Smith-Kettlewell TECHNICAL FILE

Published by Rehabilitation Engineering Center Smith-Kettlewell Institute of Visual Sciences

Bill Gerrey, Editor

Supported, in part, by Smith-Kettlewell Eye Research Foundation and National Institute of Handicapped Research

Produced by

SUBSCRIPTIONS
Braille Edition: . . . . . . . . $15 per year
Large Print Edition . . . . . . $15 per year
Talking Book Edition . . . . . . $ 8 per year

Make checks payable to:

Smith-Kettlewell Eye Research Foundation
2318 Fillmore Street
San Francisco, California 94ll5

Please address all correspondence to: Bill Gerrey, Editor at the above address

TABLE OF CONTENTS

Operational Amplifiers

The Smith-Kettlewell Active Light Probe

The Smith-Kettlewell Receptionist Mat--An Introvert Detector

The Handling and Storage of Magnetic Recording Tape

Hints and Kinks

Lexicography

Editor's Corner

OPERATIONAL AMPLIFIERS

By Albert Alden

Abstract

This is the first in a series of articles on the subject of operational amplifiers (Op Amps). This article will describe the ideal errorless amplifier and give some applications. Future articles will treat the limitations of practical amplifiers, DC errors, dynamic characteristics, and additional applications.

Introduction

The Op Amp is a high-gain differential amplifier circuit (these days almost exclusively an integrated circuit) with two input terminals, two power terminals, and an output plus a couple of other connections to be described later. The symbol for an Op Amp is a triangle, usually drawn with the base vertically on the left. The input terminals are on the base with the one labeled "-" above the one labeled "+". The output is at the apex on the right. A positive input signal on the + terminal with reference to the - terminal will produce a positive output signal and vice versa. We shall henceforth call the + terminal the non-inverting input and the - terminal the inverting input. Two power supplies are usually used, plus and minus l5 volts being a typical value. It should be noted that no power ground connection is made to the Op Amp. Ground (i.e., the common connection of the plus and minus supplies) is used as the reference for the input and output signals.

The properties of our ideal Op Amp are:

The consequences of each of the above are as follows.

Although you can't buy an Op Amp with the parameters described above, analysis based on them is usually adequate for an understanding of the working of an Op Amp circuit.

In the majority of applications of Op Amps, negative feedback is employed. The feedback is realized by the connection of a resultor from the output back to the inverting input. This is called the feedback or output resistor. With the non-inverting input grounded, any deviation of the output from ground will feed back through the resistor a signal into the inverting input with the proper polarity to cause the output the correct to zero. Or conversely the feedback forces the signal at the inverting input to be at ground. This is called a virtual ground: it is at ground potential but not connected to ground.

With the virtual ground concept and the zero input bias current feature of the ideal Op Amp, we can describe our first application.

The Inverting Amplifier

The circuit of the inverting amplifier is as follows. There is a feedback or output resistor (Ro) from the output to the inverting input as described above. The non- inverting input is grounded. An input resistor (Ri) is connected to the inverting input; the other end is the input terminal of our circuit.

If we apply a positive signal (Ei) to the input resistor, a current I = Ei/Ri will flow. Remember the other end of Ri is at virtual ground and therefore the voltage across it is the input signal. Since no input current flows into the input terminals, it must flow through the output resistor. The negative feedback and infinite gain of the Op Amp forces the output voltage to be exactly the correct value to "accept" this current; that is, Eo = -IRo. Notice the minus sign. From the above two equations we get Eo = -Ei (Ro/Ri). The relationship between the output and input is a function only of the ratio of the values of output and input resistors (with a negative sign).

A good way to think of the inverting amplifier is to use the analogy of the lever. The fulcrum is the virtual ground. The lengths of each end are proportional to the value of the resistors. Lifting the input end (a positive voltage) causes the lever to rotate about the fulcrum (virtual ground) and the output end to go down (a negative voltage). The longer the output end, the greater the output movement (larger gain). The infinite gain in the amplifier corresponds to a rigid fulcrum in our analogy.

An extension of the inverting amplifier circuit is the summing amplifier. To the inverting amplifier circuit we can add any number of additional input resistors connected to the inverting input of the Op Amp. The free ends of these resistors constitute new inputs to the amplifier. Each resistor contributes a current In = Ein/Rin that passes through the output resistor and thus adds to the output signal. The general equation is

Eo = -{Ei1 (Ro/Ri1) + Ei2 (Ro/Ri2) + Ei3 (Ro/Ri3) + ...}

The contribution of each input is determined by the ratio of the output resistor to its input resistor. Note, however, that the gain of all inputs is controlled by Ro. This ability to adjust individual gains and overall gain is very useful.

The Non-Inverting Amplifier

We shall now describe the non-inverting amplifier. This circuit retains the output resistor Ro, connected from the output back to the inverting input. The input resistor Ri is connected to ground. The non-inverting input is not ground but becomes the input to the circuit. The output resistor connection and the infinite gain of the amplifier causes the voltage at the inverting input to be identical to that at the non-inverting input. There is no virtual ground because the non- inverting input is not grounded, but there is a virtual signal input voltage at the inverting input. This virtual signal Ei (equal to the actual input signal) causes a current I = Ei/Ri to flow through Ri to ground. This current comes from the output of the Op Amp. The output must be such that Eo = I (Ro + Ri). Substituting Ei/Ri for the current I, we get Eo = Ei ((Ro + Ri)/Ri) or Eo = Ei ((Ro/Ri) + 1).

Notice two differences between this and the inverting amplifier equation:

  1. there is no negative sign.
  2. there is a +1 appended to the resistor ratio.

If Ro is shorted and Ri eliminated, the gain is unity, and we have a "follower."

Another way of looking at the non-inverting amplifier is as a backwards attenuator. The output and input resistors are connected as an attenuator where the output of the amplifier must adjust itself to give a voltage at the resistor junction (inverting input) equal to the input signal (non- inverting input). Writing the equation for an attenuator made from Ro and Ri and swapping Eo and Ei in the equation gives the equation for our non-inverting amplifier.

In our lever analogy, one end of the lever is fastened to the earth, the input is the fulcrum, and the output is the free end of the lever. As before, lengths of the lever correspond to resistor values.

Input impedance of Inverting and Non-inverting Amplifier Circuits

The input impedance of the inverting circuit is equal to Ri, while that of the non-inverting is equal to the input impedance of the Op Amp itself. This is infinite for our ideal device.

The difference amplifier

As before, the output goes through Ro to the inverting input, while Ri goes from this inverting input to one input signal. On the other hand, the non-inverting input goes through a resistor to ground (equal to Ro), while this non-inverting input also goes through a resistor equal to Ri to another input signal. In other words, we are simply adding a resistor equal to Ro from the non-inverting input to ground plus a resistor equal to Ri connected to the non-inverting input. There are now two input points and the output is equal to the difference of these two signals times Ro/Ri or Eo = (Ei1 - Ei2 (Ro/Ri). This formula may be derived by shorting one input, calculating the output due to a signal on the other input, repeating for the opposite case, and summing the two expressions. This is possible due to the law of superposition.

The derivation is left to the reader. Hint: One situation is treated like the inverting amplifier and the other like the non-inverting amplifier, but with the signal being attenuated before getting to the non- inverting input to be amplified by ((Ro/Ri) + 1).

741 op-amp

In order to try some of these circuits, here are the pin connections for the popular 74l op-amp.

The offset null connections are not used in this discussion. As mentioned previously, there is no power ground connection. The common connection between the two power supplies is the input and output signal ground.

Resistor values in the 3.3K to 47K range should be used as starters for experimentation.

Next issue: More applications and the real world.

THE SMITH-KETTLEWELL ACTIVE LIGHT PROBE

Abstract

The described unit is a pocket- sized light probe with a self-contained light source. It can be used as a "passive" detector of external light sources, or it can "actively" detect reflections of its own internal source. The latter mode permits it to be used as a print detector and for reading gauge pointers behind a glass face. The unit has a sensitivity control, allowing it to be adjusted for a wide variety of ambient light conditions. Although it is commercially available, the circuit has such versatility as to afford "special purpose" probes to be constructed, thereby enhancing its value for typists and experimenters. Modifications are included so that the circuit will provide sufficient power for tactile output, thus making it usable by the deaf-blind.

A Sketchy Background-- The Editor's Perspective

Somewhere (where I can no longer find it) is the mention of a Polish inventor who, in 1879, devised a light detector for use by the blind as a mobility aid (as I remember, his name was something like Noisiewski). Your Editor first got his little hands on a "light probe" at the ripe age of 4 years. In 1951 my father, who was also blind, contracted with a local "inventor" to build a detector which he could use in turning off the various lights and displays in his piano store--one which he could also try as an aid to mobility. This instrument was a neon relaxation oscilla- tor controlled by a selenium photoresistor. This piece of handiwork cost him $35, the crime of which is alone enough motivation to run a technical magazine to reduce the number of such instances.

By the late l950's, the work of Swail and Gunderson, as well as developments at the Royal National Institute for the Blind in England, brought forth a number of hand-held probes "containing even the whole battery!" Throughout the 23-year course of the Braille Technical Press, several units containing their own light sources were described by Gunderson, Swail, and Earl Quay; these could be used for the detection of print on a page and for even looking through a glass crystal to find a pointer beneath.

Until the early l970's, "active" light probes (those containing a light source as well as a "passive" light sensor) were made by carefully positioning a lamp next to the sen- sor, either or both of which also required a condensing lens. Soon, however, all-new solid-state sensors were devised for industrial use--made for guiding rolls of material, counting rotations of shafts which were scribed with a line, and reading holes in punched cards. These sensors made construction of the probes' "front ends" much easier.

With the advent of these new sensors, two more "inventors" came on the scene: Southwest Research Institute in Texas brought out the "paper money identifier;" it was a fine active light probe, but it was marketed for the wrong use, in addition to which its fancy wooden cabinet kicked the price well up over $100. Bill Gerrey, on the other hand, built Gunderson's ol' Transistorized Auditory Gimmick around the new sensors, and then set about finding industrial uses for this very inexpensive and almost pocket-sized instrument. (I had it doing everything from reading non-electrical gauges to positioning holes in fabric in the jaws of an eyelet- inserting machine.)

It then became the task of "inventors" (in number now rivalling those who created the lights we were detecting, including the inventor of the candle) to make the instruments small and inexpensive. The smaller we made them, the more expensive were the components and the more intricate was the machining of their cases. I may hold the record in this antiquity; my final design was only 4-1/2 inches long, l inch in diameter, had an $8 switch, a $l0 speaker, a $6 battery, and just a little machining . . . bringing its price well up to $l00 along with the rest.

By l979, the 100th anniversary of the original Polish invention, production engineering was completed on a unit which broke both the size and cost barriers. Bill Loughborough, at Smith-Kettlewell, eliminated the tubes and housings we were grappling with--he glued the parts together and dipped the assembly in tool-handle plastic, leaving the battery to stick out and serve as a handle. His is the only unit I have ever seen being routinely carried in people's pockets, and its cost is little more than what my father paid for its primitive counterpart 30 years earlier.

Description and Operation of the Production Model

The assembly of parts is done on a small PC board which is the same size as the cross- section of the 9-volt battery which powers it. The light sensor is glued to the back of the speaker, which is then cemented to the top of the 555 timer and other components at one end of the board. Next, the sensitivity control is positioned back-to-back with the speaker and glued to the tops of the components on the other end of the board. Finally, a pair of battery snaps is glued up under the bottom of the board.

Vulnerable parts of the assembly are then covered with tape in preparation for dipping the unit in self-curing plastic (Plasti- Dip*1). The assembly is snapped to the top of a dead battery, turned upside down and immersed so as to cover perhaps one-third of the battery. This procedure is repeated three times.

The resultant instrument is about the size of a cigarette lighter, with the assembly of the probe protruding about one inch beyond the top of the battery. A hole for the speaker is cut through the plastic on one side and a control knob for the sensitivity control is fitted to the potentiometer shaft on the opposite side. Plastic is cut away on the very top of the probe to expose the end of the sensor, which sticks out slightly beyond the speaker.

The sensor used is a so-called "line finder" (Spectronics SPX1404); it is slightly concave so as to accommodate a rotating shaft onto which a line has been scribed--its industrial use being that of an optical tachometer. Its phototransistor and its LED source are pointing slightly toward each other so as to intersect at a focal point a short distance ahead of the sensor. (The manufacturer claims this distance to be about l/2 inch, although it is nominally shorter than this and varies considerably between units.)

The least expensive pot we could find was used; it is a typical panel-mount unit with a split shaft designed to fit a control knob. On the light probe, this shaft has been cut so as to protrude only 3/l6 inch beyond its bushing. A special knob is machined for the resultant oddball diameter and length of the shaft. A fillister head set screw was chosen so that the screw could be used as a "pointer."

Operation

Whatever the task, I advance the sensitivity control to the point where the frequency of oscillation levels off (the point of saturation), after which I back the control down to a point where I get a moderate frequency. This latter setting is chosen so that changes in light level will bring about a pitch change in either direction. The more light which the sensor sees, the higher the frequency of oscillation, and vice versa.

On the other hand, in the passive mode a higher sensitivity will have the effect of broadening the angle of acceptance to which the probe will respond. Therefore, if you're looking for light in the middle of nowhere, turn the probe all the way up and look around till you find it; backing off on the sensitivity will then allow you to pinpoint the source exactly.

The fact that the LED and the phototransistor are pointing so as to intersect ahead of the sensor enables this device to look through a thin piece of glass; when the sensor is up against the glass, specular reflections from the glass surfaces are out of position so as to miss the phototransistor. In this way, the probe can be used to detect a meter pointer behind the glass cover, or you can detect the level of a liquid in a thin glass vessel. For best sensitivity to changes in reflectance of the items behind the glass, be sure that the sensor is up against and perpendicular to the glass surface and that the control knob has been adjusted to give you an intermediate frequency.

The "focal distance" of the sensor does mean that when you are detecting print on the surface of a sheet of paper, the probe must be spaced some distance away--perhaps l cm. This spacing must be held absolutely constant; otherwise you will hear variations in pitch that are due more to the surface coming in and out of focus than to changes in reflectivity.

I use my thumb and middle finger to pinch the probe at its forward end, thus using the ends of these fingers to space the probe at a constant distance away from the surface. You could, if you were inclined to, fit the end of the sensor with a tubular spacer. However, for the production model, this constituted machining a special part, and we had all fallen victim to that roadblock before.

Notes on Uses

Both the probe itself and its instructions (in print and in braille) can be gotten from the San Francisco Lighthouse for the Blind.*2 These instructions list household and office uses--checking for lights which are left on, checking to see that your typewriter ribbon is working, finding the letterhead on company stationery, etc. I intend only to supplement this material in this article and gear it more toward our high degree of sophistication.

A tape can be gotten from Harvey Lauer at the VA Blind Center*3 describing the use of the probe for identifying paper money. This process is slow; it is good for when you are truly on your own and perhaps as a parlor trick. (Sorry, Harvey. It is a good tape, though.)

Uses from the routine to the bizarre, such as adjusting the flame in your fireplace, are being documented by the Carroll Center for the Blind in Newton, Massachusetts*4 and will appear in their publication, "Aids and Appliances Review." This compilation will also list the numerous different brands of probes which are available.

"Science for the Blind,"*5 sells both active and passive probes built around their Audicator, as well as selling instruction tapes having live demonstrations of their uses.

Manufacturers and employment uses (up through l978) can be found in the Sensory Aids Foundation catalog, "Sensory Aids for the Employment of Blind and Visually Impaired Persons," available from AFB.*6 (The Smith-Kettlewell units listed there are Bill Gerrey's antiques and are not to be confused with the unit described here.)

It must be remembered that with the new solid-state sensors, the light source is infrared. Therefore, the reflectivity of surfaces may not correlate with what is seen visually. For example, glossy black surfaces can look very reflective indeed.

Very often, you can identify wires by their reflectivity. Different color wires will cause the unit to emit different pitches provided the distance from the sensor is carefully controlled. The red and black binding posts on various lab instruments are a cinch to identify, and I rarely bother reaching for a voltmeter to do so.

Though there is no substitute for having your own adapted instruments which you can truly read, it is sometimes handy to reach over to the front panel of an instrument using a light probe to get a relative indication from its meters. I see this as a crude shortcut, not a legitimate procedure. This system absolutely breaks down if a meter is illuminated from behind. Also, some meters have a little strip of mirror behind the pointer which you must either avoid or carefully trace with the probe; if you do use the mirror, the probe sensitivity will have to be turned way down.

I have had occasion to use active light probes in reading mechanical gauges which cannot be read with a meter reader. For example, in a pinch, I can get relative indications off a visual vacuum gauge used in repairing player pianos. I much prefer, however, getting out my trusty electronic version (see our Vocational Aids Catalogue) for getting an accurate reading.

If you ascertain that the light probe is the best way to read a given gauge, you can fashion an attachment and design a special probe for this specific purpose. For example, I could adapt my vacuum gauge by gluing a pivot and a movable hand right in the middle of the glass crystal, then mounting a Spectronics sensor on the side of the hand so that it can trace the course of the gauge needle. Finally, an outer disc containing braille markings can be fitted around the gauge, against which the position of the hand is read. In operation, the hand is turned until a dip in pitch occurs, at which point the sensor is positioned over the needle.

When it comes to checking your typing with the probe, there are times when the tiny sensor on the end of a cable would come in handy. With practice, such a specially designed unit can be used to reposition the paper into the machine so as to continue typing on the last line.

Incidentally, it is possible with practice to determine whether the last word typed was long or short. This is indeed useful in case an unforgivable dunderhead calls you on the phone and interrupts your work. With the probe, for example, you can determine whether or not the last word you typed was a conjunction or not. Once again, the small sensor on a cable can be braced against the type fork and the carriage manipulated to perform this test more easily.

Circuit Operation and Description

Although I highly recommend that you try one of the production models (I have bought a couple myself), this will give you the chance to make one to your own specifications. For example, you may wish to preset its sensitivity with a screwdriver adjustment. Our deaf-blind readers can change a couple of components to get a low frequency and enough output to be suitable for tactile indications.

Circuit Operation

The internal light source, the LED in the Spectronics sensor, is infrared and cannot be seen visually. The LED is operated in the conventional way through a current-limiting resistor.

The oscillator in the unit is a 555 timer which directly drives the tiny speaker through a resistor. (A normal-sized speaker could be used; because of its lower impedance, it should be driven through a resistor of perhaps l00 ohms.) The charging current for the timing capacitor is controlled by a "current amplifier" transistor, the base of which is controlled by the phototransistor.

The phototransistor is properly seen as a current source. The resistor network in its collector is best termed a variable "current divider;" the sensitivity control shunts current away from the network controlling the base of the current amplifier.

Circuit Description

The negative side of the battery goes through the on-off switch to ground. (In the production model, this switch is a part of the sensitivity control.) The positive of the battery goes directly to the VCC line. The cathode of the LED is grounded, while its anode goes through 390 ohms to the VCC line.

The emitter of the phototransistor is grounded, while its collector goes through two branches to the VCC line: First, the collector goes through the sensitivity control (500K rheostat), then through 390 ohms to VCC. Second, the collector goes through l.8 megohms, then through l megohm to VCC. The junction of the latter two resistors goes to the base of the PNP transistor which is the current amplifier. (This could be any PNP silicon transistor such as a 2N2907--we used 2N5l39.)

The emitter of the PNP transistor goes through 390 ohms to VCC. Its collector goes through l2K, then through 0.0luF to ground, with the junction of this resistor and capacitor going to pins 2 and 6 of the 555. Pin 7 of the 555 goes directly to the collector of the PNP current amplifier.

Pin l of the 555 is grounded, while pins 4 and 8 go to VCC. Pin 3, the output, goes through the speaker, then through 390 ohms to VCC. (The speaker used here is Panasonic No. EAFl2R0lA.)

For our deaf-blind users, the speaker or transducer can be driven through 47 ohms instead of 390; then increase the timing capacitor and/or the l2K resistor in series with it. Initially, try a timing capacitor of 0.1uF with its series resistor being increased to perhaps 39K. It will take some juggling of these components to suit your taste and your transducer.

When the frequency is slowed down for tactile output, the dynamic range around an intermediate frequency is reduced, making adjustment of the sensitivity control very critical. For this reason, you may wish to use a l0-turn 500K pot in place of the garden- variety 270dg unit.

The task remains to identify the leads on the Spectronics sensor. Two leads emerge from each element in the package. The pairs of leads cannot be confused, since the elements are separated by a little mounting flange. The LED will look like a diode on your continuity tester; the tester will sing when its positive lead is on the anode. As for the phototransistor, proper connection to your continuity tester will create a system which acts like a light probe (which it is). This will occur when the tester's positive lead is on the collector and its negative lead is on the emitter (the base of the transistor is not available).

*1 PDI, l458 West County Rd. "C", St. Paul, MN 55ll3, (6l2) 633-9633.

*2 San Francisco Lighthouse for the Blind, 1155 Mission Street, San Francisco, CA 94l03, (4l5) 43l-l48l.

*3 Mr. Harvey Lauer, Blind Center ll7A, VA Hospital, Hines, IL 60l4l.

*4 Carroll Center for the Blind, 770 Centre Street, Newton, MA 02l58.

*5 Science for the Blind, P.O. Box 385, Wayne, PA l9087.

*6 American Foundation for the Blind, l5 West l6th St., New York, NY l00ll.

Parts List

Resistors, W1/4, 5 percent:

Potentiometer with switch:

Capacitor:

Semiconductors:

Speaker:

THE SMITH-KETTLEWELL RECEPTIONIST MAT--AN INTROVERT DETECTOR

Abstract

Whether you are stationed in an office or taking your turn at manning a booth, much energy goes into wondering whether or not a non-vocal person has approached you. This unobtrusive device does nothing until someone has been standing at your station for a few seconds, after which a quick little "beep" alerts you to his presence.

Introduction

Whether it's at a cake sale or a university registration desk, we have all been asked to put in our time in "personing" the booth. About every 20th person, you get someone who simply cannot bring himself to address you first. This leads to your frequent address of mirages in order not to miss anyone-- "Hello?" you mutter nervously.

Well now, with this device you can content yourself with "woolgathering" until a bashful soul has been caught in your snare, at which point you can say, "Yes?...Well go ahead and be that way!" And you can do so with confident abrasiveness.

The system uses a "mat switch" of the kind used for automatic doors and burglar alarms. A selection of mats is given at the end of this article, including good-looking ones for on top of the floor and switch arrays which go under the carpet.

Two sections of a 556 dual timer chip are connected as one-shots in cascade. The first of these sets the length of time that a person has to stand there before he gets caught, while the second sets the duration of the beep. The beep itself is generated by a Star Micronics buzzer.

The unit can easily be built into a small minibox, which is then laid on the desk or taped up under it. The buzzer can even be shut up inside the box if the environment is fairly quiet.

It was necessary to play a trick on the second one-shot; otherwise they would both trigger as soon as the switch were closed. Its timing capacitor is not returned to ground, as is usually done; it is brought over to the battery side of the mat switch so as to put a charge on this capacitor for the initial state. In this way, the second one-shot thinks it has already just been triggered; its discharge pin immediately sets about discharging the timing capacitor in order to initiate "another" cycle.

Note that in the one-shot configuration, the Threshold and Discharge pins are often directly tied together--something you never see in the free-running oscillator connection. Thus, in the first section these pins are tied together, pins 2 and 1 respectively. However, in the second section it was necessary to buy a little time in the "false post- triggered state" mentioned above to keep this section from firing early.

Circuit

The positive of the 9-volt battery goes directly to the VCC line, while its negative side goes through the mat switch to ground. Pin 7 of the 556 is grounded, while pins l4, 4, and l0 go to VCC.

Pins l and 2, the Discharge and Threshold, are tied together; they go through 5uF to ground (negative at ground). Pins l and 2 also go through l00K, then through a l meg rheostat (delay control) to VCC. To trigger this one-shot, pin 6 (trigger) goes through 0.luF to ground, as well as going through a l00K pull-up resistor to VCC. The output of this section, pin 5, goes through 0.luF to the trigger (pin 8) of the second section. Pin 8 also has a l00K pull-up resistor going from it to VCC.

Also in the second section, pin l3, the Discharge, goes through 47K to pin l2, the Threshold. Pin l2 also goes through 0.47uF to the negative side of the battery, not ground. Pin l2 goes through l00K, then through a l meg rheostat (beep duration control) to VCC.

Pin 9, the output of the second one-shot, goes to the control pin, pin 8, of the Star Micronics CMB-l2. Pin l of the buzzer is grounded, while pin l4 goes to VCC.

Parts List

Resistors, W1/4, 5 percent:

Potentiometers (connected as rheostats):

Capacitors:

Semiconductors:

Mat Switches

THE HANDLING AND STORAGE OF MAGNETIC RECORDING TAPE

(The following was gotten from BTP February 1971; it was originally reprinted from "Sound Talk," a service to the industry from the makers of Scotch Magnetic Tape, Vol. III, No. 1, 1970.)

Abstract

Heavily excerpted, this material covers all the "right things to do" to preserve your recordings. As was suggested in the original article, pick out the ideas which you can economically afford and which do not interfere with your everyday use of magnetic media. Although this article pre- dates floppy discs, the ideas presented here are relevant to them.

The Recording Area

Ideally the equipment room of a recording studio should approach, as closely as possible, a "clean room" environment. By definition, a "clean room" is characterized by the absence of normally expected airborne dust and lint. The integrity of this area should be maintained by periodic cleaning of shelves and floors. When vacuum equipment is used for cleaning, the exhaust from this unit should be located outside the room.

It is doubtful that smoke will contaminate the tape, but ashes can. Therefore, smoking should not be allowed directly over the machines or when handling tape. Food and drink should also be prohibited for obvious reasons.

The equipment area should be such that reasonable control of temperature and relative humdidity can be exercised. Variations in temperature should be held within plus or minus 5 degrees F. of a preselected value, and variations in relative humidity should be kept to within plus or minus 10 percent. Preferably the temperature should be in the 70's with a relative humidity of 40 percent.

When recording on location or at home, it may be difficult to control the surrounding environmental conditions. At the very least, it is important to eliminate the entry of foreign material into the machine. It is recommended that the equipment always be covered during storage, and as much as possible during operation. Many of the protective dust covers provided by manufacturers permit their machines to be operated while they are in place; these should be used when in an uncontrolled environment.

Tape Storage

The temperature and humidity of the tape storage area should closely approach that of the work area. The smaller the environmental change experienced by the tape, the better will be its operation and reliability. As a general rule, a temperature between 60 and 80 degrees F. and a relative humidity between 40 and 60 percent is recommended. If the environmental conditions of the storage area vary widely from the recording area, allow time for the tape to reach temperature and humidity equilibrium before putting it into use.

Recording tape, especially cartridges and cassettes, poorly stored or casually laid on the dashboard or in the glove compartment of an automobile, can be damaged by the heat of strong sunlight. The molded cases used for some cartridges and cassettes can be permanently distorted if subjected to high temperatures. Both cartridges and cassettes use splices within their tape rolls which can be affected by heat. The splices may separate, or the adhesive may soften and "ooze" from the edges of the splice, causing adjacent layers of the tape to stick together. The exposure of the splice adhesive will also collect any contamination present in the case, causing additional problems.

Protection from accidental erasure while in the storage area is easily accomplished and is, ironically, of little concern. First of all, fields strong enough to cause erasure are just not normally found in an office or home atmosphere. Secondly, if the tape is kept as little as three inches away from even a strong magnetic source, this spacing should be sufficient to offer adequate protection. {Editor's Note: With the advent of very flat loudspeakers, fields from their permanent magnets are often strong enough to damage cassettes or credit cards. Keep magnetic storage media away from your talking clocks, "Mr. Thin" radios, and light probes.}

During storage, the tape must be enclosed in a container (original box, plastic case, tape cannister) for several reasons. One reason is to provide protection from physical damage. Another reason is to protect the tape from dust.

The closed containers should be placed into storage on edge, so that the reel is in an upright position. While they may also be stored individually, lying flat, tape boxes should never be stacked so high that there is a possibility of crushing or distorting the bottom container from the excessive weight of the stack; this could cause edge damage to the reel of tape in the bottom container.

For long-term storage, additional protection from dust and moisture can be gained by sealing the container in a plastic bag. It is generally considered good practice to clean the container before using it, so that dust which has accumulated during storage will not contaminate the recorder or tape.

Of primary importance is the way the tape is wound on the reel, since poor winding can result in distortion of the tape's backing. A winding tension that is relatively low is recommended. Three to four ounces per l/4 inch of tape width is sufficient to render a firm, stable wind on a reel configuration. This tension, while great enough, does not result in high pressures within the roll that could permanently distort the backing. Backing distortion, caused by extreme pressures within the tape pack, may result if a roll of tape which has been wound too tightly is subjected to an increase in temperature while in storage.

Too low a winding tension can cause difficulty as well. If the wind is too loose, slippage can occur between the tape layers on the reel. This "cinching," as it is called, can distort the tape by causing a series of creases or folds in the area that has slipped. When the roll is unwound, the backing will be wrinkled. When an attempt is made to use the tape again, the wrinkles and creases will disturb the necessary intimate contact between the tape and the head. Because the tape is repeatedly lifted from the head, the result will be a series of signal variations. If the tape is properly rewound immediately after cinching, there is a good possibility that the information may be saved.

Some recorders now in use do not have a method of adjusting winding tension; therefore, care must be taken while operating these machines. Sensible operation of fast- forward, rewind, and start controls can eliminate the sharp stress loading associated with starting and changing the tape direction. Tape distortion and "cinching" can be reduced by allowing a minimum of slack when threading and starting the machine. It is also good practice to allow the spinning tape reels to completely stop before changing tape direction.

Along with proper attention, another important consideration is quality of the wind. The successive layers of tape should be placed on the reel so that they form a smooth wind with no individual tape strands exposed. A smooth wind offers the advantage of built- in edge protection. A scattered wind will allow individual tape edges to protrude above others. Since there is no support for these exposed strands, they are vulnerable to damage.

It is sometimes suggested that tapes in storage be rewound at specific intervals, such as every 6 or l2 months, to relieve internal pressures. This may be recommended for tapes of marginal quality or for those with other than heavy-duty binders for securing the oxide coating. For modern tapes with polyester backing and advanced binders, this periodic rewinding may not be necessary.

A good practice, however, is to select a random sample from various areas of the library for visual inspection. The reels chosen can be examined for loose windings and dust accumulation. They should be checked for rippled edges and other signs that indicate the presence of physical distortion. If anything is found that indicates a problem may exist, additional samples should be inspected to ascertain what percentage of the library may be effected.

If the above recommendations concerning the storage environment and the actual preparation for storage are followed, no serious problems should be encountered, even in long- term storage.

Shipping of Tapes

There are certain precautions which apply to the shipment of recording tapes that should be followed to ensure safety in transit. Logically, the first consideration would be the physical protection of the tape while being transported. Also, since the tape carries magnetic information, measures must be taken to protect the reels from accidental erasure.

The outer shipping container into which the tapes are placed must have the necessary strength and rigidity to protect the tape or tapes from damage caused by dropping or crushing. While a container which is l00 percent watertight would not be necessary, it must nevertheless provide a reasonable degree of water resistance. It should, for example, be capable of protecting its contents from being damaged when left on a loading dock in the rain.

While it is always good practice to secure the free end of a reel of tape, this is particularly important when preparing reels for shipping. A short length of pressure- sensitive tape is all that is necessary. {Editor's Note: For many years, plastic clips have been available which hold the free tape end and which slide in between the flanges of the reel and the tape pack. I consider these devices mean, nasty, and barbaric, since their very insertion and their presence as the package is crushed causes serious edge distortion.}

As far as erasure of the tape is concerned, laboratory-conducted tests have determined what would constitute adequate protection from stray magnetic fields of a magnitude which may possibly be encountered in transit. It was found that field strengths within the tape of 50 Oersteds or less caused no discernible erasure. The average bulk degausser, on the other hand, produces a field of l500 Oersteds.

Sources of magnetic energy to which tape being shipped might be subjected would be motors, generators, transformers, etc. These devices are designed to contain their magnetic fields to accomplish some type of work. Bulk degaussers, on the other hand, are designed to produce a maximum external field that is strong enough to erase a tape while it is still on the reel. It is safe to assume that field strengths from other devices would not be more than l500 Oersteds.

Because field intensity decreases rapidly with distance from the source, the 50 Oersted point (mentioned earlier as not affecting the tape) is reached at a distance of 2.7 inches from a l500 Oersted source. From this it can be seen that the easiest and least costly method of obtaining erasure protection is by ensuring a 3-inch spacing between the tape and the magnetic source. It is suggested that tape being prepared for shipping be packed with bulk spacing material such as wood or cardboard between the tapes and the outside shipping container. This magnetically protective spacing can also be justified because of the excellent protection gained against physical damage to the package contents.

Recent laboratory tests concerning exposure of recorded tapes to X-ray have determined that the recorded signal is not affected by even severe exposure to this source of radiation. The tests involved a commonly used recording tape with several different frequencies recorded on it.

The X-ray machine was operated with 200mA at 110kV for a 6-second exposure time at a 36-inch distance. Testing and measuring the signal output before and after exposure indicated no signal loss or degradation.

Tape in transit may be subjected to temperature extremes. Temperatures as low as -40F might be encountered in the cargo hold of high flying aircraft. A temperature of l20F could be encountered in a motor vehicle in the summer sun. It must again be emphasized that all incoming tape should be allowed to reach environmental equilibrium before being used.

Good Operating Habits

The container in which tape is stored is probably the cleanest space in the recording studio; of course, this is the reason that tapes should remain in their boxes until they are ready for use. To maintain the cleanliness of the container, it should be closed when the tape has been removed.

The hub is the strongest and most stable part of the reel. Always handle the reel by the hub and not the flanges.

When handling tapes, use utmost caution to ensure that the tape does not become contaminated by fingerprints. Simply stated, fingerprints are nothing more than deposits of body oils and salts. These oils will not attack the oxide-binder system, but they will form excellent adhesive areas for dust and lint.

Fingerprints on the backing are just as serious as on the coating because dirt deposits will transfer from the backing of one wrap to the coating of the next wrap on the reel. When a reel has been contaminated in this manner, the tape deck itself can be affected and will spread this contamination to other clean reels of tape that are used after the dirty reel.

The above gives reason for visually inspecting the tape machine after each roll of tape has been run to determine if cleaning is necessary. If the machine becomes contaminated with dust or wear products from the tape, complete contamination of an entire roll of tape can easily be the result. Contaminants can collect on heads and guides and be dumped along the backing or coating surface of the next tape. This contamination will then be wound into the reel under pressure, causing it to adhere firmly to the surface. Each one of these deposits will appear as a "drop-out" or group of drop-outs the next time the tape is used.

Tape contamination caused by fingerprints can be reduced by remembering not to touch the tape unnecessarily. Frequent cleaning of the tape deck will reduce the chance of spreading contamination from one reel of tape to others in the library. A cotton swab or lint-free pad moistened with "Genesolve-D" (an Allied Chemical trademark) or "Freon TF" (a DuPont trademark) or similar cleaner is recommended for cleaning all components along the tape path. If other types of cleaning agents are used, they should be given time to thoroughly dry before loading the tape; this will prevent damage should the cleaner have any tendency to attack the magnetic tape. Accumulation of tape wear products on the transport can be largely eliminated by using high reliability tape.

Maintaining reel integrity cannot be over- emphasized since valuable information can be lost, not because of tape failure but because the tape was distorted by a faulty reel. Empty reels should be thoroughly inspected and cleaned before winding tape on them for storage. Reels with hub damage, such as a plastic burr, or with dirty hubs, can cause tape distortion exactly as outlined in the preceding paragraphs.

The effect of a broken or cracked flange is easily noticed, since the tape will exhibit a series of nicks or mutilated areas along one edge, and the cause can be easily detected because of the obvious defect in the reel. A bent or distorted reel, however, can also cause damage to one or both edges if the tape is allowed to rub against the flange being used.

Not only can edge damage be caused by a defective reel and by improper storage, but a similar type of damage will also occur if any of the tape machine components are misaligned. Any of these faults can result in a complete failure of a roll of tape. Of course, damage to the edge will result in loss of information on the edge track. Further, however, the debris generated from the edge damage can be redeposited onto the remaining surface across the entire width. An examination of the edges of a tape that has been damaged in this manner will disclose an accumulation of oxide debris.

While edge damage is serious, it is sometimes difficult to ascertain its cause or even to notice its effect until the damage is severe. Operators must acquire the habit of physically inspecting the tape machine in the area of its guides and heads for an excessive build-up of oxide or backing debris. This is generally the first clue that something is wrong. Excessive drop-outs on an edge track or a loss of high frequencies may also indicate that an alignment or tracking problem exists.

It is also good practice to observe the physical condition of the tape. A sure sign of developing edge damage would be a lip or distortion on the edge being injured. When wound on the reel, the effect of this lip will be cumulative and can stretch the backing. The stretched backing will be rippled and will not conform to the recorder heads the next time the reel is used.

If tape in this condition is properly rewound immediately before being put into storage, it may be possible to salvage the roll. If this is not done, the backing will be permanently stretched and will not recover. This will result in the entire roll having to be discarded.

Major Catastrophe

The discussion, to this point, has been devoted to precautions and suggestions involving the day-to-day routine use of recording tape. The final area of concern, while a remote possibility, is nevertheless of utmost importance because it affects not only just a single reel of tape but the entire recording library. This section will be devoted to two forms of major catastrophe: fire and nuclear radiation.

Fire Damage

For a substance to burn, there must be a breakdown of the organic materials contained in it. The organic materials in magnetic tape are the plastic backing and the binder. To burn, these must first vaporize-- thus increasing their exposure to the oxygen in the atmosphere--and then rapidly oxidize, producing light and heat. An ample supply of oxygen is required to sustain burning.

Since magnetic tape contains no "built-in" oxidizer, it cannot burn in the absence of air. Simply stated, its behavior can be closely compared to the way in which a tightly wound roll of paper would burn.

While the "self-ignition" temperature of polyester-backed tape is in the neighborhood of 1000 deg. F., temperatures below that point can still cause damage. Polyester film will shrink 1-1/2 percent at 300 deg. F. and 25 percent at 325 deg. F. Acetate film, because of its sensitivity to heat, will exhibit greater shrinkage and backing distortion, and is more susceptible to heat damage than polyester. If a roll of tape is heated to the approximate temperatures listed below, certain effects will be noted when the roll has cooled.

When charring occurs, the tape cannot be unwound from the reel, since it will flake when touched. The temperature limitation of present-day tapes is a function of the organic components, and not a function of the gamma ferric oxide.

Winding and storing magnetic tape properly will lessen the possibility of damage in the event of fire, since tape is a poor conductor of heat. It is sometimes possible to recover information from a tape receiving slight fire damage by carefully rewinding it at minimum tension. The information it contains should be transferred immediately to another reel of undamaged tape.

We recommend CO2-type fire extinguishers for combating burning magnetic tape. CO2 is clean, and this type of extinguisher contains no chemicals or powders that could harm the tape.

{Editor's Note: I just called our Fire Department to ask if CO2 extinguishers were even made anymore. The Captain assured me that they were, but that they are generally inefficient in fighting fires; although they are ideal for this purpose and for electrical fires. He recommended having both CO2 and powder type on hand, or planning carefully what kind of conflagration is to be set.}

If water reaches the tape, it will probably not cause complete failure, but there may be some evidence of "cupping" or transverse curvature. The amount of "cupping" will depend on the quality of the wind, backing material, and the length of time the roll is exposed. If the wind is loose or uneven, the water can more easily reach the oxide surface and the cupping will be more pronounced. The tape should be removed from the water as soon as possible--certainly within 24 hours.

After removal, the rolls should be allowed to dry on the outside at normal room temperature and then be rewound a minimum of two times. This rewinding will aid the internal drying and will also help the rolls to return to equilibrium faster. If moisture is allowed to remain within the roll, severe "blocking" (adhesion of adjacent layers) can be the result.

If a temperature increase is also incurred while the tape is water soaked, steam or at least high humidity will be present. This is more likely to cause damage than water alone. A temperature in excess of l30 deg. F. with a relative humidity above 85 percent may cause layer-to-layer adhesion, as well as some physical distortion.

Nuclear Radiation

As a general statement, it can be said that magnetic tape will be unaffected by nuclear radiation until the dosage approaches a level 200,000 times greater than that which would cause death in 50 percent of the exposed population. Radiation of this level (l00 megarep) would tend to increase the layer-to-layer signal transfer or "print-through" by about 4dB, but would not prevent information retrieval.

Nuclear radiation at the above level would also have some physical effect on the tape coating and backing. The backing will show significant embrittlement, and it is expected that the wear life could be reduced by as much as 60 percent.

It is reasoned that whatever electromagnetic field might result from a nuclear detonation would not be of sufficient intensity to adversely affect the tape; therefore, the threat of signal erasure is virtually non-existent. The effect of neutron bombardment would no doubt be limited to activation of the iron oxide in the coating. This would produce a radioactive isotope that itself might become a source of further radiation, but it is theorized that such activation would not produce a change in the overall magnetic properties of the coating.

Radioactive dust or fallout is not capable of producing the dosage necessary to adversely affect magnetic tape. The recommendations made earlier to protect the tape from normal contamination are applicable here.

Conclusion

As can be seen from the above discussion, when speaking of major catastrophes, heat and fire damage are considered much more serious.

Under proper storage conditions, magnetic tape has the ability to retain intelligence for an infinite period of time; of greatest importance is the physical preservation of the medium so that adequate head-to-tape contact can be maintained.

If any questions about this topic arise, simply write to:

Product Communications
Magnetic Products Division
3M Company
3M Center
St. Paul, MN 55101

{Editor's Note: In writing, please refer to the original "Sound Talk" article of 1970; remembering that it is now 13 years later and these people have never heard of me.}

HINTS AND KINKS

The following was submitted by John Dascenzo, W3IPD, of Fairfield, PA:

"I have always had stability problems when using a soldering gun, in positioning the point on the work and then pulling the trigger. No matter how careful I am, the action of my pulling the trigger frequently causes the point to move. I found a simple solution for this problem.

"First, the trigger of the gun is tied permanently "on" with a piece of string (color of this string not important). The gun is then plugged into a foot switch.

"My foot switch is rather fancy. I placed a microswitch on the side of a heavy metal block--about the size of a brick. It is arranged so that by merely sliding my shoe slightly, the gun goes on. An AC receptacle for the gun is also mounted on this block."

Thank you, John Dascenzo. Your point has good basis in physiology. If we look back at "Soldering, Part II" (Winter 81), an analogy is drawn between the wrist and a pulley system, with the tendons being "ropes" which connect the fingers to their muscles in the forearm. Applying tension or moving one of the fingers puts the wrist in a very unstable condition; uncontrolled movement of the hand is the likely result.

The following ideas were contributed by John Lizza of Carle Place, NY:

"Obviously, Nichrome wire in heating elements cannot be spliced by soldering the broken ends. Repairs can be made by inserting the broken ends into a crimp lug and crimping it tightly.

"I do automotive work, and it is often necessary to note small but critical differences in the voltage on the electrical system (say from 13V to a high of 15V while the battery is charging). However, my multimeter only has scales of 0 to 50 volts and 0 to 5 volts.

"I get around this problem by connecting a zenor diode (for example, a l2.4-volt unit) in series with my meter while setting its range to the 5-volt scale. In this way, the critical changes end up as a 2-volt range at the bottom of my more sensitive scale.

"To check liquid levels--brake fluid, oil, water, etc.--use a straw or a piece of tubing through which you are constantly blowing. You will know immediately when the other end becomes obstructed by the fluid."

Thank you, John Lizza. Also, it seems I missed a major point in describing his work on circuit boards with the hand in a plastic bag. He points out that this technique really shines when checking for intermittents and loose connections. Sorry.

LEXICOGRAPHY

Editor's Note: Here's what us old folks do to get around not having a scientific calculator.

The National Braille Association's Braille Technical Tables Bank provides thermoform copies of the more than 300 tables in its collection at prices substantially below cost. The collection includes standard tables found in mathematics, computer science, statistics, chemistry, physics, finance, etc.

The Pictorial Catalog (print only) of the collection consists of a written description of each braille table plus a reproduction of a sufficient portion of the print table from which it has been adapted and transcribed to permit positive identification of domain, range and number of significant figures.

Price:

Payment must be in U.S. currency or an authorized purchase order must accompany the order.

Mail to: NBA Braille Technical Tables Bank, 3l6l0 Evergreen Road, Birmingham, MI 48009.

Also available from the Book Bank of the NBA are the following titles on radio information. These titles are listed in their General Interest Catalog, which may be obtained for $l.00 from the NBA Braille Book Bank, 422 Clinton Avenue South, Rochester, NY l4620. Specify print or braille.

Available from Recorded Periodicals, 9l9 Walnut Street, 8th Floor, Philadelphia, PA l9l07 (tel. 2l5-627-4230) are the following titles. Specifics on subscription information available on request.

The following items are available on tape from The Volunteers of Vacaville, P.O. Box 670, Vacaville, CA 95696. (Catalog also available on open reel or cassette.)

EDITOR'S CORN

I got a note from Joseph Giovanelli informing me of the proper meaning of the root "lexis" and of my incorrect use in the contrivance "lexicography." "Lexis," as it seems, means "word," while "biblio" means "book." This may well explain why books are not written on polycarbonate (Lexan). It's Greek to me, Joe.

With regard to the above, I offer the following two reasons for retaining the name, "Lexicography." 1} Have you ever noted the Arctic response from a co-worker of whom you innocently ask, "Would you please type this 20-page bibliography?" 2} In my years of experience with products and organizations, I have noted that there is only one thing worse than calling it a bad name: changing it. I hope you all enjoy that column, "Lexicography."

I am horrified and incensed at the fact that several of my suggested questions were left out of the Evaluation Questionnaire on the contrived rationale that it should be objective. These questions are listed below.

H.5. Are the braille cells (please circle the appropriate response)

Too large, too small, or standard?

I.5. In the large-type version, is not the print too close to the page?

J.1 Closed-circuit phone jack with insulating washers (please circle the appropriate mounting hole). {Hmmmm, how'd that get in there?}

J.2 Don't you think that if university professors were blessed with the Editor's quick wit that our grades would have been higher?

Say Yes

J.3 Must the Editor's veiled greatness ever remain tragically o'ershadowed by his obvious sheepishness?

Yes sir, no sir, three bags full.

See you next Spring.

CORRECTION--"SCA Decoder" (SKTF, Fall 1981)

We forgot to power the RCA CA3089; pin ll goes to the VCC line, which in turn goes through a switch to the power supply.

CORRECTION--"Little Go Beep, II" (SKTF, Summer l982)

In the first mike preamp using a TL072 (top of page 33), we find, "pin 3 also goes through 4.7K to ground." No, No! In building this preamp, we discovered that its bias must be VREF1, just like pin 2. Therefore, "pin 3 should go through 4.7K to VREF1."

Sorry, a whole battery supply of engineers missed that one.

CORRECTION--"The Fowle Gimmique" (SKTF, Summer l982)

Jim Swail points out that an offset voltage in some LM358's can introduce an error at the triggering point of the "beep oscillator"-- his unit was triggering low by about 4mV. (Data sheets on the 358 list offsets of 2mV nominal and 7mV max.) There are four possible solutions:

  1. The LM358 can be directly replaced with a high-grade op-amp, National LF442ACN, whose offset is 0.5mV nominal and 2mV max. (Available for $6.l0 from Hamilton Avnet, ll75 Bordaux, Sunnyvale, CA 94086.)
  2. For the same money, you can buy several 358's so as to hand-pick one with a low offset.
  3. An offset adjustment can be built into the instrument by connecting the bottom end of the calibrated pot through a 500-ohm rheostat to the negative end of the battery, not to ground. The calibration and offset adjustments will interact, however.
  4. One could rebuild the device with two single op-amps having provisions for offset pots (See the l00K pot in the original SK Gimmick, Spring l98l.) With this arrangement, the l meg offset pot in the present Fowle Gimmique's oscillator can be eliminated.

Subscription reminder

Dear Subscriber:

Our records show that your subscription to the Smith-Kettlewell Technical File expires as of this issue. We very much want to keep you, for obvious reasons. It is true that we need the financial support, but we also would hate to miss our chance of getting nuggets of knowledge from anyone. Please don't pass us by unduly.

Actually, you are in luck as far as subscription price is concerned. It is true that the yearly rate has gone up to $l5 for braille and large print, and $8 for cassettes. However, in an attempt to simplify our mailing list situation, we are switching everyone to a purchase of "calendar years" of the magazine. This means that your renewal will actually be buying the four issues in 1984 with the remaining issue of this year being sent free in order to shift you into phase with everyone else. Act now, and you will be getting five issues for the price of four.

Please make checks payable to the Smith- Kettlewell Eye Research Foundation at the above address. Thanks for your support.

Bill Gerrey, Editor