Walls have ears but could they have mouths as well? An alternative to thermo-elastic model of microwave hearing allows for human hearing of a.m and f.m as well as pulse signals.  By Dr Chris Barnes, Bangor Scientific and Educational Consultants, Bangor, Gwynedd, Wales, UK e-mail manager@bsec-wales.co.uk

  REVISED AND RE-PUBLISHED NOVEMBER 2015                    

Abstract

A new hypothesis is developed to account for why buildings are required in some people’s perception of the acoustic phenomena known as the ‘Hum’.  Causes of the Hum are usually put down to Acoustic or radio – frequency phenomena.   In the case of the latter the field strength of pulsed signals in the offending buildings is rarely, if ever, large enough to account for Hum perception by thermo-elastic means.  Low frequency electromagnetic energy may however be directly converted to acoustic energy by rectification and a piezoelectric transducer effect in outer building walls.

 

Introduction

 

This present paper attempts to explain how some people might perceive the phenomenon known as the Hum, which is often only heard by those afflicted at night and inside buildings.  Some cases of the Hum have been explained in terms of infrasound of external original and/or low frequency noise (refs).  Others account for the Hum in terms of the so –called microwave hearing effect (refs) the only proven mechanism of which is the so-called thermo-elastic model.

 

However, there are many cases of the ‘Hum’, which as yet remain totally unexplained.   These are cases wherein there is no externally measurable sound field and no externally measurable seismic vibration.  Those advocates

of  a Hum of electromagnetic origin would in these cases advocate electromagnetic signals for its causality. (Refs) 

 

 

The only known proven mechanism by which humans are known to detect and ‘hear’ electromagnetic signals is microwave hearing by the so-called so-called thermo-elastic model. (refs). However, in order to hear such signals they have to be both pulsed and above a certain threshold radio frequency field strength.  (refs). Such pulsed microwave signals are often perceived by humans as a such pulsed microwave signals are often perceived by humans as series of clicks (ref) which is uncharacteristic of what people subjectively describe as the Hum.  Further, indeed, and as result of NRPB limitations, it    is rarely if ever the case that field strength of pulsed signals in the offending buildings is ever, large enough to account for Hum perception by thermo-elastic means. Such pulsed signals might be expected to arise from GSM and CDMA mobile phone installations, airport radar systems and DAB and DVB broadcasting.  That is not to say other mechanisms for electromagnetic hearing have not been proposed. There are several including; quantum biological models (ref) direct interaction with the cochlea or acoustic nerve (refs) and bioactive ferritin in the brain (ref). However, as far as the author is aware none are as yet tested and proved.

 

 

So what of an entirely new mechanism?  What if the conversion of electromagnetic signal to sound / infrasound takes place external to the body but within the very fabric of the house itself!

 

 

Hypothesis; Cavity wall as a radio receiver.

 

Most UK houses since as early as 1932 use so-called ‘cavity wall’ construction. Essentially the outer load bearing walls of the house feature a double brick skin separated by a layer of air and locked together at regular intervals by so called ties or tie bars.  Early   ties were flat slightly twisted steel bars and corroded readily.  Even more modern ties, which are bow tie shaped pieces of galvanised steel wire, will corrode in places where the galvanisation is compromised. In any case in the galvanised system there are two dissimilar metals, the ideal scenario for a rectifying junction. In the early system the rust /steel interface could act as a rectifying junction.  Many ceramics of which brick is an example are semi-conductors, another possibility for radio demodulation.  The ‘rust nail’ effect is well documented amongst radio engineers ( refs) and thus it is at least possible to see how a house cavity wall might demodulate a radio transmission. Such rectifier demodulation will work for AM transmission (envelope demodulation) or FM (slope demodulation).

 

However in any radio receiver transmission we need an output transducer.  These days there would be several stages of amplification driving a moving coil loudspeaker.  In the early days of radio a crystal detector would energise headphones directly the dc bias being extracted from the rf carrier wave.

 

So what of the output transducer system in our cavity wall?  Brick and cement mortar is a naturally piezoelectric material.  So here we have it. The demodulated signal from the brick/tie bar system directly energies the brick, which will then vibrate mechanically in sympathy with the driving waveform.  It would be expected that such vibrations would be of minimal amplitude and hence only perceived at night when the house was really quiet.  It would further be expected that the wall would act like a low pass audio filter for the transmission of sound. This has been shown by   Craik (ref).

 

 

The last essential element of a ‘crystal radio ‘ is tuning.  The LC element in our cavity wall is effectively a lump tuned array and very difficult to calculate accurately. However best estimates for the resonant frequencies of a single array element are based on the inductances of tie bars and the effective brick shunt capacitance.   Estimates of tie bar inductance vary from 100-600 micro-henries (ref) and estimates of the brick capacitance taking the dielectric constant of brick as 3 (ref) vary from 0.3 –5 pF. Mortar capacitance can be up to a hundred times higher.      Taking the mortar into account yields estimates of resonant frequency from 39 KHz- 2.9 MHz.  The Q of this system is not expected to be particularly high due to water losses in the mortar of the outer wall and thus reception on frequencies outside the estimated range might not be precluded.

 

Bringing the above together yields a new mechanism by which radio signals could be received and demodulated and possibly perceived at night in quiet rooms as the ‘Hum’.

 

 

The Wall Voltage

 

An amateur Hum investigator John Dawes has referred to a 50Hz a.c voltage measurable by probing the plaster of house walls. (ref) Two characteristics of this so-called wall voltage are that it is phase shifted with respect to the ac mains and varies with weather conditions.

 

A possible hypothesis here is that the wall voltage exists by capacitive coupling from the house mains wiring. However it is possible to test this hypothesis by isolating the mains electricity to house, see experimental. A possible significance of the wall voltage is that if it is of significant amplitude it may bias our brick –tie rectifier system. Two fold effects would then be expected. Firstly, enhanced sensitivity of the rectifier system to radio frequencies.  Secondly, one might expect non-linear mixing of the wall voltage frequency with any demodulated frequencies present.     This certainly complicates the hypothesis but may well provide a measurable audio response, which some hearers could identify directly as ‘their Hum’.    

 

 

 

Experimental work

 

Experiments were performed using a fundamentally very simple yet highly effective suite of apparatus. The apparatus consisted of an Acer Travel mate 2423 laptop computer running Windows XP and a shareware spectrum analysis programme know as Spectrum Lab. Spectrum lab is capable of operating as a near real time audio spectrum analyser to 5.5 KHz or as a direct conversion radio receiver with waterfall display DC to 50 KHz.  In order to test if the wall possessed efficiency as a radio antenna at higher frequencies a portable radio transceiver Vertex Standard type FT817 was also utilised in receive mode.    A coaxial wire probe was first used to analyse the a.c voltage on the house bedroom wall with reference to mains earth and was either connected directly to the line –in socket of the laptop or to the antenna socket of the transceiver. . This strategy was used as mains earth is bonded to brick in sockets.  The spectrum analysis program also allows the PC to play the audio play back in almost real-time.

 

In order to measure acoustic emissions from the wall a large loudspeaker Acoustic Solutions type AV120 MKII was used as a moving coil microphone and butted up in intimate contact with the wall.

 

 

 

 

 

Results

 

Perhaps the most dramatic result was the presentation of the audio signal derived from the coaxial wire probe (wall voltage signal) to the author’s wife, a Hum hearer.  Immediately she reported that this was one of the noises she hears in the dead of night.  Various spectra were recorded and are categorised below.

 

 

http://www.drchrisbarnes.co.uk/HUMWALL_files/image002.jpg

 

 

                                                                                                        Figure 1 coaxial probe signal 0-65 Hz.

http://www.drchrisbarnes.co.uk/HUMWALL_files/image004.jpg

 

Figure 2 Coaxial probe signal 0-400 Hz.

 

 

http://www.drchrisbarnes.co.uk/HUMWALL_files/image006.jpg

 

Figure 3 Coaxial probe signal 0-400 Hz with extra signal present

 

 

 

 

http://www.drchrisbarnes.co.uk/HUMWALL_files/image008.jpg

 

Figure 4; direct conversion receiver mode.

 

 

It can be seen that the low frequency wall voltage dominates this spectrum. However, the wall does appear to be resolving about five significant signals in the range 20-25 KHz and a further weak signal at about 46-47 KHz.  The presence of such signals is closely consistent with the theoretical resonance calculation presented above.  It is interesting to note that submarine communication frequencies dominate this frequency band.    Just lower than this frequency band power line harmonic frequencies could, also presumably,  be received by the wall, either directly or by re-radiation from space.  

 

Experiments with the radio transceiver in receive mode showed that the wall provided no useable antenna gain in the frequency range 1-30 MHz over and above that provided by the connecting wire itself.  Again this suggests that the lower estimates of resonance frequency for the wall are likely to be most accurate.

 

 

An experiment was done comparing the wall voltage function with and without mains electricity connected to the house wiring.  This was facilitated using the two-pole isolator switch located between the house electricity meter and fuse box.

http://www.drchrisbarnes.co.uk/HUMWALL_files/image010.jpg

 

Figure 5 Mains supply on and then isolated.

 

 

It can be seen from figure 5 above that isolating the ac mains from the house reduces the frequency spread and amplitude of the 50 Hz component.  Surprisingly the low frequency noise pick up 0-50 Hz is unaffected as are the harmonic components of the mains frequency.

 

The conclusion must be somehow that the other frequencies are coupling to the wall by a mechanism other than internal mains current leakage. That is the wall is acting as a very low frequency antenna with respect to mains ground. It could be it receives circulating currents present in the outside earth wherein the foundation of the wall is seated.

 

 Further spectra were recorded of the direct audio spectrum from the wall. This time the signal was considerably weaker and the Hum was not perceived directly on the PC speaker,

 

 

http://www.drchrisbarnes.co.uk/HUMWALL_files/image012.jpg

 

 

Figure 6 direct wall audio in range 0-95 Hz.

 

 

http://www.drchrisbarnes.co.uk/HUMWALL_files/image014.jpg

 

Figure 7 Direct wall audio 1 minute later.

 

 

Experiments were performed with a wider frequency window on the spectrum analysis program but very little output was yielded even at higher mains harmonics consistent with the wall as a low pass filter (ref).

 

Discussion coaxial probe signals

 

The most evident feature in figures 1 and 2 is that of 50 Hz mains component and its harmonics. This must indeed be Dawes so called wall voltage. However, finer inspection shows other significant frequencies and both frequency and amplitude inter-modulation products.  Close inspection of the three figures shows frequency components at 17, 25, 35 and 75 Hz in addition to mains frequencies and harmonics.  The 50 Hz band appears very broad as though it has been modulated in frequency space from about 37-64 Hz and there is some weak evidence of time amplitude gating.   Whereas the 200 and 400 Hz harmonics appear to be strongly time gated in amplitude at the same approximately 1Hz interval of the 17 and 35 Hz pulses.  Clearly a very complicated non-linear mixing process is at work and the sound played out of the PC speaker took on the author’s wife’s; perception of the Hum.    

 

Figure 3 is similar except that there appears to be another signal process creating sidebands on the 250 and 300 Hz mains harmonics. GSM is known to produce 217 Hz.     

 

 Discussion of wall audio output signal

 

The wall audio signal, figures 6 and 7 was found to contain some similar frequency components and some dissimilar components to those contained in the wall voltage signal. Frequency components   of 23, 34, 35, 45, 50, 54, 62, 82, 85 and 95 Hz can be seen. Interestingly all the components are pulsing on and off at a pulse repetition frequency of about 1HZ with the exception of the 54 Hz component which is continuous although may have an upper sideband which is pulsed. In the wall voltage signal the 17 Hz component was pulsing on and off at about once per second as were the outer fringes of the 50 Hz components. The higher harmonics of the mains were pulsing but at a slower rate.   The similar frequency components probably arise by direct piezoelectric conversion of the wall voltage and any received radio signals. For the dissimilar frequency components we may well need to look elsewhere. In this respect, it should not be forgotten that an outer cavity wall of a house can conduct ground borne vibrations and radiate them into a room directly as infrasound or acoustic sound.   It must be remembered that piezo-electric conversion is a two way street.

 

It has been shown previously that with a mixture of low frequency components one can perceive the Hum.   (ref)

 

Conclusions

 

It is concluded that walls do indeed have ‘mouths’ and that those mouths radiate low levels of low frequency sound perhaps dependent on the presence of external radio signals and ac mains fundamental and harmonic frequencies. However, it must not be forgotten that an outer cavity wall of a house can conduct ground borne vibrations and radiate them into a room directly as infrasound or acoustic sound.  The complex patterns of low-level infrasound and sound emitted as a result of both sources are probably what some perceive as the Hum.

 The plausible hypothesis proposed by this paper has been proven experimentally and forms a valid alternative way that humans can perceive mainly low frequency electromagnetic energy other than by the thermo-elastic    mechanism of microwave hearing!