Smith-Kettlewell TECHNICAL FILE

A QUARTERLY TECHNICAL JOURNAL FOR BLIND AND VISUALLY IMPAIRED READERS

Sponsored by REHABILITATION ENGINEERING CENTER SMITH-KETTLEWELL INSTITUTE OF VISUAL SCIENCES

2318 Fillmore Street, San Francisco, CA 94115

VOL. 2, NO.2, SPRING 1981

TABLE OF CONTENTS

Auditory "Gimmick" Circuits,Old and New

Basic Analog Meter Reader

Making Braille Dials

Auditory Volume Level Indicators (Part II)

Soldering (Part III)

Wire Wrapping, A Construction Technique for IC Projects

Auditory "Gimmick" Circuits Old and New

ABSTRACT--Two voltage-controlled oscillator (VCO) circuits are detailed here. "Gimmick" is an old Braille Technical Press term denoting the class of instruments which serve as relative indicators. Such circuits have been used primarily by ham radio operators to adjust their transmitting equipment. However, they have other applications, such as testing of electrical continuity, aligning tape recorder heads, serving as light probes, et cetera.

Unlike most other tasks involving electrical measurement (meter reading), transmitter and antenna tuning requires the operator to make a series of interdependent adjustments based on -relative meter indications. Adjustment of two or more controls is done while attempting to "guess" the outcome, and an iterative process of readjustment finally causes the meter to "land" on the desired setting. These adjustments must be made as quickly as possible if damage to the equipment is to be avoided.

"Null-type" and talking digital meters are useful' given static conditions, but they give the user little or no dynamic information; they are not well-suited for adjusting transmitters and/or antenna tuners. With a digital meter, the operator takes too long, causing overheating of the equipment or transmitter shutdown. With a null-type meter, the operator gets very little indication of relative change once the null condition has been upset; he can easily become lost as he manipulates more than one control at a time. Therefore, these systems are the wrong tool for the job.

A good relative indicator such as a voltage-controlled oscillator (VCO) gives the operator constant feedback as to what effect he is causing with his various adjustment procedures. Traditionally, a VCO of high sensitivity is paralleled across the visual meter of interest during the tuning-up process. The voltage drop across the meter (from 50mV to 500mV for a full-scale meter deflection) is used to control the frequency of the audible VCO.

A quantitative meter reading can be taken at any time by disconnecting the VCO from the meter and connecting it to a braille calibrated voltage standard which can be adjusted through the same full-scale voltage range as will be developed across the visual meter. When the calibrated standard is adjusted to bring the VCO to the same pitch as is heard when it is connected across the meter, the braille dial on the standard can be read as if it were the pointer of the visual meter.

Two VCO designs will be described here which afford simple and direct connection to transmitter meters. The first of these instruments was designed by Bob Gunderson of the Braille Technical Press, and was published under the name of "Transistorized Auditory Gimmick" in 1955. The second design, the "Smith-Kettlewell Auditory Gimmick", came along 23 years later when we adapted a VCO circuit commonly used by electrophysiologists to register the minute electrical potentials of neural activity. Both VCO circuits are sensitive enough to give audible indications when connected to low level voltages such as those developed across standard meter movements.

CAUTION! if the meters to be monitored are not at ground potential, the tuning aid circuitry must be appropriately insulated from ground in order not to short circuit the equipment. Furthermore, if these meters are connected in high-voltage circuits, the tuning aid circuitry must be well insulated in accordance with good high-voltage construction practice, and an outer shielding metal cabinet which is grounded should enclose the tuning aid.

Transistorized Auditory Gimmick

I have always felt that the brilliance of an engineer can be measured by the simplicity of the eventual design. Should I be forced to choose one example to illustrate my point, I would have to pick this circuit. It contains only two transistors, two capacitors, one resistor and a garden-variety push-pull audio output transformer. (Although they are expensive, loudspeakers with a centertapped voice coil can be obtained to make a transformerless version of the instrument).

Rivaling the genius of its inventor is the variety of innovative uses to which this circuit has been put by Bob Gunderson and readers of the BTP. With a diode detector and an antenna connected to it, the Gimmick becomes an RF Field Strength meter. By connecting this diode detector to a parallel-tuned circuit made up of a variable capacitor and one of a set of plug-in coils, it becomes a tuned absorption wave meter. With this diode detector working into a larger filter capacitor, the Gimmick becomes an ideal instrument for aligning tape recorder heads. If a phototransistor is put in place of the input transistor, the Gimmick can be converted into a general purpose light probe. If the input transistor is omitted and test probes are put in its place, you have one of the finest continuity testers imaginable. If a telegraph key is put in series with the loudspeaker, the instrument becomes a code practice oscillator. This circuit may already look familiar to you; it is used as the VCO in Swail's volume level indicator (Winter 1981), and as the auditory indicator in the Gunderson diode checker (Fall 1980).

Its disadvantages as a meter reader are not serious--they are best categorized as leaving "room for improvement." First, the input impedence is just low enough to cause a slight loading effect on the visual meter. Second, the oscillator pitch changes are most noticable for input voltages above 15 or 20mV, making use of the device marginally successful for VSWR measurement.

In operation, a "blocking" oscillator using a PNP transistor is built around the primary winding of the output transformer. The base bias of the oscillator transistor is controlled by an NPN germanium transistor operating as a DC amplifier to sense the minute voltage appearing across the visual meter.

The sensitivity (the pitch range attainable) and the audio output level are interdependent, and can be varied by changing the capacitance values in the oscillator circuit. By experimenting with these values, a "trade off" can be made to suit your liking.

Original Gimmick Circuit

The transformer used has a 500 ohm center-tapped primary winding, and an 8 ohm secondary winding which is connected to an 8 ohm loudspeaker. The center tap of the primary goes to the emitter of Q1 (PNP transistor). The bottom end of the primary winding goes through a capacitor (.01 uF) to the base of Q1, with the collector of Q1 being grounded. The top of the primary goes to the +9V line, which goes through a switch to the positive side of the battery. This primary winding is shunted by another capacitor (.01 uF). The emitter of Q2 is grounded; its collector goes through 47K to the base of Q1. (Q2 is an NPN germanium transistor). The base of Q2 is the hot input to the Gimmick, while the cold side of the input is grounded.

To isolate the circuit from RF, the base of Q2 is bypassed to ground through .01uF and this base also goes through an RF choke to the positive meter terminal. The meter side of the choke is also bypassed to ground through .01uF. The negative side of the meter is grounded. Locate the choke and its bypass capacitors near the Gimmick, at least put them inside the cabinet. The cabinet must be metal and must be grounded for RF shielding.

Two popular schemes for "calibrating" the Gimmick are in common use. The least sophisticated of these is based on the premise that there are usually two or three specific meter readings of interest to the operator; for example, the transmitter manual may call for a specific plate current for SSB, and another for CW. If this is the case, presettable comparison tones can be wired to a row of push- buttons, one button for each meter reading of interest. The second system uses a braille-calibrated potentiometer as a voltage standard which can be adjusted to match the voltage across the meter. Although the circuitry for this latter setup is simple, the builder does have to come up with a suitable pointer knob and a braille scale.

Calibration Circuits

The connection between the RF choke and the base of Q2 is broken so that the push-button switch circuitry can be inserted. The .01Uf capacitor on the base should remain at the Q2 base.

The first scheme uses a series of single pole double-throw push- button switches, one for each meter setting. The swinger of switch SW1 goes to the base of Q2. The swinger of SW2 goes to the normally- closed contact of SW1. The swinger of SW3 goes to the normally- closed contact of SW2. The swinger of SW4 goes to the normally- closed contact of SW3... The normally-closed contact of the last switch goes to the RF choke--the end which was originally connected to the base of Q2. Each normally-open contact goes to the arm of its own 10K POT. The bottom of each POT is grounded while the top of each POT goes through 560K to the +9V line.

For the second scheme, a single-pole double-throw switch has its swinger going to the base of Q2, its normally-closed contact going to the choke, and the normally open contact going to the arm of a 1K precision POT. The bottom end of the POT is grounded, while the top end of the POT goes through a 150K rheostat in series with 22K to the +9V line.

Field Strength Meter Circuit

A germanium diode (1N270, lN34A, Radio Shack 276-1123) has its cathode connected to the hot input terminal of the Gimmick, and its anode connected to a length of antenna. The cold side of the input is grounded.

Tuned Absorption Wave Meter Circuit

Instead of going directly to an antenna, the anode of the diode goes to the top of a parallel-tuned circuit whose bottom end is grounded. This tuned circuit consists of a 50pF variable capacitor (rotor at ground) and a plug-in coil wound for each desired band. An antenna can be coupled to the device by winding a 1- or 2- turn link around the coil form near the cold end of the tuning coil.

Tape Head Alignment Circuit

The hot input terminal of the Gimmick goes to the cathode of a germanium diode, and the anode of the diode goes to a low-impedence output of the playback system (a loudspeaker or ear phone output is suitable). The input terminals of the Gimmick are shunted by an electrolytic capacitor (from 1uF to 3uF). The cold input terminal of the Gimmick is connected to the cold side of the tape system output. I built a patch cord for this purpose, the plug of which contained the diode and electrolytic capacitor.

Light Probe Circuit

A phototransistor (such as the Fairchild FET 100) is substituted for the NPN germanium input transistor, that is, the emitter of the photo transistor is grounded, and its collector goes through the 47K resistor to the base of the oscillator transistor. The base of the phototransistor is unused and is left open.

Continuity Tester Circuit

The NPN transistor is removed from the circuit. The negative test probe is grounded, while the positive test probe goes through the base resistor to the oscillator transistor.

Some juggling of the circuit values may improve the performance of this circuit. In the SFB Products "Audicator", for example, the positive test probe goes through 4.7K to the base of the oscillator, and the capacitor from the base to the bottom of the transformer primary is .1uF. The capacitor across the primary is omitted.

The Smith-Kettlewell Auditory Gimmick

This circuit has three advantages over the original Gimmick: They are as follows:

  1. The input resistance of the Smith Kettlewell Gimmick is extremely high, virtually an open circuit. Thus, the instrument has no effect on the reading of the visual meter.
  2. The VCO operates at very low input voltages (within a couple of milivolts above ground). This makes the instrument ideal for reading the meters of SWR bridges. Furthermore, the inclusion of an offset adjustment POT in the circuit enables the user to preset the lowerfrequency limit of the VCO.
  3. The sensitivity of this instrument is adjustable over an extremely wide range, making the instrument usable for monitoring a wide variety of meters having internal resistance networks. The original Gimmick, on the other hand, is usable only for low-impedence meters ("current" meter). This flexibility is of special interest to owners of R.L. Drake radio equipment, since Drake transmitters use a voltmeter in their metering circuits.

In operation, an RCA CA3130 op-amp together with Q3, works as a voltage-controlled active current source. When a voltage is applied to the non-inverting input, pin 3, the op-amp strives to establish an equal voltage at its inverting input, pin 2. To do this, the output (pin 6) of the op-amp adjusts the base bias of Q3 so that the voltage drop across the Q3 emitter resistor equals the input voltage on the non- inverting input. For any given input voltage, a corresponding emitter current will be developed through the Q3 emitter resistor to assure that the inverting input's voltage follows that of the non-inverting input.

Changes in the collector load will cause the op-amp to re- establish the IR emitter voltage at pin 2, keeping the emitter current constant. Since the collector current approximately equals the emitter current, the collector of Q3 can be seen as a constant current source which can be controlled by an input voltage. Varying the Q3 emitter resistance changes the gain of the current source, thereby providing a means by which the sensitivity of the instrument can be adjusted.

The active current source drives a "current mirror" made up of transistors Q1 and Q2 which, in turn, controls the charging cycle of an NE 555 timer IC. (Q1 and Q2 are PNP silicon transistors.) In operation, current drawn from the Q1 collector-base connection causes the bias conditions of Q2 to be established for sourcing an equal current from the Q2 collector. Q2, through the current-limiting resistor of 22K, supplies the charging current for the timer's charging capacitor.

Finally, the 100K offset adjustment POT (between pins 1 and 5 of the op-amp) has been included to provide an adjustable lower limit to the pitch of the VCO.

Smith-Kettlewell Gimmick Circuit

The cold input terminal goes through an RF choke (2.5mH) to the circuit's common ground and to -9V. The hot input terminal goes through an identical RF choke, then through 220K to the non-inverting input of the op- amp (RCA CA3130) with this non-inverting input also going through .1uF to ground. The input terminals are shunted by .05uF, and the other ends of the chokes are also shunted by .05uF (this input filter is necessary because of the circuit's high propensity to RF interference.)

Pin 4 of the op-amp is grounded (to the Gimmick side of the RF choke). Pin 7 goes to the +9V line. Pin 1 goes through 30pF to pin 8. This pin 1' also goes through a 100K POT to pin 5 (offset adjustment), with the arm of the POT being grounded.

The output of the op-amp goes to the base of the current source transistor, Q3 (2N2222), with the emitter of this transistor going to the inverting input, and also through an adjustable emitter resistance to ground. This emitter resistance is comprised of 1K in series with a 25K rheostat.

The collector of Q3 goes to both the collector and the base of transistor Q1, the first half of the current mirror. The emitter of Q1 goes to the +9V line. (Q1 and Q2 are PNP silicon transistors' type 2N2907.) The junction of the Q3 collector, the Q1 collector, and the Q1 base goes directly to the base of Q2,with the emitter of Q2 going to the +9V line.

The collector of Q2, the second half of the current mirror, goes through 22K to pins 2 and 6 of the NE 555 timer, with pins 2 and 6 also going through .0047uF to ground. Pins 2 and 6 go through 10K to pin 7. Pin 1 of the timer is ground, while pins 4 and 8 are tied together and go through 10 ohms to the +9V line. The +9V line is bypassed to ground through 100uF (negative side of the capacitor at ground). Pin 3, the output, goes through 100 ohms to one side of the speaker. The other side of the speaker goes to pins 4 and 8.

Circuit Adjustments

The offset and sensitivity adjustments are interdependent.

I recommend that the op-amp first be "zeroed" (the output brought to zero with the input shorted). Short out the input and turn the offset POT to a point at which the Gimmick oscillates noticeably; then slowly bring this adjustment to the point at which the frequency of oscillation goes to zero. Open the input and connect it to the meter to be monitored. With the equipment set to produce a fairly high meter indication, adjust the POT on the emitter of Q3 (25K) to obtain a comfortably high-pitched tone from the Gimmick. Finally, readjust the offset POT to zero the op- amp, or to produce a comfortably low-pitched tone as desired.

Calibration Circuits --

a similar system can be used to provide standard voltages with which the meter indications can be compared. The circuit using the series of push-buttons to provide a series of comparison tones can be stolen directly from that used with the original Gimmick. The swinger of SW1 goes through the 220K resistor to the non- inverting input of the op-amp , and the .05uF capacitor stays with this resistor. The normally-closed contact of the last switch goes to the instrument end of the RF choke in the hot input lead. For the system using a braille-calibrated precision POT, I recommend the following circuit. The swinger of an SPST pushbutton switch goes through the 220K resistor to the non-inverting input, and also through the .05uF capacitor to ground. The normally-closed contact goes to the Gimmick side of the RF choke in the hot input lead. The normally-opened contact goes to the arm of a 10K precision linear POT, with the bottom of this POT being grounded. The top of the POT goes through 100K in series with a 1.5megohm rheostat to the cathode of a 6.8V zener diode (Radio Shack 276-561 6.2V unit is suitable). The anode of the zener is grounded. The cathode also goes through 470 ohms to the +9V line.

Circuit Modifications for Use With Drake Equipment

Many Drake transmitters employ a volt meter instead of the low-impedence current meters traditionally found in equipment of other brands. The Smith-Kettlewell circuit can easily be adapted for use with these meters. First,the Q3 emitter resistor should be increased; the 25K POT should be changed to 150K. Second, the calibration circuit must be changed to provide higher comparison voltages. To do this, the series resistances associated with the POT should be decreased by a factor of 10. For example, if the preset push-button system is chosen, the 560K resistors should be changed to 56K. If the braille-calibrated system is used, the 1.5Meg rheostat should be changed to 150K, and the 100K series resistor should be changed to 10K.

This Smith-Kettlewell circuit can be outfitted with the other various attachments described for use with the original Gimmick. The Field Strength Meter, the Wave Meter, and the Tape Head Alignment circuits can be connected to the input as before. However, because of the extremely high input resistance of this VCO, a 10K ohm bleeder resistor must be connected across its input to serve as a load into which these attachments can be properly terminated. (Of course, this resistor can be incorporated into each attachment, and should not be present on the input when the VCO is connected across a meter.) On the other hand, the circuit connections for the Light Probe, the Continuity Tester, and the Code Practice Oscillator are not immediately obvious and are described below:

If the current source is disconnected, we have a perfectly good NE 555 oscillator to play around with. To accomplish this, a double-pole double-throw two-position switch (without a center off position) can be installed to disconnect Q3 and to short out the input of the op-amp . The swinger of one pole is grounded, while position 2, the "modify" position goes to the non-inverting input of the op-amp. Position 1 of this pole is left open. The swinger of the second pole goes to the Q1 base and collector, along with the base of Q2. Position 1, "normal", goes to the Q3 collector. Position 2, "modify", goes through 100K to one side of a "continuity" jack. The other side of this jack goes to the cold input terminal, which is essentially ground (this permits grounding of the cold side of the continuity jack to the metal cabinet).

Light Probe

The emitter of the phototransistor (Fairchild FPT 100) goes to the cold side of the jack, while its collector goes through the 100K resistor to the switch. You may wish to reduce the sensitivity of this Light Probe by connecting a resistor of 1 megohm from the base to the emitter of the phototransistor.

Continuity Tester

The negative test probe goes to the cold side of the jack, and the positive test probe goes through the 100K resistor to the switch.

Code Practice Oscillator

The telegraph key can be connected across the terminals of our continuity jack, but an additional resistor of about 220K ohms must be put in series with the key to lower the pitch of the oscillator (unless your dog knows the code). One side of the key goes to the cold side of the jack. The other side of the key goes through an additional 220K, then to the hot side of the jack already going through 100K to the switch. An alternative arrangement is to connect the key directly between the hot side of the continuity jack and hot side of the input jack, thereby using the 220K resistor from the RF choke to the non-inverting input (which is now grounded through the switch). This eliminates the need for an additional resistor. A tantalizing prospect for a future article is pairing up a VCO with a digital meter reader, so that the push of button would give you a direct meter reading in speech or Morse code.

PARTS LIST (Original Gimmick)

PARTS LIST (Smith-Kettlewell Gimmick)

Capacitors:

Resistors:

Semiconductors:

Chokes:

Basic Analog Meter Reader

ABSTRACT--This circuit can be used to modify inexpensive test instruments containing analog'meters. At SmithKettlewell Institute we have used it to convert many test instruments-- industrial meters and automotive test equipment. We have used it to make instruments from scratch, such as pressure and vacuum gauges, and electronic micrometers.

Talking digital meters are marvelous; they are the only hands- free type of metering equipment available to the blind technician. However, as we forge ahead and order the speech chips or the, completed instruments, two questions should not go unaddressed. Are they useful in all testing situations? How costly are they?

It can be argued that in many cases digital meters are not the right tool for the job. Consider the following examples: Whereas a sighted person can have immediate access to information by glancing at a digital display window, the blind user does not have access to these readings "at a glance", he must wait for the display to be spoken. When adjustments are being made on a circuit, this becomes a tedious waste of time.

As another example, anyone who builds electronic projects discovers that tracking down intermittent solder connections constitutes the majority of the troubleshooting process. Whereas the sighted technician can visually "comb the bushes" to find troublesome connections very rapidly, the blind technician cannot identify them as quickly by touch--he must offset his disadvantage by "zeroing in" on these problems with measuring instruments. Talking digital meters do not begin to address this issue, since the time at which the information is presented is not the time at which the intermittency occurs. Attempting to use them for this application can cause your verbal utterances to become unseemly, and may result in premature depletion of the cooking sherry.

As it stands now, talking digital meters are expensive. Even if you were to buy the parts and build one yourself, a talking meter could not be built for less than about $300. This can buy you 25 years of the Technical File at today's prices.

I use the previous discussion to justify the following article, Reviewing the traditional auditory meter scheme, which can be implemented at very low cost. This system can be used to convert inexpensive visual multimeters, etc., and its audible tone output presents immediate feedback to the user as changes in the circuit conditions occur.

Conventional visual meters operate on the principle that the magnetomotor force associated with an electric current in a conductor is proportional to this current. Pitted against the spring, whose force is proportional to its displacement, the "deflection" of the spring can be used to measure the electric current. A visual meter movement is like a spring scale set up to measure the magnetic force around a coil of wire through which an electric current is passed. (Note that these devices measure current, not voltage.) The more sensitive and delicate these instruments are, the less influence they will have on the circuits being metered. For this reason, they are almost always much too delicate to be read by touch.

The system of measurement used by our meter reader involves comparison of the unknown quantity with a standard. Analogous to the balance scale, an unknown voltage is matched with a selectable standard voltage. (Note that this system involves the comparison of voltages, not currents.) The advantage of this approach is that the fragility or ruggedness of the standard is not an issue--as long as this standard can be set to match the unknown voltage, it satisfies the electrical criteria. This gives the instrument designer free reign over the standard's construction.

A simple but effective voltage standard can be made using a precision linear potentiometer which is fitted with a pointer knob and a braille dial. This potentiometer is connected through a screwdriver-adjust calibration rheostat to a stable voltage source (such as a zener diode). The voltage across the precision potentiometer can beset to cover exactly the range of unknowns anticipated.

We now need a way of determining when the calibrated voltage and the unknown voltage are the same, i.e. we need to know when the balance scale is balanced. If the difference between these two voltages is "chopped" (modulated) at an audible rate and then fed through an amplifier to a loudspeaker, the user can hear this difference voltage--the humming of the chopper is emitted by the loudspeaker. If there is no difference between the two voltages, there will be no difference signal being chopped, and nothing is heard from the loudspeaker. As soon as a difference between the two voltages occurs, this condition will be indicated by a tone.

In operation, the user adjusts the calibrated standard for a "null" (the point at which the tone fades away), indicating that there is no difference between the unknown and the standard. The rugged pointer knob is then read against the braille graduations, just as if it were the pointer on the meter.

In the BTP days, these instruments were said to use the "chopper-comparator" system. In the rehabilitation community, they are generally referred to as "null-type" instruments.

These chopper-comparator systems have some advantages over visual meters. They are usually more accurate, and they can be arranged to have very little influence over the circuits being measured.

Visual meter movements are plagued with mechanical problems. Variations in parameters of the spring effect calibration. Bearing friction is a major source of error which becomes more significant at small meter deflections. Manufacturers rate the tolerance of error as a percentage of úthe full-scale reading (typically two to five per cent of full scale). This means, for a meter which is good to a full-scale accuracy of 2%, that readings taken at half scale can have an error of 4%, and readings taken at one-quarter scale are only accurate to within 8%. The accuracy of our comparison system, on the other hand, depends primarily on the precision of the standard (potentiometers of 1% linearity are inexpensive and readily available), and measurement error is uniform over the whole range.

When comparison systems are used in their purest form as a "volt balance" ("The Cranmer Volt Balance", BTP December, 1960) no current will flow when the two voltages are equal, and no power will be extracted from the circuit under test. In a sense, a balancing system can be seen as an "impartial observer" of the unknown voltage.

The circuit to follow is designed to read any standard visual meter by sampling the minute voltage which appears across it. The chopper-comparator circuit acts as an impartial observer looking at the IR drop across the meter, and has no effect on the meter's reading when nulled or when turned off. The braille voltage standard can be set to cover full-scale voltages ranging from less than 50mV to greater than 500mV, making this circuit compatible with meters in common use.

If desired, this meter reader can be fitted into the space occupied by the visual meter in the converted test instrument. Often, the visual meter is mounted to the front panel by four machine screws and can easily be removed. There are cases, however, where the cabinet of the instrument serves as the housing for the delicate meter movement, and you must use your imagination in taking out the meter movement and utilizing this space. Before you ravish an instrument beyond all repair, make sure that there will be room for the precision potentiometer behind the face of the unit. Remember also, the instrument must be roomy enough to house the electronics of the meter reader and a loudspeaker (the larger the speaker the better).

If the meter in the instrument has its negative terminal- grounded, you may be in luck; it might be possible to power the meter reader from a source which is already part of the instrument. More often than not you will not be so fortunate, and an additional battery or floating AC supply will have to be included to power the meter reader. Also, the meter will need an on-off switch. Sometimes it is possible to modify or replace the instrument's power switch for the inclusion of an extra pole, so that the meter reader can be controlled through this main switch. If the instrument is AC operated, a separate 8 or 12 volt AC supply can be run in parallel with the instrument's power supply, thus eliminating the need for this extra pole or a separate on-off switch.

Finally, if the visual meter is removed to accomodate the meter reader, a resistor of equal value must be put in place of the meter to preserve the integrity of the instrument. Details on how to determine the meter's resistance are discussed later.

You may wish to build the meter reader into a separate cabinet which can be hooked up to the converted equipment by a plug-in connecting cable. This eliminates the need to perform major surgery on the instruments to be modified. Furthermore, this separate meter reader can be made multi-channel by incorporating a selectable set of input jacks into the design. In this way, several different test instruments can be converted using a common meter reader to switch between them.

I have built many of these meter readers into separate cabinets measuring 3 x 4 x 5 inches. The 4 x 5 inch dimension can accommodate a 4 inch square piece of plastic containing the braille markings. This still leaves a 1 inch margin beyond one edge of the plastic where switches can be mounted. This size cabinet allows plenty of room for the circuit to be built using point-to-point wiring or solderless breadboard techniques.

Circuit Operation

The chopper used in this circuit is an opto- isolator (G.E. H11F3), consisting of an LED shining at a photo FET. Both the LED and, the photo FET are incapsulated into a 6-pin dual in-line package.

The channel of the photo FET can be seen as a normally-open switch which is controlled by the LED. When current is applied to the LED, the channel of the photo FET becomes conductive. An NE 555 oscillator supplies a 500Hz signal to the LED, causing the channel of the FET to open and close at an audible rate.

The channel of the FET is set up to "chop" the current through a 47K sampling resistor across which the "difference" voltage is developed, that is, the output of the voltage standard goes through 47K, then through the chopper to the top of the meter. An audio amplifier (using a National LM 386) "listens" to the difference voltage across the sampling resistor and presents it to the user through a loudspeaker.

Smith-Kettlewell Meter Reader circuit

The negative side of the meter reader supply is grounded (this ground may not be the same as the instrument's ground). The positive side of this supply (from 8 to 12V) goes to the +V line of the meter reader through an on-off switch.

The bottom of the braille-calibrated precision potentiometer (10K) is grounded. The top of this POT goes through 100K, then through a 1.5Meg calibration rheostat to the cathode of a 6.8V zener diode (1N 5325). The anode, of this diode is grounded. The cathode also goes through 470 ohms to the +V line.

The arm of the 10K calibrated POT goes through 47K, then through the channel of the FET to the positive side of the meter. (pins 4 and 6 on the H11F3 are the FET's channel.) The negative side of the meter goes to ground (meter reader ground). The meter is shunted by 10uF, - with the negative side of this capacitor being grounded. (The negative side of the meter is usually nearest 0, that is, the left-hand terminal as you face the front.)

The junction of the 47K resistor and the FET goes through .1uF to pin 3 of the LM 386. Pins 2 and 4 are grounded. Pins 2 and 3 are shunted by .01uF. Pin 1 goes to the positive end of a 10uF capacitor, with the negative end going through a 1K rheostat (volume adjust) to pin 8. Pin 7 goes through 25uF to ground (negative at ground). Pin 6 goes through a 10 ohm decoupling resistor to the +V line, and pin 6 is bypassed to ground through 250uF (negative at ground). Pin 5 goes through .22uF to pin 4. Pin 5 also goes through 100uF (positive toward pin 5) to one side of the speaker, while the other side of the speaker is grounded.

Pin 1 of the NE 555 is grounded. Pins 4 and 8 are tied together and go to the +V line. Pins 2 and 6 are tied together and go through .01uF to ground. Pins 2 and 6 also go through 100K to pin 7. Pin 7 also goes through 100K to the +V line. Pin 3, the output, goes through 2.2K to the cathode of the LED (pin 2 of the H11F3). The anode (pin 1) goes to the +V line.

If you ever have trouble procuring the G.E. H11F3 opto-isolator, an alternative circuit can be built using an ordinary Pchannel JFET {such as the Siliconix J176-l8, the Fairchild 2N 4360, or the Radio Shack 276-2037). The null attainable with the simple JFET is not as complete; the optoisolator provides extremely good isolation between the oscillator and the chopper. The H11F3 is worth getting. The following circuit, however: is a pretty fair alter- native.

Circuit Using a Simple JFET

The only change is in the oscillator circuit. Instead of going directly to ground, pin 1 of the NE 555 goes through 220 ohms to ground (which prevents forward biasing of the gate junction). Pins 4 and 8 are tied together and go to the +V line. Pin 1 goes through .1uF to pin 8. Pins 2 and 6 are tied together and go through .01uF to pin 1. As before, pins 2 and 6 go through 100K to pin 7 ,and pin 7 also goes through 100K to the +V line. Pin 3, the output, goes through 100K to the gate of the P-channel JFET, and this gate also goes through .0022uF to ground. The source and the drain (the ends of the JFET channel) are interchangable and are connected as before.

Circuit For Including Multiple Inputs

Since the meters of your various instruments may not be common to ground or to each other, it is advisable to switch both sides of the meter reader's input. Use 2- conductor shielded connecting cables,permitting both sides of the input to "float." In addition, a separate calibration rheostat must be switched into the circuit for each meter.

3-conductor plugs and jacks are used to connect the meter reader to the various test instruments. The cabinets of all the instruments are connected together through the cable shielding. The "ground" or negative meter reader input goes to the swinger of one pole of the selector switch. The "hot" side (FET side) of the meter reader input goes to the swinger of another pole of the selector switch. Position 1 of the first pole goes through the connecting cable to the negative side of meter 1, while position 1 of the second pole goes through the other wire in the cable to the positive side of this meter. In the same way, position 2 of the first pole goes through l wire of the cable to the negative side of meter 2, and position 2 of the second pole goes throuqh the other wire to the positive side of this meter.

A third pole on the selector switch is used to connect a different 1.5Meg rheostat for each. meter. The swinger goes through the 100K resistor to the top of the 10K precision POT. Position 1 of this pole goes through rheostat 1 to the cathode of th.e zener diode. Position 2 of this pole goes through rheostat 2 to this same cathode, etc.

Testing the Meter Reader

Of course, it goes without saying that you can test each block as you build it; you can listen to the oscillator with a test amplifier, you can test the amplifier by touching the input with your finger, etc. However, since I have built meter reader boards a half dozen at a time (to be used in Smith-Kettlewell projects), I have refined the testing of completed boards as follows:

First, short the input, otherwise there will be nothing for the chopper to chop, and no sound will be emitted. Turn the unit on with the calibrated POT turned clockwise up from zero. If you hear a tone, turn the calibrated POT counterclockwise to see if it nulls at zero. If the tone increases in volume you probably have wired the ends of this POT backwards; try turning it to the other end. If no null can be obtained, the POT or something else is not properly referenced to ground. Measure voltages around the POT. They should be small--less than 1 volt.

If nothing is heard from the speaker, touch the amplifier input (pin 3 of the LM 386) with your finger. You should get a loud crackling, shrieking, or growl. If so, back up to the FET end of the .1Uf capacitor and try again. If you get nothing, the FET may be always on. Pull it out of the socket and try this end of the input capacitor again. If you get noise from the amplifier, this may indicate that the oscillator is "hung up" so as to cause the FET to conduct.

So long as the amplifier works, the oscillator can easily be tested by laying one finger on pin 3 of the LM 386, and another on pin 3 of NE555. Somewhere under the shrieking and humming, you should hear the singing of the oscillator.

Sorry, I can take you no further on this bus; from this point you're on your own. After all, if you are like the editor,you can be very creative in damaging parts or making wiring errors.

Calibration

If you happen to know the voltage drop which will occur when the visual meter is deflected to full scale, you can properly adjust the meter reader by using a sensitive voltmeter to measure across the calibrated POT and adjusting the rheostat to match this full scale voltage. This can be done without the meter reader being connected to anything. (Some laboratory voltmeters are not sensitive enough to measure these minute voltages, making this scheme impractical.)

If this voltage is not known or if your test equipment is not capable of measuring it, get the visual meter to deflect to some convenient value (say half or full scale). Set the calibrated POT to this same point on the braille dial. Adjust the calibration rheostat for a null.

Determining the Resistance of the Visual Meter Movement

If the meter is left in the converted instrument, the steps in the preceeding paragraph can be carried out without knowing the parameters of the meter. However, if the conversion is such that the meter is removed from the instrument, the movement's internal resistance must be ascertained so that a resistor of equal value can be installed in its place. CAUTION Never measure a meter's resistance directly with an ohmmeter; the comparatively high current sourced by the ohmmeter will burn out the meter movement long before you get a reading.

Test Set Up

Start by assuming that you have a sensitive current meter. Connect the negative side of the meter to the negative terminal of a 9V battery. Connect the positive end of the battery to the clockwise end of a rheostat of perhaps 500K ohms (linear taper). Be sure that the arm of the rheostat is at the other end or counter clockwise end, and connect the arm through a current-limiting resistor of 100K to the positive side of the meter. Be sure you know what you are doing. You can blow the meter to Fiddler's Green by making a careless mistake.

Carefully decrease the resistance of the rheostat (advancing it clockwise) with the intent of getting a full-scale reading. If a full-scale reading cannot be obtained, the test set up resistances must be changed, try reducing them in value by a factor of five (a 100K rheostat and a 22K limiting resistor). Be careful and make sure that the meter is being closely observed. If a full-scale deflection still cannot be obtained, reduce the values again by a factor of five (a 20K rheostat and, a 4.7K limiting resistor).

Once the meter has been made to indicate full-scale, connect another rheostat across the terminals of the meter--choose one with a value of perhaps 1% of the total resistance in the series network. For example, if the 500K rheostat and 100K limiting resistor did the trick, the parallel rheostat should be about SK for starters. Reduce the resistance of this latter rheostat with the intent of bringing the meter reading down to half-scale. If the adjustment is too sensitive to permit setting this reading accurately, pick another rheostat whose value is lower (by a factor of five) and try again.

Paralleling a rheostat across the meter splits the current into two branches, one through the meter and the other through the rheostat. When this rheostat is adjusted so as to cut the meter reading in half, the currents in the two branches are equal; the rheostat and the meter are now equal in resistance.

Disconnect the parallel rheostat from across the meter and measure it with an ohmmeter; it has the resistance parameter we have been looking for.

While the series network is still connected to the meter (causing it to read full-scale), the full-scale current of the meter can be read with a micro ammeter.

Meters containing internal resistance networks come in a wide variety. Highcurrent meters with very low-resistance internal shunts are common, expecially in radio transmitters. I have also run across DC and AC voltmeters in automotive test instruments which were converted here at Smith-Kettlewell Institute.

If you are getting nowhere with the series network, try measuring the voltage across the meter terminals. If the meter is a voltmeter, you will see some voltage across it. If you read nothing, it is a high-current meter, and you must dredge up a suitable power supply and a low-value series rheostat to cause its deflection.

Parts List - Basic Meter Reader

(Note--If a P-channel JFET is substituted for the optoisolator as discussed in the text, this list should be amended to include additional 1/4W resistors of 220 ohm and 100K, along with a.0022uF capacitor:

Capacitors:

Resistors: (5%)

Potentiometers:

Semiconductors:

*1 Precision POTs

(Note--the model number refers to the case style and other physical characteristics. Normally, resistance of units must be specified when ordering; 10K units are in stock at Newark, and this stock number should be sufficient for ordering from them.)

*2 Supplier's

(Note--our thanks to Mr. Rich Lago, Product Manager at Hamilton Avnet, for stocking HllF3's especially for this project, and for waving the minimum order policy for readers of the Smith- Kettlewell Technical File. Most suppliers require individuals to drum up a large purchase before they will ring up the sale, a policy which throttles the small-scale experimentor. Perhaps by delicately courting this relationship with Hamelton Avnet, we can overcome this stumbling block).

Making Braille Dials

In a future issue, I will reprint a distillation of BTP articles which was collected and recorded by science for the blind in about 1963. Many truly ingenius items were detailed there. However, since I edit this fine magazine my idea first.

Using drafting tools is not very convenient when your available potentiometer has a rotation angle of something like 315 degrees, 352 degrees, etc. Furthermore, you cannot generate a non-linear dial with drafting tools.

Why not mount a plastic blank onto a precision potentiometer and let a meter reader indicate the placement of the braille markings. For generating linear dials, you can connect the meter reader to a calibrated voltage divider made up of a string of resistors or a 10-turn potentiometer mounted on a protractor. For non-linear dials, you can electrically cause the visual meter whose scale you are copying to deflect to the desired readings. (If you wish, you can temporarily disconnect the visual meter from its instrument and operate it from a circuit such as that used to measure its resistance, detailed in the previous article.)

Braille Dial Considerations

Anything on on which dots can be made is suitable for making a braille dial. Plastic, such as Lexan (polycarbonate), mylar, celluloid, Thermoform paper, is all suitable.

Do not attach the dial to the knob so that it turns with the potentiometer; the dial itself should be stationary to promote good spatial orientation when reading it. After completion, it can be secured to the panel of the instrument with double-sided Scotch tape.

My feeling is, don't clutter up the dial with braille figures. Mark it like a clock, three dots for primary divisions, two dots for other major divisions, and single dots for minor divisions. Usually, I make linear scales with ten major divisions having two dots (sometimes putting triple-dot markings at zero, half-scale, and full- scale), and I prefer to have only one single dot between the double-dot markings. An occasional braille "K" may come in handy for a unique marking.

All markings should be at the same radial distance from the center, i.e. put single dots in the nearest position of the cell.

A pointer can be made using a piece of sheet metal or plastic about 50/l000ths of an inch thick. This pointer can then be cemented to the bottom of a suitable knob. You can even attach a pointer to a knob by threading a long machine screw into one of the set screw holes. Cut off the head of the screw and file it to a point (use a knob with a metal bushing for this trick).

Mounting the Blank for Brailling

A fixture for mounting the blank to the precision POT can be made using a panel bearing (H.H. Smith No. 119). This bearing has a 3/8 inch threaded outside portion with a matching nut which can be used to secure the blank to it. To fix this panel bearing to the potentiometer shaft, a knob or the metal insert from a knob can be cemented to the hexagonal flange of the panel bearing.

Find either a knob with a flat top surface or a metal insert which can be broken free. (If the whole knob is used, drill the hole for the shaft all the way through.)

Thoroughly clean the knob and the bearing with alcohol in preparation for cementing them together. Slip them both onto a piece of 1/4 inch shafting which has been lightly greased with Vaseline to prevent them from sticking to it. Apply cement to them and mash them together, being careful not to spread Vaseline around which will ruin the bond. Because of the Vaseline, you should be able to tap the shaft out of this new fixture after the glue dries. Remove any excess glue with a 1/4 inch drill bit.

Punch a hole in the center of your blank and enlarge it to 3/8 inch with a tapered reamer. Secure it to the panel bearing as intended. (Sandwiching the plastic between washers is good practice.)

Brailling Fixture

A fixture must now be made to put the blank in relation to a braille slate so that it can pass through the slate. Perhaps the easiest and most expensive way of doing this is by modifying a board slate.

The precision POT is mounted below the bottom edge of the clipboard using a small extension plate of sheet metal. This small plate should be screwed to the underside of the board to leave room under the blank for the potentiometer's mounting hardware.

In order to accommodate large size blanks, you can put this mounting bracket below the bottom right corner of the board; however, you must also remove the sharp little locating pin that punctures the paper at the open end of the slate.

Try to locate the POT directly below a column of cells on the slate, then notch the slate adjacent to this column at both upper and lower edges to make finding these cells easier.

Circuit for Generating Linear Dials

The following voltage standard can generate a dial with 20 divisions. It uses 14 1K precision resistors, and an 11-position single-pole rotary switch to get 10 major divisions, and a DPDT toggle switch to offset its output by one half a division.

Ten 1K resistors are connected between adjacent positions on the 11-position switch; position 1 goes thru 1K to position 2, which goes through 1K to position 3, etc. Position 1 of this switch goes through 2-1K resistors in parallel to ground. Position 11 goes through 2-1K resistors in parallel, then through 120K to the cathode of the zener diode in the meter reader.

The swinger of pole #1 of the DPDT toggle switch goes to position 1 of the rotary switch, while the swinger of pole #2 of the toggle switch goes to position 11. Also on the toggle, position 1 of pole 1 is grounded. Position 2 of pole 2 goes to the junction of the 120K resistor and the paralleled 1K resistors.

The arm of the rotary switch goes to the hot input of the meter reader, with its cold input being grounded.

Operation

First things first, the meter reader must be calibrated so that full-scale on the above circuit or full-scale on a meter gives you a null with the precision POT fully clockwise. (In the above circuit, do this with the toggle in position 1.)

Now set the system up for half-scale and square the blank up with the board (which puts mid-scale at the middle of the top of the blank).

Adjust the board slate so that a cell in the aforementioned column gives you the desired radius.

Make sure your set screw is tight and set the equipment up for zero. Step by step, change the input, null the blank, and make your markings.

You can change the scale factor by increasing the series resistance (now 120K) so as to make dials and portions of dials with different angles of rotation.

Auditory Volume LevelIndicators (Part II)

The stereo version of the Smith-Kettlewell circuit still only has three IC's; an LM324 quad op-amp (four opamps in a package), an NE556 dual timer (two 555's in a package), and a uA7805C voltage regulator.

The one advantage this circuit has over earlier schemes is that the two channels can easily be matched. Rather than hand picking matched components for the two VCO circuits, an adjustment can be included in one of the controlling op-amps to aid in matching their sensitivity.

Smith-Kettlewell Stereo Indicator Circuit

The negative side of the battery is grounded along with the common terminal of the uA7805C The input of this 7805 goes to the +V line (from 7 to 30V). The output terminal of the 7805 is the 5V line, and is bypassed to ground (near the regulator) by .01uF.

Pin 7 of the 556 and pin 11 of the 324 are grounded. Pin 4 of the 324 goes directly to the +V line. Pins 14, 10, and 4 of the 556 go to the 5V line.

The 5V line also goes through two resistors in series to ground (470K), with the junction of these two resistors going to pins 3 and 12 of the 324. Pins 3 and 12 are also bypassed to ground by .05uF in parallel with 5uF (negative side of the electrolytic grounded).

The cold side of each audio input is grounded, along with the bottom end of each 100K calibration POT. The hot side of the channel 1 input goes to the top of its calibration POT, and the hot side of channel two goes to the top of its calibration POT.

The arm of the channel 1 POT goes through .47uF, then through 47K to pin 2 of the 324, with pin 2 also going through 910K to pin 1. Pin 1 goes to the anode of a diode (1N914). The cathode of this diode goes through 10K, then through the parallel combination of 820K and 1uF to ground (negative side of the capacitor at ground). The top of this parallel RC combination goes to pin 5 of the 324.

The arm of the channel 2 POT goes through 0.47uF, then through 47K to pin 13 of the 324, with pin 13 also going through 910K to pin 14. Pin 14 goes to the anode of a 1N914. The cathode goes through 10K, then through another parallel combination of 820K and 1uF to ground (negative of the capacitor at ground). The top of this parallel RC combination goes to pin 10 of the 324.

Pin 3 and pin 11 on the 556 are the "control voltage" terminals from which we get the 2/3V firing point of each oscillator (see Part I). Pin 3 of the 556 goes through 100K to pin 6 of the 324, with this pin 6 also going through a 100K feedback resistor to pin 7 of the 324. Pin 11 of the 556 goes through 100K to pin 9 of the 324, with pin 9 going through a variable feedback resistor to pin 8 of the 324. This variable resistor consists of 82K in series with a 50K rheostat between pins 8 and 9, and the rheostat can be used to match the responses of the VCO's for levels exceeding 0 VU.

Pin 7 of the 324 goes through 47K to pin 1 of the 556, with pin 1 also going through 10K to pins 2 and 6 of the 556 which are tied together. Pins 2 and 6 also go through .022uF to ground.

Pin 8 of the 324 goes through 13 of the 556, with pin 13 also through 10K to pins 8 and 12 of which are tied together. Pins 8 go through .022uF to ground.

One side of each loudspeaker goes to the 5V line. The other end of the channel 1 speaker goes through 47 ohms to pin 5 of the 556, the other end of the channel 2 speaker goes through 47 ohms to pin 9 of the 556. If you wish, the outputs of the 556 can each go through 47 ohms to a common loudspeaker. As before, 10uF coupling capacitors can be used so that the speaker can be referenced to ground.

Measuring the Program Level

I will now steal Mr. Swail's input attenuator which allow us to measure the level of the program material. A string of resistors is set up around a rotary switch to provide five steps of input sensitivity at 3dB intervals. For the stereo version, two such attenuators can be ganged together by building them onto a double-pole five-position switch.

The hot audio lead goes to position five of a five-position selector switch. Position five goes through 6.8K to position 4, which goes through 2.2K to position 3, which goes through 680 ohms to position 2, which goes through 220 ohms to position 1, which goes through 100 ohms to ground and to the cold side of the audio line. The arm of the switch goes to the top of the calibration POT at the input of the level indicator.

The next installment in this series will cover circuits of LED VU meters and their adaptation for use by the blind and visually impaired.

Parts List (Stereo Indicator)

Capacitors:

Resistors: (1/4W)

Resistors: (1/2W)

Semiconductors:

Soldering (Part III) Tinning Stranded Wire

I have learned from experience that neglecting to tin stranded or braided wire before soldering it to a terminal is no shortcut. If the bundle of strands is not soldered together beforehand it is very likely that a few individual strands will not be involved in bonding to the terminal. This weakens the connection, and individual strands may subsequently stray over to an adjacent terminal.

In order for all the strands to be presoldered together, the stripped end of the wire must be bathed in fresh solder. As it is with tinning the iron, there is hardly such a thing as applying too much solder. A generous amount of solder can be wrapped around the end of the wire and heated. I keep a roll of very thin solder on hand for this purpose (about .03 inches in diameter). For small wire (above 22 gauge), I use a length of solder about two times as long as the stripped-back portion of the wire. I double this amount when soldering larger wire such as lamp cord or braid.

Don't be stingy with the length of wire from which you remove the insulation--remember, you will have to find this stripped end with the tip of a hot iron.

Give yourself a good 1/2 inch of bare wire beyond the insulation.

Twist the stripped end of the wire between your fingers to tightly bundle the strands. Using the two-to-one rule, fold the solder back against itself about an inch from the end, thus forming a long skinny hook of solder. Slide the wire all the way into the hook and position the bend of the hook just ahead of the insulation. Wrap the 1 inch long piece of solder in a neat coil around the bundle of strands, progressing towards the end of the wire.

Finding the wire with the tip of the iron is no easy task. Hold the wire in one hand the iron in the other, and rest both hands against a familiar referencing object. While bringing them closer and closer together, gently rock one of them up and down so that they are sure to connect as soon as they cross.

When they have met, put the bundle of strands on top of the tip of the iron and wait for the solder to melt. Melting of the solder will be indicated by the disengagement of the wire from the solder leading back to the spool, and by the usual "squeakiness" of solder-wet surfaces.

At this time slowly slide the iron out from under the end of the wire, allowing the wire to flip downward and shake off its excess solder. (This will tend to splatter solder, so don't do it in the direction of other people or the family pet.)

Tinning braided conductor deserves special mention. Taken together, the many strands in the braid present a large amount of surface metal with which the solder can alloy. In addition, these strands are in firm contact with each other--efficient heat transfer occurs between them. The result is that braid acts like a "wick" in the molten solder; bathing the end of the braid in solder will cause a surprising amount of the braid to become stiffly soldered together.

If you wish to tin merely the end of the braid without impairing its flexibility further back, one or two clip-on heat sinks can be attached just behind the length being tinned; this causes an abrupt decrease in temperature beyond which the solder cannot alloy with the strands of the braid. Remember to reattach a heat sink to the braid when soldering it to a terminal, since this "wicking" of the solder can still occur.

Wire Wrapping, A Construction Technique for IC Projects by David Plumlee

The mounting of IC's in a project can be a big stumbling block for many sightless technicians. How do you guarantee that you make all the connections you want, and only the connections you want on those devilish little DIP sockets with pins that sometimes seem closer together than the hairs on your head? If you have good luck soldering IC sockets into a board, congratulations, I envy you. My luck is more like burned fingers, solder bridges, and cold solder joints on those little pins. Plug-in solderless breadboards are rather expensive to use in all your projects.

Perhaps after reading this article you may want to explore wire wrapping for your IC projects. This technique is particularly useful when building digital electronic projects in which components other than IC's are rarely needed. However, projects containing many components external to the chips can be built by wire wrapping pig tails to the IC sockets and soldering the far ends of these leads to the other components where there is more room on the board.

Incidently, I feel strongly that no IC should be soldered into the circuit directly. You can easily damage it with the heat, and its presence in the circuit will probably invalidate any ohmmeter reading-taken to test for a solder bridge (two pins may even be common on the chip). If a chip later goes bad, servicing the unit is a real headache.

It must be understood at the beginning that some manual dexterity is needed to do wire wrapping. However, with practice I feel that many of you will quickly be able to acquire this skill.

In this article, we will consider only a low-priced wire-wrap tool--namely the Radio Shack #276-1570 (Priced near $6.00). It is capable of both wrapping and unwrapping. This is the tool I use for my wire-wrap work, and it will-provide an inexpensive way to see how you and wire wrapping get along.

If you want to try a more elaborate unit, there are many to choose from. Some tools even make a tiny slit in the insulation of the wire as they wrap, thus eliminating the need for wire stripping. I would suggest talking to a friend or dealer who is very familiar with these tools. (Radio Shack sells a battery-operated unit for about $22.00, #276-1571, and a hand-operated slitting type tool for about $13.00, #276-1572.)

In our discussion we need to consider the post, the wire, the tool, and the technique.

The posts are 25/1000 of an inch square, and are about an inch long. This square shape is necessary so that the wire will form square corners as it goes around the post, thus preventing the wire from unwrapping spontaneously. The sharp corners of the post actually nick the wire as it bends around them, thus assuring extremely firm contact of the metals. The posts themselves are either tin or gold plated so as to resist oxidation.

You can buy wire-wrap IC sockets from many suppliers. Be sure to specify "Wirewrap" when you order them. I have heard that such things as terminals, coil forms, etc. can be obtained in wirewrapping hardware. I am only using wire wrapping for IC's. (For this it is a big help, especially if a 28-pin device is to be installed.)

Ground Rules:

  1. DO NOT attempt to use anything for posts other than standard wire-wrap hardware. If you attempt to wrap to something else--the lead of a resistor, a small lug on a coil form or potentiometer, etc.--the wrap will not be shaped properly, and you may damage the tool in your attempt to get it on a pin for which it was not intended. You can, as mentioned, solder the far end of the wrap wire to a resistor, switch or component.
  2. Use only wire intended for wirewrapping. This wire is insulated with a tough plastic called "Kynar". This material resists abrasion when accidently brought up against the sharp corners of adjacent pins. In addition, its insulation will not stretch and contract as much as some, leading to fewer bare-wire segments next to pins. As for that other wire you have around the shack--well, this ground rule means what it says.

A good wrap will have about eight turns of the bare wire wound neatly around the square post with the turns touching each other. I emphasize here that these connections are not soldered, therefore, you must insure a firm wrap in order to have a reliable connection.

The wire for wire-wrapping is about 30 gauge. This wire is available in 50 foot spools, or it can be bought from some suppliers in assorted pre-cut lengths, 3 inches, 4 inches, 7 inches, etc. I use a spool of it for my work.

The tool has an overall length of about six inches. It has a hexagonal body with a tubular projection extending axially from each end. The shorter tube is used for unwrapping connections, and the longer one is for wrapping. You will notice that there is a groove running longitudinally along the outside of the wrapping end. When we later load the tool, the bare wire will be located in this groove (as we begin to turn the wrap).

Near the center of the tool's handle, a hole is drilled perpendicular to the axis. At one entrance of this hole, a rivet can be felt. This rivet secures a U-shaped blade designed to strip the insulation from the end of the wire. The notch of the U is only wide enough for the wire to pass, and the blade is sharp enough to cut through the insulation. Thus, as we shall soon see, wire forced into this notch can easily be stripped.

Using the Wire Stripper

To strip the wire, unreel a convenient length of it from the spool--do not cut it off. Hold the tool vertically; oriented so that the rivet is toward you and is below the hole. Perform the following actions in sequence:

Insert the wire into the hole, entering at the side facing you. Feed the wire through until about 3/4 of an inch of insulated wire is showing on the other side of the tool. To do so, you will need to keep the wire near the top of the hole (above the U-shaped blade).

When the wire is positioned as indicated, grip the wire near the rivet side of the tool. Pull down firmly--you will feel a resistance as the wire is forced to the bottom of the U-shaped notch in the blade. You have now cut and captured the insulation as implied above.

Pull down and out (toward you) to remove the wire from the tool. The insulation will remain with the tool (on the far side), and must now be removed from it. This procedure, if correctly performed, should give you about one inch of bare wire.

Practice stripping the end and clipping it off alternately a few times and you should become proficient at it. If you are tempted to grab that knife or conventional stripper, forget it-you will probably nick the wire, and it will break very quickly. The blade on the wire-wrap tool is hard to beat.

Loading the Tool

To better visualize the operation of inserting the wire into the wrapping end, let's draw a comparison between the tool and a crank. Think of the post around which the wire is wrapped as the axle about which the crank turns. Think of the longitudinal groove in the wrapping tube as the crank handle. Our job is to insert the bare wire into the crank handle from the open end of the tube. You may work out your own technique for inserting it, but let me guide you by suggesting one:

Hold the tool horizontally; oriented with the wrapping end pointing to your left, and with the longitudinal groove exactly at twelve o'clock (at the top of our imaginary circle). With your left hand, bring the end of the bare wire to the opening of the tube at the twelve o'clock point. Gently begin feeding the wire into the opening of the tube, making sure that the top of that wire is actually touching the twelve o'clock point of the inside of the tube. By guiding the wire gently with the left hand, you can actually move it around and search out the upper opening which you cannot touch by hand.

If you keep the wire at the top of the tube, you should enter the hole which is the handle of our hypothetical crank. Continue feeding the wire until the bare portion disappears into the tube. To verify correct positioning of the wire, rotate the tool about its axis with the right hand and observe the behavior of the wire with your left hand. If the wire's position is correct, it will ride around the circle described by the crank's rotation. If you have entered the axle hole by mistake, the wire will not be forced to describe this circle. Rotating the tool back and forth through one rotation, you can tell if the wire is positioned correctly.

Once again, position the wire at twelve o'clock (the top of the circle). Grip the insulated portion of the wire just beyond the entrance of the wrapping tube. Bend the tool up to a vertical position so that the wire makes a 90 degree bend at the end of the tool. Holding your hands in a constant relationship to each other, guide the tool to the post you want--then set it on the post. Do not rotate the tool yet: you must verify that you are on the correct post of the socket.

Pin Verification

I recommend double verification for any except an end pin. I illustrate this process as follows:

Suppose you want to wrap on pin 4 of a 16 pin socket. Once you have seated the tool on the post, locate pin 1 and count with your fingernail, "one, no; two, no; three, no; four, yes." Then from the other end, count "eight, no; seven, no; six, no; five, no; four, yes. " In practice, this verification can take place very rapidly.

I recommend that you strictly carry out the above verification sequence, at least in your first wire-wrap project. It may save you the trouble of having to correct a wiring error, or save you from having to buy a new chip because power to it was misapplied. Pay careful attention to the way pins are counted on the socket; remember you are working with the pins facing you.

Chip Orientation

I also suggest a rule of uniformity for chip orientation in your own construction projects. My suggested system is as follows:

Designate one edge of your board as the "head", and one edge at right angle to it as the "alternate". As much as possible, try to point all your IC sockets with pin 1 toward the "head" edge. If you must mount an IC at a 90 degree difference from this orientation, point its pin 1 toward the "alternate".

For clarification, suppose this page in the book were a circuit board lying on the bench with the pins of the sockets pointing up. If we designate the top margin as the "head", and the left margin as the "alternate", we can mount chips so that pin 1 would ideally be nearest the top. If it is necessary to mount a chip at 90 degrees to this orientation, we will have pin I near the left margin. (Remember that this view is with the pins up, that is, the component side of the board is lying on the bench.)

You may want to mark the head and the alternate sides with tape.

I realize that extremely high-frequency projects may require exceptions in the interest of short leads, but if you can use a rule of uniformity, it will simplify construction and troubleshooting.

Enough meditation on IC orientation and pin verification on company time.

Making the Wrap

Check to see that the bare wire rests in the longitudinal groove, now vertically oriented on the tool. Grip the insulated portion of the wire with your left hand and twirl the tool clockwise with the right. The bare wire will disappear down the groove as it wraps around the post.

CAUTION--exert no strong forces upward or downward on the tool as you wrap. Doing so may result in the following:

If you lift your hand as you twirl to help the wire climb in a coil, you will introduce spaces between the turns; these spaces will allow the connection to become loose. You may even lift the tool free before all the wire has been wrapped, and the tail thus formed may short to an adjacent pin.

On the other hand, if you exert a downward force on the tool, you may cause a second layer to form on the coil. The wire may bind and break after only two or three turns have been wrapped, giving you an unacceptable connection.

If there is any wire remaining in the wrapping end of your tool when you remove it, you have certainly broken that wire off the part you have stripped previously. I recommend that if that remnant is anything greater than 1/4 inch, you should unwrap it and do the connection over.

To help prevent riding the tool, I twirl the unwrapping tube between my right thumb and index finger. There is enough flexibility in these finger joints to allow the tool to climb by itself as it lays the coil. The cranking resistance imposed by the wire in the tool is surprisingly low, so the drive method I am advocating should be more than adequate.

Unwrapping

Unwrapping a connection is rather easy. Find the desired pin; place the unwrapping end of the tool on it and verify your position. If correct, spin the tool counterclockwise. This time, you can load the tool slightly downward (toward the board). The tool will catch the end of the wrap and begin to unwind it. If you can locate the wire you intend to remove (and be certain of it), grip its insulated portion and place a small amount of tension on it, as if to get it to unroll from around the post. Then, wonder of wonders, it will more than likely unroll, and you can feel it unwinding. If it does not unwind, it should at least loosen, so that you can place a fingernail alongside the post and work the loosely-wound coil up the post.

When you unwrap a connection, the end of the wire will have nicks and kinks in it where it bent sharply around the square post. I recommend that you do not re-wrap this same end of wire. Rather, I advocate measuring the wire so as to leave a safety margin, so that in case of an error you have enough wire to strip and wrap again. Measure your wire as follows:

After wrapping one end, lead the other end to the terminal or wire-wrap pin to which it is to be attached. Allow one extra inch for the wrap and one more for safety. As you gain proficiency, you may elect to pinch this safety allowance a little, or perhaps eliminate it all together.

Miscellaneous Considerations

You can generally wrap three or four connections on any pin. With a little practice, you can start a second one just above the top of the first one, and so on. Remember, at the top you will need a little margin to keep a turn from hopping off.

If one pin of several chips is common (such as VCC) , you can wrap on the first chip, then lead the end of this wire to the second chip. Another wrap on this pin of the second chip can give you a new wire to lead to a third chip., A second lead on this: pin of the third chip will prepare you for running to the fourth. Make your connection to theVCC line of the power supply to a chip in the center of this string of chips in order to minimize the lead length to any given chip. This pin in the middle of the string will have a third wrap on it to connect it to the power line. Remember that the wire is fine, so don't try to power ten current-hungry chips from one chain. Break it up and feed two wires back to the power line--otherwise, you will have problems with voltage drops and unwanted spikes introduced to parts of your project.

If multiple wraps on pins become a problem, you may consider wrapping only one wire on each pin and soldering these wires to nearby push-in terminals or leads of other components.

Of course, if one pin of a chip goes to another pin of the same chip, you should use a short length of wire to connect them, rather than cluttering the board with long leads running to a common point elsewhere.

Be consistent in your handling of connections to VCC in projects containing logic chips--gates, etc. You may set up a VCC terminal near each chip, or run a VCC bus around the board, or run individual VCC leads from each chip to a main VCC pin. Choose your method, then follow it through on all your chips. This consistency will simplify trouble shooting.

In regard to CMOS chips, I point out that you should not apply signal to such devices when power is not present. You do not want to open then line carrying power to the VCC terminal on the chip while still having other pin on the chip at logic "1" or VCC. In other words, if you remove power from a chip, make sure you can remove the voltage on its inputs.

Finally, don't forget about bypass capacitors. You will need about 0.1uF to ground from the VCC of perhaps every third or fourth chip down the line. If you are using shift registers or other chips with latch data, it would be well to bypass VCC of each of these chips, otherwise the data on the chips may change erratically. Inadequate bypassing can spoil your project and cost you several hours of de-bugging.

If you have a cooperative dealer in your area, you may get good advice from him in case of difficulty. I encourage you to establish good rapport with your supplier. If Radio Shack is your supplier, try to pick a store where the personnel seem helpful. If you study electronics and learn it well, you can generally earn the respect of a dealer you patronize frequently. In turn, he will often introduce you to new components and techniques.

I hope this article will pave the way for some of you who have hesitated to do construction of IC projects.