Feedback (Jan. 1987)

Home | Articles | Forum | Glossary | Books

Or not, as the case maybe I recently bought a hairdryer. As an integral part of the hairdryer a cable was supplied which is to be connected to a mains plug by the user. The instructions read as follows: IMPORTANT

The wires in this mains lead are colored in accordance with the following code:


The wire which is colored blue must be connected to the terminal marked with the letter N or colored black. The wire which is colored brown must be connected to the terminal marked with the letter L or colored red.

Can any of your readers help me convince a member of the opposite sex that engineers really do think logically or are we all quite mad?

W. Scott, Highgate, London N6.

Heat transfer I refer to a letter by Mr G. Nalty in the October issue in the Feed back column. I am afraid that Mr Nalty, like so many others, has been confused by Specmanship in his interpretation of the thermal performance of semiconductors and especially transistors.

Mr Nalty claims to have heard the difference in distortion in an amplifier when changing from a power transistor rated at 12.5 watts to one rated at 50 watts due to lower junction temperature. I would suggest that the truth of the matter is that the difference was due to the actual transistor characteristics rather than junction temperature. Let us ex amine the facts.

Mr Nalty's circuit conditions are such that the device is dissipating 60 milliwatts. Presumably the transistor is mounted in free air for such a small dissipation. What is important in this condition is not the Rth junction to case (Rth;,) but the Rth;_a (junction to ambient) which is considerably different and very much larger. Furthermore, this figure is in the main independent of package for a given die size, although not necessarily true for all die sizes. It is easy to find devices in 10220, SOT82 and SOT32 (T0126) in the SGS Power Transistor Databook which use the same die and exhibit very different Rthi_c but the same Rthj_a. But let us be generous and use some figures which do vary, for example an Rthj_a of 62.5 deg C/W and 83.4 deg C/W which can be found for certain devices in TO220/SOT82 and 10126 respectively. We have here a difference of about 20 deg C/W which for the quoted dissipation of 60 milliwatts will yield a junction temperature difference of a mere 1.2 degrees centigrade between the two de vices. Taking this to its logical conclusion and using figures for more representative types we would have negligible tempera ture difference, certainly much less than the normal measurement tolerance capability, tending to prove my previous statement that the distortion difference is due to transistor (die) type alone.

Mr Nalty might like to note that it is not thermal resistance which decreases with increase in power dissipation, but rather the other way round. The Rth defines the maximum allowable dissipation with reference to the maximum allowable junction temperature. It is the first link in the chain from the die to the outside world, so to speak.

Let us look at this phenomenon a little further and take as an example the devices TIP110, TIP120, TIP130, SGS110, SGS120 and SGS130. The three TIP--- types use increasingly larger die assembled in TO220 while the SCS--- types are the equivalent die in SOT32 package, which ís physically the same size as 10126 (or SOT32) but without the fixing hole, which thus enables a larger die to be accommodated. The Rthj_a is identical for all six types (regardless of die size or package type), but the Rth;_c is another matter. Die for die (e.g. TIP110 to SGS110) the Rth)._, is the same but it decreases from 2.5 through to 1.78 deg C/W as the die size increases from 110 to 130. As a parallel to this the figure for 7800 regulators in 10220, which use very much smaller die, shows a RthJ_c of 4 deg C/W.

I hope this sheds some light on a rather ill-understood topic.

Nigel Pritchard, Field Applications Engineer, SGS Semiconductor Ltd, Aylesbury.

Filters and pulses I was interested to read at the end of Mr C.F. Coleman's letter (July issue), the question "Why, if the Fourier transform of a single square pulse can represent the zero level before the pulse arrives, should the calculated out put which it generates when applied to a low pass filter begin before the leading edge of the pulse arrives?" I remember as a student (many years ago) being puzzled by precisely this anomaly. The answer which emerged after a great deal of effort, mathematics not being my strong suit, was that the ideal pulse and network postulated were not physically realizable. If one utilized in the calculation the characteristics of a pulse and a network which were physically realizable, then no output occurred before the input pulse was applied, and the effect was delay the rising edge of the output pulse with respect to the input pulse.

No doubt your more mathematically inclined readers could furnish a convincing proof.

J. Hollingworth, Grand Caymen, B.W. Indies.

S5/8 I regret that I am still unconvinced about the proposed S5/8 system, despite the enthusiastic flag waving by its proponents.

The worst point about this system is that it is not a properly designed transmission system; it takes no account of the properties of the transmission line, and it is not terminated. It is thus very prone to interference pick up and ringing; a problem that must increase as line length is increased. It is obvious that lengths in excess of 1 meter will be required. Mr Tomlinson's point regarding the t.t.l. parallel interface to disc drives is very valid. These are used only on short runs, but the cable is at least terminated! I am not impressed with arguments about an extra pound or two for components, nor even an extra 100 mA or so. I am sure that anyone who has lost valuable programs or text from a computer system would gladly pay the price for a reliable sys tem. Such a properly designed interface need not tie up software time with CRC checks etc. These need only be used when unstable media such as radio link or magnetic recording is involved.

A wired system should be as reliable as possible.

I would be very interested to learn what system is employed for consultation between BSI and the engineering community before standards are proposed.

Clearly this proposal would have benefited by input from engineers in the television, radar or audio field where the points I have listed have been well known and understood for many years.

Perhaps wider publication of such proposals in Wireless World might be advantageous.

Les Hayward, Eastpoint Ltd, Corfe Castle, Dorset.

German v.l.f.

I am engaged on a program of research into a very low frequency transmitting radio station established in Germany, late autumn 1941.

The station called "Goliath", was near the village of Kalbe ( Saxony, Prussia) and partially straddled the River Milde.

I seek any information, no matter how small or seemingly trivial. Any drawings, photo graphs, technical details or personal memories by German or Allied personnel would be especially welcome. Any documents loaned to me would be treated with care and postage refunded.

T.F. Bernascone, Goliath Research Project, Teeside Polytechnic, Borough Road, Middlesbrough, Cleveland TS1 3BA.

Relativity In the October issue of E&WW, Alan Watson of Mallorca challenges skeptics to provide physic al evidence against the second postulate on which Albert Einstein based his relativity doc trine. Perhaps I can help.

The postulate was presented in the second paragraph of Einstein's famous 1905 paper in which he introduced his Special Relativity. He wrote in German.

Translated into English, the postulate reads:

"...light is always propagated in empty space with a definite velocity c which is independent of the state of motion of the emitting body." Statements about motion and velocity are meaningless unless they specify (or unambiguously imply) a frame of reference. In later years, Einstein and his followers interpreted the postulate to mean: The velocity of light measures as c under all conditions.

Consider the radar equipment used by traffic police to catch people driving their automobiles too fast. Typical is the case in which a police vehicle sits at the side of a highway, aiming a radar beam at approaching traffic.

One person is driving toward the police vehicle at a speed (a velocity) of 100 km/h (relative to the road surface) in violation of the law. The radar beam bounces off the offending vehicle, bounces back to the police vehicle, and records the illegal land speed in the police radar equipment. The police officer signs a citation, and the offender pays a fine.

How does the radar work? The radar beam emitted by the police vehicle travels at c (nearly 300,000 km/s) relative to its source. It is reflected back to the waiting police vehicle. How fast does the reflected beam travel? At c, relative to the vehicle from which it is reflected.

So far, Einsteinians and anti Einsteinians agree. But then we ask the critical question: What is the velocity of the incoming reflected beam relative to the police vehicle? It is only c, the Einsteinians reply. But the anti Einsteinians say no, insisting that the returning beam comes in at a velocity of c+100km/

The radar equipment in the police vehicle shows the Doppler effect, a slight shift in the frequency of the incoming beam, and calculates the oncoming vehicle's illegal land speed of 100 km/h, which is equal to

If the incoming beam's velocity were only c (as Einsteinians contend), there would be no shift in frequency.

But there is a shift, and Einstein was mistaken.

Now a bit of direct evidence.

Scientific American, in its issue of November 1961, published an article describing the two-mile linear accelerator to be built at Stanford University in California (it was built a few years later, and has been in operation for many years). On page 50, the article said the velocity of particles would be 0.9999999997c relative to the accelerator. Then, a few words later, on the same page, came this:

"....the accelerating (electro magnetic) field must travel at a velocity close to that of the particles. Some slippage between the two is allowable, so long as the particles stay near the wave crests (the points of maximum accelerating force). To bring the wave velocity close enough to that of the particles, the inside of the pipe is designed with a series of ridges. These have the effect of slowing the electromagnetic waves travelling through the pipe. If the pipe were perfectly smooth, the waves would travel faster than c and would be unable to transfer energy to the particles... A proper choice of the dimensions of the ridged structure slows the wave to the de sired velocity."


1. A Einstein, "On the Electrodynamics of Moving Bodies," in The Principle of Relativity, translators (1923) W. Perrett, G.B. Jeffery (Dover, New York, 1952), p.38. (Also, General Publishing, Toronto, and Constable & Co., London.) Lee Coe, Berkeley, California.

I was very interested in H. Aspden's letter concerning relatively and the Sagnac effect.

STC in London are developing a optical-fibre gyroscope which depends upon this effect for its operation. However, what is detected is a rotation of an optical-fiber coil, not linear motion, thus the general theory and not the special theory is involved. In E.W. Silvertooth's experiment as described, surely what is being detected is a rotational effect as the earth rotates around the sun.

D. Marquis, Westerham Hill, Kent.

As every schoolboy surely has been taught, it is perfectly possible to "prove" (mathematically, of course) that 2=1. Permit a reminder:

Let a=bi and multiply both sides by a, so a2=ab.

Subtract b2 from both sides-

a2-b2=ab-b2 Factorize (a+b)(a-b)=b(a-b) Cancel (a+b)=b

But a=b

Therefore 2b=b' 2=1 It is quite like removing every vestige of energy from a system so that a transform may work in the static state of mathematics, simultaneity having broken down, so that the assumption may be made that the speed of light is infinite to correct the situation, so that the interval in space-time between two events on the world line of a photon may be made zero, in order that the said transform may work as intended! Of course, it is just another Ezekiel's wheel revolving round an axle of presumption.

One is then left with the infinite speed of light when the system is devoid of energy: very neat! This is further explained by the strange statement that mass and energy are the same thing, whence energy must have inertial mass because it is non existent.

Unfortunately for the mathematicians, both these are interrupted by catastrophe which makes both the numbers and their symbols totally valueless.

What this demonstrates is the lack of value of mathematics unless they are fully backed by an active dynamic logic which may be expressed in words: that is to say, the mathematics must be done without first freezing the system solid by removing every vestige of energy, and existing thought is sheer poppycock. Sorry Mr Coleman, but at least Mr Snoswell should be pleased, along with Mr Aspden and Mr Abdullahi when they get down to it, all in November Feedback.

Incidentally, Mr Editor, and presumably for the benefit of your (what used to be called) "comps", the word is "couch" as recorded in the Book of Job: to "couch" implies servility but "couch" implies a sleepy state of static boredom! James A. MacHarg, Wooler, Northumberland.

I was interested in the comment in the letter from M.D. Abdullahi of Zaria, Nigeria in your issue of November, where he obtained the energy of a charge as (mc2/2, rather then mc2.

mc2/2 is the finite limiting value of the translational kinetic energy of the electron, and we can obtain this as equal in magnitude to the finite limiting value of the spin angular kinetic energy of the electron as the translational velocity approaches zero, which is the rest energy.

As the translational velocity of the electron approaches its finite limiting value, the fluctuations in the magnitude of the spin angular velocity increase indefinitely, and the electron liberates an infinite sequence of photons of indefinitely increasing energy. Thus the electron can take as much energy from the accelerating field as the field can give it, and most of this energy is radiated as an infinite sequence of photons of indefinitely increasing energy.

During the fluctuations in the spin angular velocity of the electron, the resultant angular momentum is conserved by fluctuations in the lateral component of the translational velocity of the centre of mass, this representing the commutation relations of translational and angular velocity of quantum mechanics. At the minima of the total kinetic energy as the translation al velocity approaches its finite limiting value, the magnitude of the spin angular velocity approaches zero, and the translational kinetic energy, which is approaching its finite limiting value, has almost entirely re placed the finite limiting value of the spin angular kinetic energy when the translational velocity approached zero, which is the rest energy, and the magnitudes of the two finite limiting energies are equal.

R. Fricker, Surbiton, Surrey.

In the November issue, M.J. Snoswell asks "what happens to a top"? The answer to that must surely depend upon how one treats the top. One result is shown in the picture. The 24 pound top is spinning at 2000 r.p.m. and is rising quite rapidly up a helical path.

The downward pressure on the finger is no more than two pounds and so its seems that some 22 pounds of gravitational mass has mysteriously vanished.

A further aspect of the affair is that the calculated centrifugal force in the horizontal plane from the centre of the helix was to have been 17 pounds but in the event there was no c.f. at all.

Again it seems that all of the inertial mass vanished.

Perhaps some dedicated relativist who fervently believes that to increase the energy of a body is to increase its mass will tell us where the mass got hid.

It may be that if such an explanation is forthcoming we shall know how the Cheshire cat's grin got out of the bag in order to join MacHarg's menagery or, is it perhaps that we have a new science and Albert was wrong? Alex Jones, Alderney.

Alan Watson (April Letters) rightly emphasizes that the cur rent experimental evidence for the time dilation predicted by Special Relativity establishes the effect beyond question. However the work which he actually cites was presumably based on the use of atomic clocks in fast flying aircraft or in satellites. Even for the latter the predicted effect is only one part in 1010, so that the experiments are open to the criticism that some small systematic effect may have been neglected, and the gravitational effects on clock rates are of com parable magnitude.

Much more dramatic are the measurements on fast-moving unstable subatomic particles such as the muon and pion. The measured lifetimes of these two particles at rest are 2.20 µs and 26.0 ns respectively, and when they are in motion the values should, according to Special Relativity, be increased by the time dilation factor. The earliest measurements of this kind involved comparing the muon component of the cosmic rays at the top of the atmosphere and at ground level, and were made in the late thirties. The muons travel downwards at almost the speed of light, and to account for the large fraction that survive the journey it was necessary to assume a lifetime several times larger than the estimated value for muons at rest.

Precise quantitative measurements became possible with the advent of high-energy particle accelerators fifteen to twenty years ago. Measurements on the fraction of positive pions which survived transit along a twelve meter flight path gave a time dilation of 2.40, within 1/20 of the Special Relativity prediction for pions of the selected energy'.

In principle, this measurement corresponds closely to the earlier cosmic ray measurements. Relativistic muons, however, travel about 700 meters in 2.2 uSec, so that to observe time dilation in the lab. it was necessary to force them to travel repeatedly round a circular path, and to measure the survival fraction by sending off bunches of muons at suitable intervals, and observing the number that decayed near a fixed position on the path as a function of time after starting. Such measurements have been made for muons of two different energies. The more recent set gave an increase in lifetime by a factor of 29.3, within 0.1% of the Special Relativity prediction 2.3. Because the muons are forced to return time after time to the same position in the lab., these last experiments actually measure differential ageing, i.e. the phenomenon involved in the so-called 'twin paradox'.

The meson experiments test aspects of Special Relativity which are completely inaccessible to measurements made with atomic clocks, and confirm its predictions with unparalleled precision. Moreover they are virtually immune to the effect of the earth's rotation, of the earth's orbital motion round the sun, and of its gravitational field. May we now forget the Haefele Keating experiment?

C.F. Coleman, Grove, Oxfordshire


1. D.S. Ayres, A.M. Cormack, A.J.

Greenberg, R.W. Kenney, E.F.

McLaughlin, R.V. Schafer, D.O.Caldwell, V.B. Elings, W.P.

Hesse, and R.J. Morrison, Nucl.

Instrum. Methods 70(1969)13

2. J. Bailey and E. Picasso, Prog.

Nucl. Phys. 12(1970)43

3. J. Bailey, K. Borer, F. Com bley, H. Drumm, F. Krienen, F.

Lange, E. Picasso, E. Picasso, W. von Ruden, F.J.M. Farley, J.H.

Field, W.Flegel, and P.M. Hat tersley, Nature 268(1977)301.

Apart from the problem he poses (Feedback, November, 1986) Mr Abdullahi should expect to encounter discrepancies involving factors of 1/2 from general considerations. The reason is that our perception of the physical world is based on sampling of our external environment by our sensory apparatus. Even with the very large number (10 to 15 thousand million) nerve cells in the brain, we cannot sample at twice the rate at which some external events occur in order to comply with the requirements of Nyquist sampling frequency and thus avoid 'abasing' as a necessary component of some aspects of mental activity. In addition, the electrical and chemical activity of the central nervous system is quantized as trains of voltage spikes (called 'action potentials' by neurophysiologists). From this, it would be expected that, at some (very low) threshold of perception and cognition (called ‘limen' by psychologists), events of the physical universe would only detected with probability 0.5, constituting a possible explanation of, for example, Bell's inequality.

B.E.P. Clement, Crickhowell, Powys, Wales.


(adapted from: Wireless World , Jan. 1987)

Top of Page

PREV.   NEXT   Guide Index HOME