HENRY MARROWS, AN INSTRUMENT DESIGNER of New York City, quarrels with the cliche that the primary tools for scientific research ultimately boil down to just a pencil and a piece of paper. "A brain limited to this method of recording data," he writes, "would certainly arrive at a narrow concept of the universe!" Over most of the spectrum of natural phenomena it is impossible to record observations without the assistance of more rapid, more sensitive, more tireless or more accurate recording instruments. Marrows explains: "A pencil propelled by human muscle can keep up with only about five events per second, and a human being cannot stay awake long enough to make a continuous record of events lasting more than about 24 hours. In the parlance of instrument designers, the muscle-driven pencil has a sharply limited frequency response. The frequencies of nature stretch across a spectrum paced at one extreme to the slow evolution of galaxies, covering billions of years, and at the other extreme to the spin of an electron, taking place within a millionth of a microsecond.

"At the low-frequency end, nature provides some ready-made recorders. The age of galaxies is recorded in their slowly changing form. The eras of the earth's evolution have been registered as strata in the rocks. The cycles of climate and other annual events can be read in tree rings and varves (lake-bottom mud deposits). And for recorders of more rapid events we have the deposits left by flash floods, the craters left by meteorites and the fused substances marking points where lightning has struck.

"There are certain important gaps in the spectrum where nature fails to furnish a record. Most of the interesting ones lie in the frequency band between a few hours and an infinitesimal fraction of a second. To bridge these gaps modern instrument makers have devised various types of recorders. All of them consist essentially of a pen which automatically traces a record of the observed event on a chart of some kind. The pen may be linked to any kind of detecting or observing instrument-a barometer, a thermometer, a seismometer, a telescope or whatever. The sensing instrument translates what it observes into a signal- mechanical or electrical-which drives the recording pen. A recording instrument has two paramount advantages over a pencil-using human recorder: it can operate tirelessly around the clock and it is capable of responding to much higher frequencies.

"The chart on which the instrument's pen writes may be a fixed sheet, a moving ribbon, a cylinder or a disk. The pen may be powered directly by the energy of the event itself (as in an old-fashioned seismograph) or may be driven by a motor actuated by an amplified signal from the event (as in a starlight photometer).

"The varieties of events to which a recorder may be applied are legion. Albert G. Ingalls, formerly editor of this department, spent his first summer in retirement recording the tidelike seiches (vibrations) of Seneca Lake in New York by means of a mechanically-actuated pen recorder which he constructed during the winter months. Studying his graph, 90 feet long, he found that the 35-mile lake had a long-period vibration of 56 minutes. He is now analyzing higher-frequency modes which appear as tiny peaks on the graph. The study promises to disclose interesting information about the shape of the lake bottom, as well as clues to the forces responsible for the seiches. Another amateur, John Ruiz of Dannemora, N. Y., recently constructed a seismograph of the velocity type and is recording movements of the earth's crust through the range of 30 seconds per vibration to half a second. His home-built recorder employs electronic amplification. John Wilke of Chicago has fitted his eight-inch reflecting telescope with a photoelectric cell photometer and an electronic recorder which enable him to plot the decay of starlight during occultations in increments of two hundredths of a second. Radio hams and high-fidelity addicts study vibrations in the range from about 20 per-second to billions per second. You find these amateurs investigating everything from the characteristics of transistors and home-made loudspeakers to the electromagnetic resonance of chemical compounds.

"Amateurs frequently ask for advice about the best kind of recorder to buy or build. Usually they have in mind a universal recorder-one that will accept signals from many types of sensing instruments. The criterion of selection is frequency response. The pen of the ideal recorder should be capable of deflection for frequencies 10 times higher than those anticipated in the variable under measurement. Thus if it takes a tenth of a second for the moon to occult a star, the photometric recorder should have a minimum response time of a hundredth of a second. Such a recorder will give a 'flat' response: that is, at every frequency the signal will deflect the pen by the same amount, thus tracing a straight line. Instruments of poor response, in contrast, will not respond uniformly to all frequencies, and therefore the graph will be a curve.

"A flat response is not necessary for all purposes. Sometimes a less than ideal instrument will do. One amateur biologist now investigating the circulatory system of the frog wanted to make recordings of the electrical impulses liberated by the beat of the animal's heart. He was interested in frequencies up to 200 cycles per second. He built an instrument around a war-surplus oscillograph but found that the pen covered only 7 per cent of its full-scale swing at 200 cycles and 50 per cent at 60 cycles. By taking this limited response into account, however, he could interpret the records as faithful measurements of the frog's heart action.

"The high cost of direct-writing recorders has discouraged many amateurs from taking up such hobbies as variable star observing, seismology, micrometeorology and other avocations in which precise records are essential. A good chart drive and pen motor sells for about $300. A companion amplifier and power supply adds another $500. Equipment for dividing or multiplying the incoming frequencies and thus extending the range of the basic combination, plus equalizing amplifiers to compensate for distortion, can shoot the investment into the stratosphere. But if you have access to a drill press and a small lathe, you can make a good wide-range recorder at a cost of less than $250.

"You start with the pen motor, the heart of any recorder. A number of basic types are shown in the drawing on page126. Most commercial instruments utilize the d'Arsonval movement in one form or another. Its construction, however, is strictly a job for the experienced instrument maker. In contrast, pneumatic pen drive is relatively easy and inexpensive to make. It consists of a simple bellows linked to a pivoted lever which carries the pen on its outer end. The pneumatic motor may appear crude, particularly to devotees of electronic technology. Yet when coupled to a nozzle of the type shown above, the pneumatic motor is capable of astonishing sensitivity.

"The nozzle consists of a pair of coaxial tubes. Its outer tube ends as a small orifice from which a minute jet of air escapes. As an example of how it works, suppose we use it with a seismometer. To the pendulum of the seismometer is linked a vane, which is placed close to the orifice and moves toward it and away from it with motions of the pendulum. When the vane comes within a few thousandths of an inch of the jet, back pressure builds up inside. As the vane moves toward or away from the orifice the back pressure varies in direct proportion. The pressure is communicated through the small central tube to a bellows, which in turn moves the pen. The pen therefore swings in direct proportion to the movement of the pendulum. With air under a pressure of 30 pounds per square inch flowing from a sixteenth-inch orifice, the device is sensitive to a change of two millionths of an inch in the distance between the orifice and the vane. It gives a directly proportional response for distances up to 15 thousandths of an inch. The obstruction need not be a vane, of course. The pen will indicate changes in distance between the orifice and any smooth object.

"A single tube with a 'T' near the center works as well as the coaxial jet over distances up to four thousandths of an inch. One end of the tube is connected to the air supply and the other becomes the jet. Back pressure is taken from the 'T' connection.

"It is interesting to arrange a pair of 'T' jets and companion bellows mechanisms for operation as a push-pull pen motor [see drawing in Figure 4 ]. The obstruction can be a vane of light metal supported between the jets by a pivoted arm. Mechanical signals from the observing instrument are coupled to the vane through a conventional linkage. A regulated supply of compressed air enters the sensing nozzles at 80 pounds per square inch and impinges on opposite sides of the vane. The spacing between the surface of the centered vane and each orifice should not be more than two thousandths of an inch. As the vane is moved back and forth by incoming signals, back pressure is communicated to the bellows alternately, causing the spring-loaded pistons to swing the pen. The pen is centered by adjusting the spring tension. The vane floats free between the opposing jets, hence little energy is required to move it. The result is a combination pen motor and pneumatic amplifier capable of power gains of 10,000 to one and higher.

"Electrical signals also can be fed into the device, but they must first be converted into mechanical movement by a device such as the d'Arsonval motor. (This unit is sold by the Sanbon Company, the Brush Electronics Company, the Edin Company Inc., and others.)

Alternating-current signals of five hundredths of a volt fed into the d'Arsonval unit will deflect the pneumatic pen to the limit of its travel. Moreover, the response is uniform for all frequencies from one to 60 cycles per second.

"What is perhaps more appealing than the sensitivity of the unit is its remarkable stability and reproducibility. It requires no controls for keeping the pen centered on the chart or for adjusting its sensitivity to various frequencies. In many applications extremely small mechanical signals can be fed directly into the vane. Because of the push-pull arrangement, small variations in the pressure of input air do not perceptibly affect either accuracy or stability. A small compressor of the type used for a paint sprayer will supply all the pressure needed, and the only electrical power required is 110-volt alternating current for working the compressor, heating the stylus and driving the chart motor

"For the chart a direct-writing recorder may employ writing paper, wax paper or paper that changes color when an electrical current is passed through it. Each has characteristic advantages and limitations. The pneumatic pen motor develops enough power to drive any conventional writing tip, even pencil lead. Unless the chart movement is slow (a few inches per hour) pencils are inconvenient because they require frequent sharpening. An inking pen, preferably in the form of a glass tube pulled to a point and rounded against a fine carborundum stone, is convenient in low-frequency applications at any chart speed. But a pen is messy, clogs easily and at high frequencies tends to throw ink. On some tracings it may be accelerated at the rate of 16,000 feet per second.

"Waxed paper (inscribed by a heated stylus) behaves nicely at high acceleration, but it costs more than untreated paper and the thickness of the trace varies somewhat with frequency.

"The heated stylus is more difficult to construct than an inking pen. It also requires a transformer for stylus voltage and a rheostat for controlling the heat. However, the heated stylus has the important advantage of making a trace that moves across the chart in substantially a straight line, instead of in an arc as conventional recording pens do. The impression is made at the point where the stylus crosses a knife edge that supports the chart. During lateral excursions the radius of the stylus arm is increased.

"An electrical stylus writing on paper sensitive to current also works well at high frequencies, but it is plagued by its own set of disadvantages. Its resolution is not as good; it requires special transformers, voltage-regulating devices and sensitized paper, and it is subject arcing and pitting.

"For propelling the chart the available mechanisms range from weight-driven cylinders such as are employed in the old Wiechert seismograph to perforated tapes moved by high-speed sprockets geared to synchronous motors. Seismograms have even been made on the smoked surface of a paint bucket turned by the hour shaft of an alarm clock. The problem of the drive has been simplified by the recent development of inexpensive synchronous motors of the fractional horsepower type.

"It is now common practice to record events at frequencies between 200 and 100,000 cycles per second on magnetic tape. Several basic tape-pulling mechanisms, together with recording, reproducing and erasing heads, are now available for less than $100. They must be appropriately housed and modified for the frequency range in which they are to work. With this type of recorder high-frequency events can be replayed in 'slow motion.' The low-frequency signal of the reproduction is fed into the electropneumatic pen motor and transcribed as a chart. Frequencies up to 100 cycles per second are recorded by means of frequency modulation, and those above this band by the conventional amplitude techniques. When driven at 60 inches per second, magnetic tape registers signals from 200 to 80,000 cycles uniformly with an error of not more than about three decibels. At high frequencies speed regulation of the tape-propulsion mechanism becomes critical. The speed is sometimes regulated by generating a 60-cycle signal from a tuning fork and amplifying it for power to drive the motor of the tape puller. The frequency stability of commercial electric power is generally adequate for recordings up to 10,000 cycles.

"Construction details of an experimental pneumatic recorder equipped with a heated stylus and a Sanborn transducer are shown on page 125. The specifications are not rigid: an ingenious experimenter doubtless will find ways to modify this plan. Aluminum stock is specified for the chassis because it is easy to cut with hand tools. For sufficient strength the aluminum should be a quarter of an inch thick. Amateurs with limited facilities may adapt war-surplus apparatus for some of the parts.

"One such item available as a starting point is a McElroy telegraph tape puller. It is equipped with a 110-volt motor drive and sells for about $20. Its aluminum casting will serve for constructing the chassis of a two-inch chart. Another basic unit is a magnetic wire recorder magazine which sells for about $18. Another piece of surplus gear, designed as an element of a photographic processing machine, looks promising as a part for a recorder equipped with a pair of two-inch charts and companion pen motors-a so-called two-channel recorder. It consists of a magnesium casting with a side compartment housing gears on detachable shafts. These could be moved around to form almost any type of gearing arrangement. It is fitted with rollers which could be modified for the chart drive. This unit is priced by my instrument maker at $8.

"The most difficult part of the peneumatic recorder to make is the pen motor. If the recorder's frequency response is not critical, a pressure transmitter made by the General Electric Company can be used when modified. It will work on a back pressure in the neighborhood of 10 pounds per square inch. My instrument maker has a limited number of these in stock priced at $10."

This department will forward the address of Marrows' instrument maker to any reader who sends a stamped, self-addressed envelope.

The imminent attempt to launch the first man-made satellite into space undoubtedly will heighten interest in the planets and satellites of the solar system and their motions. One of the most interesting refresher courses you can take on these matters is to construct an orrery. This classic model of the solar system, long a fixture of physics classrooms and planetaria, was named by its inventor, George Graham, after his patron, Charles Boy]e, Fourth Earl of Orrery. An English engineer, Frank W. Cousins, submits the design for a simple orrery pictured here [see Figure 5].

"The orrery," writes Cousins, "shows all known planets-except the minor ones -from Mercury to Pluto, with their satellites. You will find an imitation pearl necklace with beads of various sizes a splendid source of spheres to represent the planets. The scale of sizes here is based on a bead one eighth of an inch in diameter for the earth. The satellites are too small to be scaled to this standard and therefore are represented by small beads of uniform size.

"A circular scale mounted on the base gives the zodiac, the months and the right ascension from zero to 24 hours. The sun sector, attached to the tip of the central spindle, is made to scale and gives some idea of the great size of the sun when compared with that of the planets.

"The planets are mounted on arms with devices for showing their inclination to the plane of the ecliptic. In the cases of Mercury and Pluto, this is managed by means of tilted washers on the central spindle. For the earth, Mars, Saturn? Uranus and Neptune we use Z-shaped axis rods turning in deep sockets at the ends of the orbit arms.

"The central spindle and washers are of brass and the orbit arms are of silvered steel.

"As you doubtless know, we English have a custom of naming the things we make. I call this instrument the 'Urameton Orrery.' I have had a lifelong interest in the physics of astronomy, and when uranium 235 came into prominence I chanced to remember the Metonic cycle (235 lunations equals 19 years). My address is 235 Bilton Road, Greenford. I was so impressed by the coincidence that I immediately christened my observatory Urameton!"

Bibliography

DIRECT-INDICATING RECORDING INSTRUMENTS. S. R. Gilord in Electrical Manufacturing, Vol. 52, No. 5, pages 114-121; November, 1953. Vol. 52, No. 6, pages 120-128; December, 1953.

Suppliers and Organizations

The Society for Amateur Scientists
5600 Post Road, #114-341
East Greenwich, RI 02818
Phone: 1-877-527-0382 voice/fax

Internet: http://www.sas.org/