GB2220061A - Power meter - Google Patents

Abstract

A meter for measuring electro-magnetic power available in free space, for example, in the near field of a radio frequency transmitting aerial, includes three detectors each comprising two pairs of sensors. The sensors of each pair are an electro-optic sensor (124, Figures 4,5) and a magneto-optic sensor (114, Figures 4,5) arranged in series and with their light transmission axes orthogonal to one another. The order of the sensors is reversed in one pair relative to the other pair. The three detectors are arranged so as to provide outputs dependent on the components of the Poynting vector in three orthogonal directions. The meter is thus suitable for use in free space where the field orientation is not known and is insensitive to fields other than those induced by the transmitter. <IMAGE>

Description

POWER METER The present invention relates to meters for measuring electromagnetic power available in free space, for example, in the near field of a radio-frequency transmitting aerial.

The National Radiological Protection Board ("NRPB") has issued guidelines on the maximum levels of radio frequency signals which can be allowed before exposure to them becomes a health hazard. To ensure that levels are safe for personnel there is a requirement for a meter to measure the power available in the near-field of a transmitting aerial. It is important that this measurement is done accurately. Obviously, safety is of first importance, but if personnel were prohibited from areas which were actually safe (because of over cautious hazard meter readings) maintenance and other work could be severly restricted. This could be an even greater problem if the new, more stringent, safety limits proposed by the NRPB are adopted.

Most r.f. hazard meters measure either the electric (E) or magnetic (H) fields individually. The power flux due to the radiation is then calculated assuming the relationship which exists between E and H in a far-field plane wave. The relationship used is E = Z H (1) 0 where Z0 is the impedance of free space (377err). This leads to a power flux density of P - E2/Zo s ZO H2 (2) This relationship is not valid in the near field of a transmitting antenna. As well as radiative fields there are reactive components of E and H in the near field. Thus the total E and H fields can be larger than would be expected from expression (1) above.If a meter is calibrated using expression (2) above it can give an over cautious reading in the near-field since the reactive components die away when the field is loaded by the presence of a human body. These components would however be detected by the meter.

Another difficulty of near-field measurements is that the measuring device itself can alter the field distributions.

However, even in the near-field, the r.f. power flux per unit area is equal to the Poynting vector E x H. (Here the 'x' sign represents the vector cross product.) It is proposed that E and H should each be measured as three Cartesian components (i.e.~Ex, Eys Ez, Hx, Hya Hz). Then if i, 1, k are unit vectors in the x, y, and z directions respectively.

E x H 1 k ki k l x Ey Ez I I Hx Hy Hz I Expanding (3), E x H (EyHz - EzHy) + j(EzHx - ExHz) + k(E=Hy - EyHX) (4) and the magnitude of E x H is

This expression (4) can be simplified if the probe is aligned manually with the power flux. If, say, the power flow is in the z direction (along k) then E x H = k(ExHy - EyHX) (6) It is proposed that the magnitudes of the three components of the electric and magnetic fields be measured using optical sensors.

In these sensors the E and H fields modulate the intensity of incident beams of light. Electric and magnetic fields can affect the way polarised light propagates through solids, liquids and gases. These phenomena are known as electro-optic and magneto-optic effects. If the changes in polarisation due to these effects are measured, the electric and magnetic fields can be calculated.

There are several different electro-optic and magneto-optic effects. However, we have appreciated that two of them are particularly useful in the context of field measurement and can be much larger effects than the others. These are the Pockels and Faraday effects. They are both linear with respect to the applied fields.

When a material shows different refractive indices for incident light of, say, vertical and horizontal polarisation, the material is said to be birefringent.

The Pockels effect is a linear electro-optic effect in which a birefringence is induced in a material in proportion to the applied electric field. It occurs in certain crystals, depending on particular symmetries of the crystal structure.

The relationship between the difference in refractive indices for the two polarisations and the electric field can dependent on the orientation and properties of the particular crystal. However, typically:

where 9 i - relative angular delay of one polarisation in radians, nO - 'ordinary' refractive index, r - electro-optic constant in xV 1 (actually a term, r63, in the electro-optic tensor) = O - vacuum wavelength in m, V = potential difference across crystal.

It is convenient to define a 'half wave voltage' t. This is the voltage needed across a particular crystal to give a phase difference between the two polarisations oftr.

Some typical electro-optic constants are: Material r n units 10-12 m/V kV ADP (NH4H2P 4) 8.5 1.52 9.2 KDP (KH2P04) 10.6 1.51 7.6 KDA (KH2A504) 13.0 1.57 6.2 KD P(RD2P04) 23.3 1.52 3.4 Also, the plane of polarisation of linearly polarised light passing through, say, a piece of glass is rotated when a magnetic field is applied in the direction of propagation. This is called the Faraday effect.

The angle e through which the plane of polarisation is rotated in proportional to the magnetic field: e s V H 1 (9) where H - component of magnetic field in the direction of propagation, 1 = length of medium traversed, V - 'Verdet' constant in m-1 (Am ) typical values of the Verdet constant are: material Verdet constant at 589.3 nm (aAl Light flint glass 6.5 x 10 Double extra dense flint glass 2.2 x 10-3 Quartz 3.5 x 10-4 Optical detectors, such as photodiodes, can detect intensity changes but, in general, their output is independent of polarisation. So, to detect a change in polarisation the change has to be converted into an intensity variation. One simple way of doing this is to use a linear polariser.

If plane polarised light is incident on a linear polariser, with an angle 0 between the plane of polarisation of the light and that of the polariser, then the fraction of light transmitted is given by Malus' law: Transmission = Transmitted intensity 2 2 8 (io) Incident = Incident intensity A meter for measuring power flow in transmission lines utilising sensors based on the Faraday and Pockels effects is described in a paper published by Kiev State University in "Pribory i Tekhruka Eksperimenta" No.3 pp 110-112 May-June 1972 (Budovskii et al). However, this meter is not suitable for measuring the power available in free space because it relies on the fact that the direction of the power flow is determined by the path of the transmission line conductor and, hence, the orientation of the induced field is known.

More importantly, the device decribed in the paper mentioned above gives an incorrect reading in the presence of an electric or magnetic field other than the induced fields, for example, the earth's magnetic field.

In these circumstances, the device would indicate the existence of a power flow where, in fact, there is none. The Budorskii meter is intended for use in measuring flows around 50kW; at this order of magnitude, the influence of the Earth's field would be small.

In accordance with the invention there is provided a detector for measuring power flow due to electro-magnetic radiation, the detector comprising two pairs of sensors, each pair comprising an electro-optic sensing means and magneto-optic sensing means arranged in series, the light transmission axes of the sensors of one pair extending parallel to the light transmission axes of the sensors of the other pair but the order of the sensors being reversed in one pair relative to the other; the outputs of the two pairs of detectors being combined to provide a signal proportional to the component of the Poynting vector in a direction orthogonal to the light transmission axes. We have appreciated that a true indication of the power flow along any given direction can only be achieved by a complete determination of the component of the Poynting Vector in that direction.This result is insensitive to external magnetic fields whose effects simply cancel out. Preferably, the detector comprises three groups of two pairs of sensors arranged so as to provide outputs dependent on the components of the Poynting vector in three orthogonal directions. Utilizing such an arrangement of sensors1 the power available in free space can be ascertained even through the orientation of the electric and magnetic fields is not known.

A power meter in accordance with the invention will now be described in detail, by way of example, with reference to the drawings in which: Figure 1 is a schematic diagram of a magnetic field sensor utilising optical techniques; Figure 2 is a schematic diagram of an electric field sensor utilising optical techniques; Figure 3 is a diagram showing a Wollaston prism; Figure 4 is a schematic diagram of a power meter in accordance with the invention; Figure 5 shows a modified power meter in accordance with the invention; Figure 6 is a schematic diagram showing the arrangement of sensors in a complete power meter in accordance with the invention; and Figure 7 is a block iagram of the output circuitry of the meter of Figure 6.

A magnetic field sensor 10 using optical techniques is shown schematically in Fig. 1. Incident unpolarised light is plane polarised by means of a linear polariser 12 before passing through crystal 14. A polariser 16 at the other end of the crystal 14 is set at 45" to the first polariser 12. Some crystals show natural optical activity. That is to say, with no external field, the plane of polarisation of incident light its rotated by an angle proportional only to the path length in the crystal. The field- dependent rotation is added to this. If a crystal of this type is used in the field probe, the second polariser would be mounted with its axis at an angle of B + 450 to that of the first, where # is the natural rotation angle.The effect of this is to 'bias' the light output so that magnetic fields in both directions (for and-against the direction of propagation of the light) can be aeasured. It also means that the response of the probe is very close to linear for small changes in plane of polarisation.

It can be shown that the light output intensity It as a out function of the component of the H field parallel to the direction of propagation of the light is

where Iin is the intensity of the incident light.

For small variations this is very close to a linear relationship

Fig. 2 shows an electric field sensor 20. Polarised light from linear polariser 22 is incident on the crystal with its plane of polarisation at 450 to the 'fast' and 'slow' axes of an electrooptic crystal 20. Once again, some crystals exhibit natural as well as induced (Pockels) birefringence. If necessary this can be compensated for (as in the H probe). The light emerging from the crystal 24 is analysed by a quarter wave plate 25 and a linear polariser 26.

It can be shown that

Once again, for small variations this is approximately linear, so

where c and d are constants, and E is that component of the electric field parallel to the direction of propagation of the light.

It is important that the H field sensor does not response to electric fields, and vice versa. It should be possible to find a material which has a large Verdet constant but exhibits no linear electro-optic effect (because of the crystal symmetries). To make the E probe insensitive to H fields it should be possible to use a material which has an electro-optic effect significantly larger than its magneto-optic effect. Also, if circularly polarised light were launched into the electro-optic crystal, Faraday rotation would have little effect.

To measure the power flux density E x H, products of spatially orthogonal components of E and H are required (for example E H ).

xy This multiplication could be done electronically but is more easily achieved by multiplying the signals opticall.

Consider the responses of the E and H field sensors described above: For E probe oriented along x axis

For H probe oriented along y axis

If the light output from, for example, the E field probe is used as the light input to the H field probe them the overall transmission is

Now let E - E sin wt (16) ox and H = R sin (wt + d) (17) y oy then 1out = Iin ( a + bHOysin(wt+ ) ) (c + dE sin wt) (18) oy ox = 1in ( ac + cbH sin (wt + #)+ adE sin oy ox + bdE H sin wt sin (wt + #) ) (19) ox oy - dc term + terms at frequency w + cross term Expanding the cross term using standard trigonometric identities:: E H sin wt sin (wt + d) a # (cos d - cos (2wt + d)) (20) ox oy ox oy So the overall transmission of the sensor is I t I Iin ( ac + cbH sin (wt + d) + ado sin out in oy ox + #bdEoxHoy (cos d - cos (2wt + #)) (21) oy To summarise, it is now clear that the light output of a detector consisting of an electro-optic E field sensor 20 followed by a magneto-optic H field sensor 10, as described above, would have - a constant term - terms oscillating at w - terms oscillating at 2w - a term proportional to E H cos d ox oy when subject to E and H fields in the x and y directions respectively with frequency w.

The components at frequencies w and 2w can be filtered out of the detector output signal. This would leave the constant term plus a term proportional to Eo=Hoy cos 6. To provide an output signal dependent only on EoxHoycos 6, the constant term must be eliminated.

The sensors 10 and 20 outlined above use linear polarisers to convert the changes in polarisation caused by the electric and magnetic fields into variations in light intensity. Because the sensors 10 and 20 are biased to have a linear response there is an offset at the detector output. In other words, there is a light output when E and H are both zero. This output is represented by the constant term referred to above.

One way of removing this offset and, hence, eliminating the constant term is to analyse the output of the sensors using a Wollaston or Rochon prism.

An ordinary polarising prism transmits only one plane of polarisation. Wollaston and Rochon prisms transmit both polarisation components, but in different spatial directions as shown in Fig. 3. An r.f. power detector, similar to the one described above, could utilise one of these prisms in place of the output polariser 16. It can be shown that if the output of- the prism in one polarisation varies as I + aE H cos 6 then the other o ox oy would vary as Io - aEoxH cos d. (High frequency oscillatory terms oy have been disregarded). Thus the offset can be removed simply by taking the difference of the intensities of the two output beams which have opposite polarisations. This can be done electronically using a differential amplifier.

As well as the difference, the sum of the intensities of light in the two polarisations could also be found. This would serve as a confidence check that optical links between the detector components are unbroken. An assurance that the meter is in fact working is very important, because a meter with a broken link would always give a zero power flux reading. If the operator were not made aware of the fault he could enter a hazardous field believing it to be safe.

A detector using a Wollaston prism in this way is shown in Figure 4. The detector comprises an electro-optic sensor 120 similar to that shown in Figure 2 whose output forms the input to a magneto-optic sensor 110 similar to that shown in Figure 1 and arranged about an axis orthogonal to that of the electro-optic censor 120. The electro-optic sensor 120 includes a linear polariser 122 at its input end, an electro-optic crystal 124 and, at its output, a quarter-wave plate 125 and polariser 126. The beam output by the sensor 120 is deflected through 90 by a reflector 130 and enters the magneto-optic sensor 110.

The sensor 110 again includes an input polariser 112 and a magneto-optic crystal 114 but the linear polariser at its output is replaced by a Wollaston prism 132.

Light from a light source 134 passes along an optical link 136 and enters the detector in which is passes through the electrooptical crystal 124 and magneto-optical crystal 114 in succession.

At the detector output, the Wollaston prism 132 splits the transmitted light into two beams having different polarisations.

These pass via optical links 138 to detectors 140 which detect the intensities of the two beams. A differential amplifier 142 provides an output signal proportional to E H yx A possible disadvantage of a Wollaston prism is that it may not be capable of sufficient resolution of small changes in the plane of polarisation in some circumstances. If this is indeed the case then an alternative construction is proposed. Two separate beams are used, with opposite polarisations; the beams being applied to the input in parallel. The result is similar to that obtained using the Wollaston prism in that the difference in the intensities of the two beams at the output of the detector will give an output proportional to E H cos 6.

ox oy A detector using separate beams is shown in Figure 5 of the drawings. In most respects the detector construction is the same as that shown in Figure 4 and like components have been indicated by the same reference numerals.

At the input of the detector of Figure 5, a beam splitter 160 is positioned between the optical link 136 carrying light from the source 134 and the linear polariser 122. The beam splitter 160 splits the incident light into two beams which are transmitted separately through the sensors 120 and 110. At the detector output, the Wollaston prism 132 is omitted and separate polarisers 117 provided for each of the two output beams. Again, the differential amplifiers 142 provides an output signal proportional to EyH.

In order that the probe should have the smallest possible influence on the fields being measured, the probe head should be linked to the processing electronics by a long (perhaps 10m) optical fibre link. This would mean that the light sources (LEDs or lasers), detectors, and readout device would be remote from the measuring head.

It could be argued that there will be electro-optic and magneto-optic effects within the fibres themselves. However, these would be of no consequence. This is because a) the 'uplink' to the probe head would carry unpolarised light, and b) the only information carried by the 'downlink' would be a light intensity, not a particular polarisation.

The output of the single detector shown in Figures 4 and 5 is affected by the presence of an external magnetic field. To avoid this, it is proposed that the detectors be used in pairs, the individual detectors of each pair having their electro-optic and electro-magnetic sensors reversed relative to one another so that the difference of the detector output gives a complete indication of the Poyning Vector component in relevant direction. For example, using a pair of detectors oriented as in Figures 4 and 5 would provide an indication of the component of the Poynting Vector in the z direction: E H -E H xy yz Any effects due to external magnetic fields such as the Earth's magnetic field would be cancelled out.

Figure 6 is a schematic diagram of the complete r.f. power meter for use in free space where the orientation of the electric and magnetic fields is not known. Each of the three pairs of detectors is orthogonal to the other two so that so that the magnitude of the Poynting vector E x H is measured.

A probe using this detector arrangement should not be affected by electric and magnetic fields not associated with a flow of power (for example the earth's magnetic field). Consequently, the probe's output would be zero if the power flux were zero.

The outputs from the six detectors forming the meter illustrated in Figure 6 would be combined as shown in Figure 7 using known circuitry and techniques.

The safety limit in use at present limits r.f. exposure to a power density of 100 W 2. If the new NRPB limits are adopted this will be reduced to 10-50 Wm 2 over the VHF and UHF bands. To measure these power densities will require a probe sensitive to E fields of 60 Vc and H fields of 0.1 At 1.

Optical sensors have been reported in the journals for several years capable of this degree of sensitivity and it should be possible to construct a probe sensitive enough to measure these fields at VHF and UEF frequencies.

Additional advantages of the detector probe described are: ) An all dielectric sensing head could be made. This would cause minimal disturbance to the fields being measured.

b) The meter should give a reliable measure of r.f. power flux density, even in the near field.

Claims (8)

1. A detector for measuring power flow due to electro-magnetic radiation, the detector comprising two pairs of sensors, each pair comprising an electro-optic sensing means and magneto-optic sensing means arranged in series, the light transmission axes of the sensors of one pair extending parallel to the light transmission axes of the sensors of the other pair but the order of the sensors being reversed in one pair relative to the other; the outputs of the two pairs of detectors being combined to provide a signal proportional to the component of the Poynting vector in a direction orthogonal to the light transmission axes.

A detector according to claim 1 for measuring power flow in free space, the detector comprising three groups of two pairs of sensors arranged so as to provide outputs dependent on the components of the Poynting vector in three orthogonal directions.

3. A detector according to claim 1 or 2 in which the light transmission axes of the electro-optic and magneto-optic sensing means of each pair are orthogonal to one another.

i A detector according to claim 1, 2 or 3 in which the electrooptical sensor comprises a crystal in which the Pockels effect can be induced by the application of an electric field and a linear polariser at the input end of the crystal.

5. A detector according to any of clans 1 to 4 in which the magneto-optical sensor comprises a crystal in which Faraday effect can be induced by the application of a magnetic field and a linear polariser at the input end of the crystal.

6. A detector according to any of claims 1 to 5 having at the output of each pair of sensors a Wollaston or Rochon prism, means for detecting the intensity of the light beams output by the said prism and means for combining the output signals from the means for detecting intensity.

7. A detector according to any of claims 1 to 5 having at the input of each pair of sensors a beam splitter for splitting incident light into beams of different polarities and, at the output of each pair of sensors, means for detecting the intensity of the light beams transmitted by the pair of sensors and means for combining the output signals for the means for detecting intensity.

8. A detector for measuring power flow due to electro-magnetic radiation, the detector being substantially as hereinbefore described with reference to any of the drawings.