On the issue of PEMIN estimation of analog signals SI low-frequency scattering fields.

Author: Kondratyev A.V.

In all previous recommendations, when it came to measuring field strengths in the low-frequency range (audio frequencies), it was always stated that these issues were well developed, fully covered in existing methods and therefore were not specifically considered. From communication with colleagues, especially the younger generation, it turns out that not everything is so rosy…

Therefore, we again have to return to this issue in the form and volume that is permissible in open material.

So, field measurements in the low-frequency range. Precisely fields, since in near zone, that is, at distances of the order (0.1÷0.2)λ The electrical and magnetic components of the field (this was already written about at the very beginning of the cycle) exist practically independently of each other. For a frequency of, for example, 10 kHz, the wavelength is 30 km. Consequently, everything is closer than 3÷6 km – near zone, and even more so our usual units – tens of m –.

Since we often have to deal with the magnetic H component, let’s start with it. Measurements of the magnetic component of the field are carried out, almost without exception, by various models of loop antennas. All of them are one or more turns of a conductor (most often – round, although there are several models with a square frame), reliably shielded by a metal (usually tubular) screen with one dielectric gap. This provides good shielding from the electrical components of the field and, at the same time, does not reduce sensitivity to the magnetic field.

The frame diameter varies greatly between models, from several tens of mm to several m. The general rule arising from the law of electromagnetic induction (Michael Faraday, 1831) is that, other things being equal, the EMF at the coil terminals (at a constant magnetic field strength) is directly proportional to the area of the coil, which, in turn, grows in proportion to the square of its radius.

In differential form, Faraday's law can be written as follows:

(in the system SI)

or

(in the system GHS).

In integral form (equivalent):

or

Where:

E — electric field strength,

B — magnetic induction,

S — arbitrary surface,

δS — its boundary. The integration contour.δS is assumed to be fixed (fixed, sis!).

That is, effective height of the antenna (conditional coefficient having dimension m (for an electric antenna, for a magnetic antenna - further) and connecting the EMF at the antenna output with the field strength in A/m and, Ka, antenna coefficient; Ka= V /(A/m) = m*Ohm) will be the better the larger the antenna area (turn area).

This is one side of the issue.

But any coil, even a single turn –, is an inductance. And not idealized, but quite real – therefore with some parasitic capacity. This capacitance consists of the interturn capacitance, the parasitic capacitance of the screen, connectors, cable and the input capacitance of the measuring instrument to which the antenna is connected. Moreover, this capacitance is connected in parallel with the inductance. What are we getting? Correct – parallel oscillatory circuit. This means that an inevitable resonance will occur at a certain frequency. And this thing is extremely unpleasant, since on the one hand, resonance sharply increases the sensitivity of the antenna. On the other hand, it makes it extremely frequency-dependent and unstable due to the influence of many different, poorly taken into account factors on this resonance.

That is why developers and manufacturers of loop antennas always consider the frequency range to be much lower than the first resonance as the working area. Only there can the stability of characteristics be ensured throughout the inter-verification period.

For very similar reasons (performance stability), antennas with ferrite cores are very rarely used. They are very sensitive to temperature changes, impacts, etc.

The amplitude-frequency response (aka Ka) of the conventional loop antenna is given in Figure 19.1

Thus, taking into account some limitations of current NMDs, we must use loop antennas with a coil diameter(s) from 0.15 to 0.8 m, with the exception of certain special cases. Less – unacceptably decreased sensitivity. More – and inconvenient to use and recommendations NMD do not allow.

So, the framework of different times and peoples, the most basic information about the models available on the market is given in part 5 of this series of articles, where we talked about field strength measurements (general information).

In addition to the above, you should know that loop antennas can be either passive (that is, just a turn, a coil) or active (with a pre-broadband amplifier built into the design). Passive antennas have extremely low levels of intrinsic noise, virtually unlimited dynamic range, but their sensitivity is limited by the above expression for electromagnetic induction and the noise of the measuring instrument. But they can also be used as radiating agents, the need for which is found in research practice.

Active antennas have a dynamic range of no more than 120÷140 dB, slightly larger noises (nothing can be done, all active elements are noisy), but noticeably greater sensitivity across the field (that is, less Ka value). Naturally, they are irreversible and cannot be used as emitting ones.

A very important feature of a classic loop antenna is its directionality. The LPC (directional coefficient, dimensionless value) graphically, in a plane running perpendicular to the plane of the turn, represents two circles touching at the point where the antenna is located. In space, these are two spheres that also touch at the point where the antenna itself is located (Figure 19.2). Moreover, the line passing through the centers of the spheres is perpendicular to the plane of the frame. It follows from this that to ensure that the antenna is oriented to the maximum of the field vector, its design must allow free rotation, ideally in three, but at least two planes on ± 2¶ radian.

And the last thing that has been written about more than once, but it’s worth recalling. For a turn in a field, it does not matter at all due to this that the magnetic flux through it changes, due to a change in the direction (or magnitude) of the vector of the field itself, or due to the movement of the turn relative to this vector. This means that any fluctuations in the frame will affect the readings. Including vibrations of the frame (or conductors in it) due to the impact of acoustic (vibration) effects on it. That is, the frame itself can serve as a classic electrodynamic microphone (a turn in the magnetic field of the earth). And the signals generated due to such an effect (as well as due to several other physical effects) cannot be distinguished from the signal caused by a similar change in the measured field vector itself. Antennas, unfortunately, microphonite, sometimes – very, very... Such antenna models for measuring very small signals in the area of AEP low-frequency signals are simply unsuitable. Moreover, none of the manufacturers of standard measuring antennas sets themselves exactly this task (reducing the microphone effect).

About magnetic (frame) antennas, perhaps everything.

For measuring E components, with virtually no exceptions, shortened ones are used (naturally, at a wavelength in km!) dipoli. And, again, with virtually no exceptions – active. Otherwise, with such a short length of dipoles, their sensitivity turns out to be too low. And they microphone, sometimes no less than their frame sisters....

Directionality of the dipole, in the plane the same as that of the frame (Figure 19.2), but in the three-dimensional representation these are not two spheres, but donut, a toroid (it is clear that when the plane intersects, you will still get two circles). But the toroid is placed in a plane perpendicular to the line of the dipole axes.

About antennas – everything. Let's now move on to the measurement features.

First of all, we note that most often field measurements in the NP area have to be performed from technical means in the circuits of which dangerous signals have a noticeable value. That is, first of all, from OTSS (TSPI). Sound amplification, conference and public address systems, voice-over room systems, multimedia processing of non-public information. The sources of noticeable fields are, again, first of all, nodes and blocks with a large signal – output stages of VLF, loudspeakers (speakers), their connecting lines. All of the above applies to both magnetic and electric fields. It is very important for lines – whether they are symmetrical or not and the degree of asymmetry that symmetrical lines always have (at least due to the flow of currents along the shielding braid or simply along the ground). The higher the symmetry –, the smaller the scattering field, that is, our favorite PEMIN.

In not very common cases, visible fields are claimed and VTSS. For example, an ordinary loudspeaker (column) with a short-circuited (locked) input, due to acoustic influence, is capable of developing a current in the voice coil of a very decent magnitude, generating an equally decent scattering magnetic field.

It is extremely useful to immediately imagine from which block (node, line) we are looking for a signal and how, approximately, the field vector can be directed, therefore, how exactly to orient the antenna in advance. Lenz's rule and gimlet keep mentally before your eyes you need to constantly and actively use this (oh school physics! How necessary are you!!!). The electric field vector, as a rule, is directed from the conductor, its generatrix (block, node) along the radius, outward (in the sphere, of course). The dipole assumes maximum levels then (especially in the near zone) when the axis of the dipoles is collinear to the direction of the field vector, along it.

Measurements are complicated by the usual situation. At any point in space near the vehicle, a field is recorded not from one source, but from several random emitters, that is, a certain superposition of fields. This effect most often occurs in the electric field, but if there are several dynamic heads of direct radiation in a certain column, then the same will happen in magnet. There may also be a sum of fields from the poorly balanced supply line and from the column itself.

In any case, measurements must be carried out at several points located on the generatrix of the sphere around the vehicle under study (node, block, vehicle device) in order to identify the point where the mentioned superposition of fields will give the greatest result. It is always measured maximum maximum from all results to assess security precisely by the maximum of the dangerous signal. As we can see, general approaches to SI in this area are traditional.

Such measurements (studies) in accordance with current NMD are carried out on a tone signal. More precisely, since all PEMINs are frequency dependent – on a certain frequency grid within a set range. The choice of this frequency grid is determined, first of all, by the results of preliminary, estimated measurements. That is, it is strongly recommended to walk manually through the range with the source (generator) of the test signal and look at the high-quality view of the PEMI frequency response, so to speak, at its spectrum. This is especially convenient when the measuring instrument connected to the antenna has a panoramic indicator and the entire operating range is immediately visible on the screen (for example, as is done by industrial analyzers UP-300, UPL, etc.). It is also convenient (in fact –more convenient) to use specialized tools, for example Талис-НЧ-Лайт.

It would be useful to mention that if OTSS (TSS) are tested, then the value of the tone test signal in the circuits should be maximum. At any of the operating frequency ranges. In any case – the maximum product provided for by the established operating mode (fits into the Regulations!).

Picture on the screen shows the entire spectrum or part of it that interests the operator, with all the interference. The desired signal, in the form of a stick or bell (if you look at a narrow range), is usually clearly visible and moves in accordance with the adjustment of the test signal generator. You can immediately see at which frequencies it is higher and at which it is lower.

It is especially convenient if the test signal source can operate as a swing frequency generator (SFG) with a given period (again Талис-НЧ and its descendants).

Just don’t forget to check the antenna orientation at different frequencies, sometimes – it requires correction! Thus, using the smooth tone method, the frequencies (frequency regions) in which it is already necessary to perform measurements are easily determined, with the exact placement of the antenna, its orientation, control of the distance to the emitter, etc.

Let us remind you that all of the above seems simple, quite logical and normally executable. But when the dangerous signal is almost equal to the interference, when there are power tips, in short – in real conditions, this already requires a creative attitude to the matter.

There is one feature of measurements in this particular frequency range. Wherever we measure, everywhere space is literally impregnated with fields with the frequency of the industrial power grid and its harmonics. Moreover, the levels of these fields are often an order of magnitude higher than what needs to be found and measured. A regular picture on the spectrum analyzer screen looks like it is illustrated in Figure 19.3.

What is the most important thing that can be inferred from this image? There are two conclusions:

• In reality, at least something can be measured only in the interval between the harmonics of the electrical network;
• To obtain the minimum errors from the network, you need to measure it in the narrowest possible band;
• Without special measures to suppress at least the first 5 harmonics of the network, measuring anything in their frequency range is most often pointless.

By the way, the same rules apply when studying AEP NPs in lines.

Of course, if an additional adaptive digital filter for suppressing network frequencies is built into the analyzer, then the picture will look almost perfect (Талис-НЧ-М1). But it is still impossible to measure at power grid frequencies, there is also a signal choke.

That’s why a regular grid looks like this:

275;

327;

425;

475;

...

875;

etc. That is, it is optimal to measure strictly between the harmonics of the network. By the way, above 1000 Hz they drop so much that they practically do not interfere.

In the electric field, everything looks exactly the same, only worse (in the sense that there is a lot more interference, including from the network). For measurements according to E, a very narrow-band device is especially important, which can also accumulate and average the result for many implementations, with wavelet or other anti-interference filters, etc. Without this, it is much more difficult to work with a conventional not tailored analyzer for such tasks.

There is another technique, in one of the previous publications it was mentioned, however, in a different connection. In some cases, the following turns out to be effective:

Twist the antenna, orienting it based on imaginary interference (including –network). And, then, we begin to twist the product (column, for example) around the stationary antenna to the maximum signal. This happened even with a line (cable). Of course, not at the site (where the line is usually fixed), but at the stand, at the laboratory SI. We managed to win back up to 20 dB from interference, and this is sometimes fundamentally important!

And the last thing you should know about sishnik.

When measuring H components, if its source is – point, that is, its linear dimensions are much less than the wavelength – which is always satisfied, and commensurate with the dimensions of the frame (better – less), which is not always performed, then the attenuation of the field with the distance to the source quite accurately follows the law 1/r³. If this is a line, several dozen winds long, and even curved – it’s difficult to predict anything in advance. Indicator at r can vary from 2 to 3.

When it comes to the electrical E component, everything is much more complicated. Practice shows that, without paying attention to any theory of E component, with a distance (within 3-15 m from emitter) it attenuates proportionally (on average) as 1/r^1.5. Real (measure and set real attenuation –required!!!) the value of the indicator at r ranges from 1 to 2.5. It happens that due to interference, re-emissions from several sources (from all sorts of conductive structures) with distance, the signal increases! Each such case must be understood individually.

Thus, as in the AEP LF, the meter’s task is to identify all PEMIN maxima in the operating frequency range for both field components and measure their values at a normalized distance when the antenna is oriented to the maximum (at each frequency!).

Usually there are 3÷15 such maxima. If there is a group of maxima close in frequency with different amplitudes – the maximum is measured. If there is some plateau that is sufficiently long in frequency with a constant level , measure at its beginning, middle and end. With a very long plateau (3-5 kHz) – take at least 4-5 points distributed evenly (in frequency).

All these recommendations are not written anywhere, in any NMD. This is the result of experience and not only my personal, senior comrades taught a lot:)

When individual, significant maxima of fields are measured, they are easy to collect, calculate the equivalent signal value (already one number) and for it, in accordance with the experimentally established or accepted law of attenuation, calculate the distance at which the equivalent signal will weaken below normal. We get ordinary R2, which will unambiguously characterize our OTSS (TSPI) or VTSS.

Actually, the most important thing that awaits sishnik when measuring fields in the low-frequency range is – everything. Then, in real life, dozens and hundreds of difficulties and misunderstandings will be encountered, but firmly remembering physics and the basics of radio engineering, you can figure everything out. Nature is not going to meet us halfway, but it is not malicious! Know, be able and reason sensibly – and everything will form a consistent, correct picture of measurements and their expected results.

Which is what was required:)