM cell.
| |
EUROTEM® 2 |
EUROTEM®- Antenna,
short
FALC |
GTEM 1750 |
FALC
(3m site) |
| Maximum EUT size (metres) |
0.3 x 0.3 x 0.3 |
1.8 x 1.5 x 1.0 |
1.0 x 0.5 x 1.0 |
1.0 x 1.5 x 1.0 |
| Polarization switch h/v |
yes |
yes |
no |
yes |
| Facility size (metres) |
1.0 x 1.0 x 1.3 |
3.0 x 3.0 x 6.0 |
4.0 x 2.4 x 8.0 |
4.0 x 3.0 x 7.0 |
| Input power for 10 V/m, 80% AM at 80
MHz |
1.6 W |
16 W |
17 W |
65 W |
| Input power for 10 V/m, 80% AM at 30
MHz |
2.5 W |
25 W |
26 W |
250 W |
| Input power for 100 V/m automotive at 80
MHz |
ca. 25 W |
252 W |
525 W |
2 kW |
| Input power for 100 V/m automotive at 30
MHz |
ca. 77 W |
630 W |
800 W |
(7.5 kW) |
| Facility investment without
equipment |
ca. 27% |
ca. 70% |
ca. 60% |
100% |
| FALC = fully absorber lined chamber | |
Table 1: Comparison of Test Facilities
The EUROTEM® antenna has been mainly developed for integration or retro-fit of an exisiting fully absorber-lined chamber. The lower corner frequency is only limited by the characteristics of the balun. The upper frequency limit is given mainly by the absorber performance of the EUROTEM® termination and the existing absorber chamber. A comparison to the conventional radiated immunity equipment using a 3m test distance versus the EUROTEM® antenna is given in table 2.
| |
Conventional solution for a 3m site using
broad band antennas |
Solution using EUROTEM®-
Antenna, stand alone |
| Antenna |
DM 13,000.-
(X-Wing BiLog) |
DM 50,000.- |
Antenna mast
(including control
unit) |
DM 31,000.- |
not applicable |
| Power amplifier |
DM 54,000.-
(0.01 - 100 MHz, 500 W)
DM
100,000.-
(80 - 100 MHz, 200 W) |
DM 61,000.-
(1 - 1000 MHz, 100
W) |
| Total |
DM 198,000.- |
DM 111,000.- | |
Table 2: Differences between Conventional 3m Solution and Using a EUROTEM®-Antenna
The prices indicate mean net values. Additionally there is the need for measurement software, a controlling computer and a generator to be considered. For precise calibrations, one would need an additional field strength sensor system.
Top
REFERENCES
- S. Berger: "A User’s perspective on GTEM testing." In: EMC Test & Design, USA, December 1993, pp. 38-43.
- J. Nedtwig: "Vermeidung störender Vielfachreflexionen in TEM-Wellenleitern: GTEM to TRIGATEM." In: EMV 92, 3, International Fachmesse und Kongress für EMV in Karlsruhe, February 1992, pp. 243-251.
- F. Leferink: "A Triple-TEM cell: three polarisations in one set-up." In: EMC 1993, 10th International Zurich Symposium and Technical Exhibition on EMC, Zurich, March 1993, pp. 573-578.
- L. Jendernalik, D. Peier: "Expanding the bandwidth of a TEM-cell with a planar terminator." In: EMC 1993, 10th International Zurich Symposium and Technical Exhibition on EMC, Zurich, March 1993, pp. 579-582.
- W. Bittinger: "Properties of open striplines for EMC measurements." In: IEEE 1993 International Symposium on EMC, Dallas, USA, August 1993, pp. 120-125.
- P. Wilson: "Higher-Order mode field distribution in asymmetric TEM cells." In: URSI International Symposium on EM Theory, Stockholm, August 1989, 3 sides.
- L. Carbonini: "A transmission line device for EMI susceptibility measurements with enhanced field uniformity." WTEM, In: EMC 92, International Wroclaw Symposium on EMC, pp. 219-223.
- L. Carbonini: "Comparison of analysis of a WTEM cell with standard TEM cells for generating EM fields." In: IEEE transactions on EMC, Vol. 35, No. 2, May 1993, pp. 255-263.
- D. Hansen, D. Ristau et al.: "Analysis of the measured field structure in a GTEM 1750." In: 1994 IEEE International Symposium on EMC, Chicago, August 1994, pp. 144-149.
- G. Moenich: "A new conical active absorber terminated TEM cell for time-harmonic and transient use." In EMC 1995, 11th International Zurich Symposium and Technical Exhibition on EMC, Zurich, March 1995, pp. 599-602.
- R. Lorch, G. Mönich: "Mode suppression in TEM cells." In: IEEE 1996 International Symposium on EMC, Santa Clara, August 1996, pp. 40-42.
- L. Jendernalik, D. Peier, R. Schaller: "TEMpact: Resonanzfreie TEM-Zellen in Kompaktbauweise." In EMV 96, 5. Int. Fachmesse und Kongress für EMV in Karlsruhe, February 1996, pp. 317-324.
- A. Podgorski, G. Gibson: "Broadband Electromagnetic Field Simulator." Canadian Patent No. 2047999, 1991.
- A. Podgorski, J. Baran: "New concept of emission and susceptibility testing." In: IEEE 1997 International Symposium on EMC, Austin, August 1997, pp. 497-499.
- L. Carbonini: "A new TEM cell: Test results up to 3 GHz and a comparison to alternative solutions." In: ITEM UPDATE 1997, ROBAR Industries Inc., R & B Enterprises Division, West Conshohocken, PA, USA, pp. 31-40.
- D. Hansen, J. Funck, D. Ristau, S. Moessler: "Comparing the field quality of the new EUROTEM® to GTEM and fully absorber lined chambers." In: Proceedings of the IEEE 1998 Symposium on EMC, Denver, Co., USA, pp. 132-136.
- D. Ristau, D. Hansen: "Correlating fully anechoic to OATS measurements." In: Proceedings of the 13th Wroclaw EMC Symposium, 1996, pp. 402-405.
- D. Hansen, D. Ristau: "Comparing the measurement results in a fully anechoic chamber to those on four different OATS." In: Proceedings of the 14th Wroclaw EMC Symposium, 1998, pp. 206-209.
- T. Jahn, D. Hansen: "Are fully anechoic chamber emission measurements in compliance with the standards?" In: International Product compliance, launch issue Jan. - Feb. 1998, published by James & James, London, ISSN 1461-1422, pp. 25-29.
Top
|
