LNB Detects
Coherent Thermal Radiation

This web page describes some experiments which were carried out in parallel with a discussion that took place in the usenet discussion group uk.tech.digital-tv in January and February 2011. It is a work in progress, and is intended to be seen as an example of science in action. It is to be hoped that any wrinkles in my reasoning will be ironed out over time and that a detailed explanation for the phenomenon which I have modestly dubbed the "Legon Effect" will be forthcoming.

At the very least, my experiments provided a fascinating insight into the hidden world of the interpenetrating electromagnetic fields and forces that surround us everywhere. When I started this project, I hadn't expected to be able to detect my own body heat through a solid oak door using nothing more than the LNB from a satellite dish and a cheap satellite finder, but such is the physical realm in which we live.

The detection of "coherent thermal wave effects" by satellite LNB was a chance observation which I made when investigating a question that had been raised in the above-mentioned discussion group in January 2011. One poster wanted to know why satellite receivers give a high signal reading even when a satellite dish is not pointing to any satellite. The signal quality will be zero yet the signal strength - depending on receiver - can be over 90%. Another poster replied that the signal was LNB noise, but I thought this was only part of the story. It is well know that LNBs pick up thermal background radiation - the Earth's thermal noise - as can be demonstrated by pointing a satellite dish towards the ground instead of the sky. The increase in noise level is around 3 dB. The world is awash with thermal radiation, but the sky is relatively "cool".

Out of curiosity, I thought it would be interesting to see what sort of signal I could get by connecting a spare LNB to a satellite receiver indoors, and pointing the LNB around the room. The receiver bar-meter displayed the seemingly aberrant signal strength reading of 99%. I then wired up a satellite finder and looked for some variation between different objects. Since I was drinking tea at the time, the hot mug was an obvious target, so I adjusted the finder to just below the threshold and got a satisfying response: the finder started to squeal as the LNB detected radiation coming from the mug. But then I noticed something that took me by surprise: the audible tone revealed a series of clearly defined peaks and troughs in the signal as I moved the LNB towards or away from the mug.

Having had some training in physics, it was obvious to me that I had detected some kind of coherent wave effect, perhaps comparable to the interference fringes as seen in Newton's Rings or Young's double-slit experiments. But how could this be, when thermal radiation is incoherent and is not known to give rise to such effects? I decided to raise the question with the discussion group, and made a short YouTube video to illustrate the effect I was seeing. I described the experiment in the following terms:

A universal (ku-band) LNB for a domestic TV satellite dish is connected through an ordinary satellite meter to a satellite receiver, which supplies DC power through a coaxial LNB signal cable. As the mug containing hot water is brought closer to the feedhorn of the LNB, the meter responds with both audible squeal and visual LED display. Further movement towards the LNB reveals a series of peaks and troughs in the signal level.

First Video - 6th January 2011.

Careful measurement of the intervals between the successive peaks and troughs showed that the radiation which was being detected by the LNB had a wavelength of 27.7 mm, and hence a frequency of 10.8 GHz. This is in the ku microwave band. The LNB uses the heterodyne principle to down-convert satellite signals to an intermediate frequency, or IF, which is sent through the coax to the receiver. It has a local oscillator which in this instance was set to the frequency of 9.75 GHz. The incoming signal mixed with the LO frequency to give a difference product or intermediate frequency of 1.07 GHz. This is within the normal operating bandwidth of the LNB IF.

The demonstration also works with cold water in the mug - and indeed with other objects - but as shown in the next video, hot water in the mug is more energetic and gives a much stronger response:

Second Video - 8th January 2011

Two identical coffee mugs, one filled with hot water and the other with cold water, are placed in turn at some distance in front of a ku-band LNB. The distance is reduced until the satellite meter responds with an audible squeal and the LEDs light up. Both hot and cold mugs give a response, but the mug containing hot water is detected at a greater distance from the LNB than the cold mug. The general effect with hot water is more "energetic".

My tentative explanation for this phenomenon was that all objects with a temperature above absolute zero emit electromagnetic radiation in a continuous spectrum, which extends into the 3 cm satellite waveband at room temperature. A continuous range of frequencies was therefore present in the radiation emitted by the mug, and through a resonant effect, a signal peak was detected when there was an integer number of half-wavelengths in the distance between the mug and the LNB feedhorn. To have this effect, the radiation for a given frequency had to be emitted with a constant phase relationship over distance - a characteristic of coherent radiation. The findings seemed to be consistent with the hypothesis that the energy being emitted by the coffee mug was coherent.

Some commentators in the discussion group asserted that the wave effect was caused by the leakage of local oscillator signal from the LNB being reflected back from the coffee mug. Several tests have shown, however, that the effect of LO leakage is very small. This was established by bringing up a second identical LNB face to face with the first LNB, and toggling the power on and off. With the two LNBs set to the same frequency band and polarization, it was possible to detect a change in signal, but this could only observed at close range and was at least an order of magnitude less than the coherent wave effect. The LO signal is always present inside the LNB at a much higher level than any signal that might be reflected off a coffee mug, especially since both the ceramic of the mug and the water are good absorbers and poor reflectors of microwaves; and in any case, two signals of the same frequency do not give a difference product in the IF waveband.

The answer, it seems, is that coherence is an observed behaviour rather than a property of electromagnetic radiation as such. Radiation which is considered to be incoherent can exhibit coherent behaviour when the bandwidth - or range of frequencies over which the observation is made - is limited.

In my coffee mug experiment, the bandwidth is limited in two ways. First, the distance between the mug and the LNB acts as a "tuned gap filter", which resonates at a fundamental frequency and a series of harmonics. These resonant frequencies are separated by an interval equal to (velocity of light)/(2 * distance) cycles per second. Secondly, the bandwidth is limited by the RF and IF stages of the LNB, as well as the tuned characteristic of the satellite finder. This selection of frequencies from a broad spectrum of noise gives rise to the "coherent wave effect", and may be described as imposed coherence.

The Research Continued

While the Philex satellite finder was very good at demonstrating the coherent radiation effect, it was not much help when it came to quantifying it, so I traced the circuit and identified a point where I could obtain a measurable signal voltage.

The signal detector is a tiny surface-mounted transistor which is biased close to the point of cut-off. It thus behaves as a rectifying diode which develops a voltage in the collector load resistor of 680 K ohms. To measure the output, I made a simple dc amplifier which I connected to the finder as shown in the photo above. The yellow/white coax goes to the detector and ground, and the white wire goes to the 10 V supply from the voltage regulator.

In order to plot the "coherent wave effect" against distance as objects were moved towards or away from the LNB, I connected the amplifier to the ADC in the expansion box of an old Acorn Electron computer. A few lines of code were written to give a display akin to that of an oscilloscpe, and the Electron screen was captured on a PC using a video dongle. Plastic bottles containing hot water or cold water were used as the target, and moved steadily away from the LNB as the signal was sampled.

Here is the result for water at room temperature:

The horizontal axis of symmetry towards which the waveform falls away, corresponds to the signal level obtained when no object is placed directly in front of the LNB - when the LNB sees the thermal noise background of the room. The waveform for hot water is similar but is shifted upwards with respect to the horizontal axis, as the LNB receives more thermal noise:

Here the plots for hot and cold water are combined:

In my initial experiments, the satellite finder was used as a kind of threshold detector which responded when the radiation exceeded a certain level. I found that for hot water, the LNB picked up the wave effect at nearly twice the distance possible with cold water. In the subsequent usenet discussion, this distance ratio was supposed to show that the effect that I was observing could not be caused by thermal radiation, since it could not be reconciled with the difference in energy level between hot and cold water. The graphical displays, however, show how the difference in the detection range arises. If the signal threshold of the satellite finder is adjusted such that a response to hot water is obtained at point "A", then the bottle of cold water needs to be moved to point "B", at about half the distance, for a response to be obtained.

I later determined that the height of the computer display was equivalent to a difference in signal level of about 2.6 dB. The vertical shift in the waveform was consistent with a difference in thermal noise of about 0.5 dB between water at room temperature and hot water at around 360 K.

In the usenet discussion, an expert in coherent wave theory concluded that I had constructed the analogue of a 'Fabry Perot' resonator, which behaves as a form of reflection interference filter. The frequency selection, and hence coherent effect, results from having multiple reflections all adding in phase due to the path delays.

Measuring Bill's Kitchen

One of the participants in the usenet discussion was a well-respected installer of TV aerials and satellite systems, and the author of numerous articles in popular publications such as What Satellite magazine. Bill was keen to investigate the thermal effect that I had reported, and set up an experiment in his kitchen using a kettle, an LNB, and a professional spectrum analyser. The analyser was set to display the full satellite IF band and to store the maximum reading for a period of 60 seconds, and snapshots were taken with the kettle full of cold water, after it had heated up for a while, and when it was full of boiling water.

At first sight the results were rather disappointing - there seemed to be little or no difference in the analyser display regardless of whether the kettle was full of cold or boiling water. It was pointed out, however, that the polished steel surface of the kettle was highly reflective of thermal radiation, and by the same token it was emitting little heat. In effect, Bill's analyser snapshots were taken not of his kettle but of his kitchen, as reflected in the shiny metal surface.

Nonetheless, whether the kettle was hot or cold, the analyser displayed a distinct pattern of peaks and troughs, and it later occurred to me that this might be a vestige of the thermal resonance of Bill's kitchen as a whole. A little work with a computer graphics program showed that the peaks were at fixed intervals across the screen, and that the difference in frequency between the peaks could be calculated since the frequency range across the display was known.

Since the above analyser display covers the full IF band, the range in frequency will be 1200 MHz, and the interval between the peaks is found to be 58.3 MHz.

Now my theory of thermal wave resonance predicts that a frequency interval of 58.3 MHz requires a tuned gap of about 2.6 metres, as shown by the formula f = c/2*d . Since this is a plausible dimension for a room, I asked Bill about the size of his kitchen, and he kindly supplied a plan with the exact dimensions marked in millimetres - noting that the kettle and LNB had been placed on the worktop to the left of the sink. Consequently, I was able to report that the resonance was taking place between the kettle and the wall directly opposite. The width of the kitchen was 2820 mm, and subtracting 200 mm for the distance from the kettle to the wall behind gave an effective width of about 2.6 metres, in agreement with my prediction.

John Legon