GB2359934A - Noise absorbing filter which can be switched between two filter operating modes - Google Patents

Description

1 2359934

TESTING EMISSION OF AND IMMUNITY TO ELECTROMAGNETIC FIELDS

This invention relates to methods and apparatus for testing emission of and immunity to electromagnetic fields in which radio-frequency energy is coupled through electromagnetic radiation in the immediate environment of the equipment under test. Within this specification references to "radiation" should be taken as including both outward radiation and inward reception of radiation unless the context explicitly disallows this.

Reference should be made to European standards EN55022 and EN61000-43:1995, which explain the terminology of this technical field and give practical details of how to do such tests. These referenced documents discuss the importance of precisely defining the cable lengths and layouts of the equipment under test (hereinafter MUT'). This is important because these cables form an important part of the antenna system by means of which the EUT radiates or absorbs electromagnetic energy. As is well-known in the art of antenna design changes in dimensions of say, one- tenth of a wavelength in antenna element length or spacing can have a large effect upon antenna gain and directivity.

It was shown in UK Patent Application G132 179 502 how to limit the antenna effect of cables by the use of ferrite-cored common mode chokes, and recently it has been described by E Ristig in the IEC Committee draft CISPR/A125 I/C1) how an "absorbing clamp" (which is a well-known contemporary common-mode ferrite device fully described in Appendix G to lEC standard CISPR 16) may be used to limit the length of each EUT cable that might act as an antenna. According to Ristig an absorbing clamp is to be placed around each cable as it passes through a conducting plate that forms the boundary of the test region. It should be noted that the ferrite-cored devices employed in these two references magnetically couple a complex impedance in series with the cable in common-mode. This impedance is typically inductive at low frequencies, resistive at medium frequencies, and of a low inductive reactance at a high frequency The exact impedances and transition frequencies depend upon the configuration, dimensions, and ferrite material used. For applications such as desribed it is of primary importance that the modulus of the common mode impedance should be at least a few hundred ohms: The exact phase angle of this impedance is less important. It follows that these two references describe essentially similar known arrangements in different but related applications.

For the definition of tests of electromagnetic compatibility that can facilitate international trade it is desired to secure measurement conditions that are easily repeatable yet approximate to practical use. Accordingly reasonable worst-case conditions for such cable-antenna effects should be established and these effects should not be reduced too much by the configuration of the cables in the test set-up.

Whilst Ristig's proposal referenced above points towards a method for defining the effective radiating length of each cable leaving the region of the ELTT, it does not address the question of how it may be ensured that the radiation from different portions of such a single cable, or from a number of such cables, is not subject to destructive interference and so cancelled out or reduced to an extent that allows the EUT to pass a test that it should reasonably have failed. In the specifications referenced above it is taught that this should be done by moving the cables about on a trial-and-error basis until a worst-case is obtained. However this is time-consuming and leads to significant differences in measurement results because of the different cable layouts used adopted by individual test engineers.

2_ It is therefore desirable to adopt some method of identifying the effects of destructive interference between the various cables acting as antennae. It has been suggested to keep the cable positions strictly fixed in standardised positions and to employ common mode chokes on the cables selectively to secure emission free from cancellation effects. These chokes might either be positioned close to the EUT so as to decouple the cable from the EUT, or be positioned at a distance from the EUT where they act by damping or detuning the resonance of the cable. Hitherto this has been an unattractive approach because of the time taken to identify and test to the worst case selection of choke positions, which may be different at difFerent frequencies.

According to the present invention there is provided a switchable cable decoupling device including a common mode choke coil with an auxiliary winding, said auxiliary winding being short-circuited upon demand by an electrical or electronic switching means.

A specific embodiment of the invention will now be described by way of example with reference to the accompanying drawing in which: - Figure 1 shows a test set up for a table-top EUT together with the possible locations for switchable cable decouplers according to this invention., Figure 2 shows a detailed circuit of a switchable cable decoupler hereinafter denoted by ccSCD5.

Figure 3 shows an interface circuit between the SCD of Figure 2 and a fibre-optic control cable.

and Figure 4 shows in outline the circuit of a remote controller that can sequence up to four SCDs each as detailed in Figure 2 via an optical fibre and the interface of Figure 3.

3 Referring to Figure 1 the EUT comprises two units la and 1b connected together by a multicore cable 2 and to the outside world by multicore cables 3a and 3b. If for example the EUT comprised a computer and monitor then these cables might comprise power supply and local area network connections. These latter cables leave the EUT region through a filter panel 4 set into the conductive ground plane upon which the table stands. It should be noted that such a filter panel usually bonds the cable earth wire or outer shield to the ground plane, securing a very low common-mode impedance to ground at this point. The above elements and layout may be found described in International Standards such as EN55022.

It is now proposed to add an SCD 5a or 5b on either or both of the cables 3a and 3b. Alternatively, if the test specification so requires, SCDs may be positioned at 6a and 6b where they function by raising the common-mode impedance so as to damp or detune the resonance of the cable sections between the ground-plane and the EUT. 7 shows an additional possible position for an SCD that might advantageously be specified.

It is to be understood that the test set-up described above might be upon an open site or within a screened room, an anechoic chamber, a GTEM cell, or any other test environment.

These SCDs may be switched in turn to decouple each of these cables from the EUT. It may prove sufficient to sequentially set just one SCD at a time to the high-impedance state - that is 4 states for 4 SCDs. Alternately it may be preferred to work through all possible combinations of impedance states for the SC1)s provided - that is 16 combinations for 4 SCDs. For emission testing each state might be maintained for approximately 2 milliseconds, which would be just long enough for a quasipeak detector according to 1EC standard CISPR 16 within the measuring receiver to register the strength of the emitted field.

Figure 2 shows any one of the SCDs 5, 6, or 7 in diagrammatic detail. Each comprises a tubular sleeve 10 of high-frequency lossy ferrite material such as Philips grade 4S2, which may be split lengthwise for convenient assembly around the EUT cable 5, 6, or 7. A conducting loop 11 forms a turn around the sleeve via a decoupling capacitor 13 and a PIN diode 12 which may be an Hewlett-Packard type HS14P3890. The PIN diode is biased by a dc potential applied between connection points 15 and 16. Resistors 14a and 14b limit the dc current flow and provide radiofrequency isolation between the SCD and its associated drive circuit.

If point 15 is held positive by, say, 7 volts with respect to 16 then the PIN diode is reverse biased and presents an impedance of some 0.3pF so that the loop is effectively open-circuited and the inherent impedance of the ferrite sleeve is realised so as to reduce any common-mode current flow in the EUT cable. The magnitude of the bias voltage that needs to be applied depends upon the voltage that may be impressed upon the PIN diode by the radiated field.

If on the other hand 15 is held negative with respect to 16 then the PIN diode conducts and assumes a low resistance so that the loop is completed and acts as a short-circuited turn to reduce the common-mode impedance coupled into the EUT cable.

t+ It is to be understood that the high-frequency performance of this basic circuit may be enhanced by the use of multiple capacitors in parallel in place of decoupling capacitor -13, and by the reduction of leakage inductance by replacing loop 11 by parallel loops spaced around the periphery of sleeve 10. A further reduction of leakage inductance may be secured by adding further switched-loop assemblies comprising duplicates of 11, 12, 13, 14a and 14b connected in parallel at points 15 and 16 with that shown. It is particularly appropriate to employ one such assembly on each half if core 10 is split.

It is also within the scope of the present invention to switch the loop impedance by an electromechanical relay, or by an alternative solid-state device such as a junction transistor or field-effect transistor with electrical or optical input.

It will be understood from Figure 1 that if an SCD is provided in positions 5a, 5h, or 7 then any control wires leading to points 15 and 16 would seriously disturb the radiated field that is the subject of the test. Accordingly Figures 3 and 4 show how a control scheme including a fibre-optic control cable may be implemented. If however an SCI) is provided in positions 6a or 6b then the control wires leading to points 15 and 16 may be routed directly through the adjacent filter panel 4 without disturbance of the radiated field. The necessary control arrangement would then be simplified by the omission of the fibre-optic link components, and can be understood from the following description without further explicit reference.

Figure 3 shows the interface circuit that may be provided immediately adjacent to the SCD of Figure 2, and which is to be connected to the correspondingly-identified points 15 and 16. In the quiescent state the 9volt battery 25 provides a positive bias to point 15 via resistor 2 7 and a negative bias to point 16 via resistor 26. Negligible current flows in any part of the circuit since the transistors 21 and 22 are held nonconducting by resistors 23 and 24. Therefore a local on/off switch is not required.

In the active state the phototransistor detector 20, which may be a Siemens type SFI-13 50V, is turned on by infra-red radiation fed to it along a 1000 micron acrylic fibre from the source to be described later. The current through 20, limited by resistor 28, turns on the complementary transistors 21 (type BC479) and 22 (type BC 109) so reversing the polarity applied between 15 and 16 and setting the SCD switch into its low-impedance state. The resulting current flow also illuminates the red LED 2 7 to indicate accordingly.

The various capacitors shown in figure 3 serve to improve the electromagnetic compatibility of this circuit.

It is to be understood that phototransistor detector 20 may be replaced by a photodiode, and complementary transistors 21 and 22 may be replaced by field-effect or MOS types. The cicuit described uses a single battery and two semiconductor switching elements 21 and 22 to reverse the output polarity: It may be replaced by an arrangement with two batteries and a single semiconductor switch. In this case the switch may be chosen to be directly photo-sensitive so as to avoid need for the separate phototransistor detector 20.

5_ Figure 4 shows in outline the circuit of a remote controller that can sequence up to four SCDs. A clock oscillator 30, which might be constructed using an astable configuration of the 555 integrated circuit timer, produces a rectangular pulse output at 550Hz. This waveform is presented to a two-stage frequency divider 31a and 31b which might be implemented with a 74HC74 integrated circuit. The true and complementary outputs from each of these stages are presented to the multichannel amplifier 32 whose outputs provide the 50mA current requirement of Siemens SFH450V infra-red transmitter photodiodes 33a, 33b, 33c, and 33d which are each coupled to a fibre-optic guide leading to one of the SCD interface circuits described earlier. This circuit configuration energises up to four SCDs in all possible configurations, with a dwell time of 1.82 milliseconds in each state, to meet the most severe functional requirement outlined in the context of Figure 1.

It is to be understood that the simpler sequential energisation of four SCDs might be achieved by a modification of this circuit using only one divider stage and decoding its Q and /Q outputs with a 2 to 4 line decoder type 74HC 13 9.

For a system that requires only a single SCD the clock oscillator 30 may drive a single amplifier 32a and infra-red transmitter photodiode 33a directly, the other circuit elements being omitted. In this case the clock oscillator timing might be chosen to energise the SCD for 2mS in a cycle of 9.13mS. This low duty-cycle economises in battery power and allows study of the EUT emission to identify which state leads to the greater emission.

In general it is preferred that the frequency of switching an SCD be chosen to be distinct from the mains supply frequency and its harmonics so that any amplitude modulation effects from these two sources may be distinguished.

It will be evident that these various logic configurations may be combined with manually-operated selector switches to provide for variable switching patterns for a variable number of SCDs.

The above description is by way of illustration and does not limit the invention.

Claims (7)

1) A switchable cable decoupling device adapted for assembly around a target cable, said. device including a common mode choke sleeve with an auxiliary winding, said auxiliary winding being short-circuited upon demand by an electrical or electronic switching means.

2) Apparatus according to claim 1 in which said common mode choke sleeve is split so as to facilitate assembly around the target cable.

3) Apparatus according to any preceeding claim in which said auxiliary winding and/or said switching means are duplicated around the circumference of said sleeve.

4) Apparatus according to any preceeding claim in which said switching means includes a semiconductor switching device.

5) Apparatus according to any preceeding claim in which said switching means is remotely controlled by and/or supplied with power by means that include a fibre-optic cable.

6) Apparatus according to any preceeding claim in which said switching means is operated so that either each period of short-circuiting or each period of open-circuiting is of a duration between 100 microseconds and 100 milliseconds.

7) Apparatus substantially as described herein with reference to Figures 1 to 4 of the accompanying drawing.