The Smith-Kettlewell Technical File

A Quarterly Publication of The Smith-Kettlewell Eye Research Institute’s Rehabilitation Engineering Research Center

William Gerrey, Editor

Issue: sktf-Winter-1984

Original support provided by: The Smith-Kettlewell Eye Research Institute and the National Institute on Disability and Rehabilitation Research

Note: This archive is provided as a historical resource. Details regarding products, suppliers, and other contact information are original and may be outdated.

Questions about this archive can be sent to


A Versatile Electronic Braille Dial Maker

An Audible Speedometer for Your Exercycle

Pedal Your Portable Music System

A Simple Cable and Pipe Finder

The Simplest Auditory Capacitance Bridge

List of Electronic Aids Designed at N.R.C. Canada

Bulletin Board


By Tom Fowle, WA6IVG

[Editor's note: Additional notes for this article were published in a later issue. Please click here to read these notes.]


A refined instrument is described which contains a "chopper-type" auditory comparison system and provides for a choice of two voltage divider systems. A wide variety of dial diameters, resolutions, and even nonlinear dials can be made with this instrument. [References: "Making Braille Dials," and "Basic Analog Meter Reader," SKTF, Spring 1981.]

The construction of circular braille dials to fit a wide variety of applications has long been a tricky mechanical problem for those who build instruments. A number of elaborate schemes have been used, most of which require precision machined parts that are expensive and difficult to obtain for the average tinkerer.

In the referenced article (see above), a board slate was modified by adding a precision linear potentiometer to which a plastic blank could be attached. The voltage at the arm of this pot was compared with a "standard" using a "null-type" meter reader. The standard was made up of a string of precision resistors which allowed making dials with a resolution of 5%. A rheostat placed in series with the precision divider string allowed the "full-scale" arc to occupy any desired portion of the circle.

The unit I built is self-contained (not a modified board slate), and several improvements have been incorporated. I wanted the ability to make dials of any diameter to fit existing equipment (the timer on my microwave oven, for instance)--the board slate was limiting in this respect. I wanted to be able to make dials having higher resolution than those from the original unit. Two methods of obtaining higher resolution dividers can be included:

The first, my favorite, is a 10-turn pot whose shaft is fitted with a pointer knob and a 20-division braille scale. Ten turns having 20 divisions per turn gives a possible 200 divisions, one every 0.5%. This is the system I included in my instrument.

An alternate scheme is to make a divider with fixed resistors; the circuit for this is also included. This system has three switches which work as follows: A 10-position rotary switch selects one through 10 major divisions. An 11-position switch divides each of these major sections into minor divisions. A toggle switch shifts the entire divider by 0.5% so as to generate midpoints for the minor divisions.

Photograph of the Braille Dial Maker

Physical Description

The unit is built on and in a heavy Bakelite box measuring 8-1/2 by 8 by 2 inches. The precision 360-degree pot which carries the blank is mounted through the top of the box near the lower middle. A slide which carries a braille slate is secured to the top of the box above the pot shaft so as to form the radius of a circle whose center is the pot shaft.

This slide is made from a piece of 1-inch aluminum angle stock and is about 7 inches long; the end opposite the pot shaft protrudes off the top edge of the box by about 2 inches. The slide carries a plastic block" to which is glued the braille postcard slate.

The controls and precision divider are mounted along the right side of the box. The controls are, from front to back: volume, mode switch (power control and scale position selector), top of scale adjust, and the precision divider. The meter reader board and battery are inside the box. The speaker is mounted on the close end of the box, so the blast of sound when you get out of null can hurt your ears and make you lose track of which dot you just punched.

With this arrangement, the left hand can be used to manipulate the dial blank while the right hand moves between punching dots through the slate and adjusting the precision divider and other controls on the side of the box. Since the divider in my unit (a 10-turn pot) is mounted on the side of the box, its dial is limited in size to the 2 inches dictated by the height of the box. Since the dial has only 20 divisions in 360 degrees, it is quite easy to read.

Building the Braille Slate Assembly

It is necessary to mount a braille slate at right angles to the aforementioned aluminum slide such that a braille letter "1" (dots 1, 2, and 3) written on the slate will be on the radius of a circle whose center is the shaft of the precision pot. The diameter of the dial to be manufactured is set by the position of the slate along the slide.

I used a braille postcard slate. The length of the line of the slate has some bearing on the maximum size dial which can be made. If a very large blank having square corners is used, the corners of the blank might run into the slate hinge as the blank is rotated. (This should rarely be a problem, however, unless you plan to make dials for Big Ben.) I glued the backside of the open end of my slate to a block of Plexiglas which is fitted to the aluminum slide and held in position with a bolt and wing nut so that it can be secured appropriately for the desired diameter.

To make the slide, I obtained a piece of aluminum angle stock about 8 inches long and 1 inch on a side. I also cut a piece of Plexiglass 1 inch square--long enough to match the width of the slate.

Place a piece of sandpaper inside the aluminum angle with the paper side toward the aluminum and put the plastic block inside the sandpaper. Rubbing the plastic block along the paper and aluminum vigorously will match the shape of the block to the aluminum so it will ride smoothly and not rock in the slide.

Remove the sandpaper and fit the block into the slide, rocking it back and forth in all directions to test for a good fit. Repeat this sanding process several times with finer grades of paper to make a smooth-running slide.

You must now decide which of the outer surfaces of the angle stock will be mounted on the top of your box. We will call this surface the bottom of the slide; the slate will be glued to the opposite surface of the block. You must now cut a slot along the side of the slide (the vertical portion of the angle stock) which will carry the bolt that secures the block in position. The procedure for making the slot is as follows:

Clamp the block into the slide at one end, placing the clamp around the bottom surface of the slide and the top of the block. On a drill press table, place this unit on its side (aluminum side down) and drill a hole straight down through the center of the block and the side of the slide. Start with a small drill and work up to a 1/4-inch hole.

It is now necessary to continue the hole made in the angle stock along most of the length of the slide to form a slot. This can be done in several ways:

I used a lethal little gadget known as a "drill saw." This has an end like a 1/4-inch bit, with a shank full of sharp, abrasive teeth, so that it can be pulled sideways through the material. On the drill press table, I made up a jig with a block of wood and C-clamps, which allowed the angle stock to slide so that the drill saw would cut a straight line along it beginning with the hole drilled together with the block. This needed some careful alignment and required a good deal of force. In using this tool, be sure you stop before the drill saw cuts through the entire length of the slide.

A somewhat safer but more tedious approach would be to use the block as a guide for drilling a number of holes along most of the length of the slide, then removing the material between the holes with a hacksaw or a coarse rat-tail file. Finish the edges of the slot with a fine file and emery cloth, so that the bolt which comes through the block rides easily along the entire length of the slot. Make sure that the ends of the slot are not so close to the ends of the angle stock as to impair its strength.

A 1/4-20 bolt fitted with flat washer and wing nut passes through the block and the slide to secure the block in any desired position.

For mounting the slide to the box, holes must now be drilled through the bottom of the slide. To allow the block to pass over the screws, countersink them for flat head screws. The position of the slide in relation to the pot shaft should be chosen so that one of the columns of cells falls directly on a radius of a circle whose center is the shaft. You might choose to file a notch in the top segment of the slate adjacent to this chosen column of cells for easy identification.

It is now necessary to epoxy the braille slate to the top of the block. I roughened both the back of the slate and the top surface of the block slightly with a coarse file. Clean the surfaces thoroughly with alcohol and arrange a clamp to hold them in place once you have applied the glue. Since the surface of the slate is bumpy, there will not be a great deal of contact area, so everything possible should be done to ensure a good bond.

Building a Clamp for the Dial Blank

It would be nice if your linear precision pot had a shaft a couple of inches long. If not, an extension with a shaft coupling must be used in order that the shaft extend up, well past the slate. This jig is made as follows:

The center of a plastic dial blank is fitted over a threaded panel bushing, the bottom of which is glued to a thick shaft collar made from the metal center of an old knob. This assembly is slipped over the extended shaft of the precision pot and adjusted in height so that the dial blank rests in the braille slate without bending. The set screws in the shaft collar should be replaced with thumbscrews to make positioning the blank easier.

Find a knob with a metal insert and crack the knob so as to remove the insert. You are in luck if one end of this collar is flat so that it can be cemented to the panel bearing. If it is not, file one end until a flat rim is available. This collar and the panel bushing must be cemented together in proper alignment -so that they will slide over the pot shaft and rotate easily.

Apply a thin coating of Vaseline on a piece of 1/4-inch shafting; this will prevent cement from sticking to the shaft. Slide the panel bearing and shaft collar on the shaft, apply Epoxy to the mating surfaces and press them together firmly. (Of course, the threaded portion of the panel bushing should be away from the shaft collar.)

After the glue has set, remove the greased shaft and clear away any cement inside the assembly with a 1/4-inch drill bit. Fit the threaded panel bearing with a pair of large flat washers, between which the plastic blank may be clamped; this assembly is secured with the original nut from the panel bearing.

Circuit Operation

Both the single-turn pot which carries the dial blank and the precision divider network (the standard) are supplied from a Zener diode regulated voltage reference. The standard supply is made variable so as to provide a "top-of-scale adjust." The setting of this adjustment determines the number of degrees covered by the full braille scale.

The voltages on the arms of the one-turn pot and precision divider are compared using a chopper amplifier system. The difference voltage between the two determines the volume of a tone heard in a speaker. Thus, when the voltages are identical, the tone is minimized, producing a null.

To facilitate alignment of the dial blank, a switch is provided which allows rapid selection of fixed positions at the top, bottom, and midpoint of the scale. Rather than going directly to the arm of the precision divider, the chopper input actually goes to the swinger of a 4-position rotary switch. The first position of this switch connects to the bottom of the precision divider; the second position of the switch connects to the center point of the divider; the third position of the switch connects to the top of the divider; and, finally, the fourth position goes to the divider's arm. As expected, the other chopper input goes directly to the arm of the one-turn pot.

Actually, the aforementioned 4-position single-pole switch is 5-position 2-pole switch, with the other pole being used as an on-off switch. Positions 2 through 5 of one poles are wired as mentioned in the previous paragraph, while positions 2 through 5 of the other pole are tied together and go to the VCC line. The arm of this latter pole goes to the positive of the 9-volt battery, leaving position 1 (off) of both poles unused.

Circuit Description

The negative side of the 9-volt battery is grounded. The positive battery terminal goes to the arm of pole No. 2 of the 5-position switch (see above); positions 2 through 5 are tied together and go to the VCC line.

Pin 1 of an NE555 is grounded, while pins 4 and 8 are tied together and go to VCC. Between pins 1 and 8, and located next to the chip, is 0.1uF. Pins 2 and 6 are tied together and go through 0.01uF to ground. Pins 2 and 6 also go through 100K to pin 7. Pin 7 goes through 100K to the VCC line. Pin 3, the output, goes through 2.2K to the cathode of the LED contained in the FET Opto-Isolator (this cathode being pin 2 of a GE H11F3). The anode, pin 1, goes to the VCC line.

Pins 2 and 4 of an LM386 amplifier chip are grounded. Pin 6 goes through 10 ohms to the VCC line, with pin 6 also being bypassed to ground by 250uF (negative at ground). Pin 7 is also bypassed to ground by 25uF (negative at ground). Pin 5, the output, goes to the positive end of a 100uF electrolytic capacitor, with the negative end of this capacitor going through the speaker to ground. Between pins 4 and 5, and located close to the chip, is 0.22uF. Pin 3, the input, goes to the arm of a 10K audio taper volume control; the bottom of this control is grounded. Pin 3 is also bypassed to ground by 0.01uF.

The swinger on pole No. 1 on the 2-pole 5-position switch mentioned above goes through 47K to one end of the Opto-Isolator's FET channel (pin 4 of the H11F3). The other end of the channel, pin 6, goes to the arm of the one-turn pot which carries the blank; the bottom of this one-turn pot is grounded. Pin 4 of the H11F3 also goes through 0.1uF to the top of the volume control.

The top of the one-turn pot goes through a 5K rheostat to the cathode of a Zener diode (5.6, 6.2, or 6.8V). The anode of this Zener is grounded, while its cathode goes through 470 ohms to VCC.

The cathode of this Zener goes to the top of a 10K linear taper pot (top-of-scale adjust) whose bottom is grounded. The arm of this pot goes to the top of a 10-turn 10K calibrated pot whose bottom is also grounded (this is the precision divider which is fitted with the braille scale). The junction between the arm of the top-of-scale adjust and the top of the precision divider goes to position 4 of pole No.1 of the aforementioned rotary switch. The arm of the precision divider goes too position 5 of this pole. Position 2 of this pole is grounded.

My 10-turn pot came nicely equipped with a center tap which I connected directly to position 3 of the rotary switch, giving a handy middle-of-scale reference. If you are not so lucky, you can connect a divider made up of two well-matched resistors of any value over 10K. In other words, two precision resistors are connected in series across the 10-turn pot. The junction of these two resistors goes to position 3 of pole No. 1 of the rotary switch.

The second divider scheme is done with switches instead of the 10-turn pot. This has two advantages: One is that a braille dial need not be made for the calibrated divider. The other is that these switches are perhaps easier to keep track of than the easily jostled 10-turn pot. This system has a very significant disadvantage; resistors required must be of extreme precision, 1/10% being preferred, and the switches must also be of high quality. However, the switching network is interesting, and you may find uses for it -- if not here, elsewhere.

This divider contains three switches: S1 is a 2-pole 10-position rotary switch which selects increments of 10%. S2 is a single-pole 11-position rotary switch which selects increments of 1%. S3 is a double-pole double-throw toggle which offsets the entire divider by 0.5%.

Circuit for Switchable Divider

The arm of the 10K pot (top-of-scale adjust) goes through a 50-ohm resistor to position 10 of deck B of switch 1. Position 10 of deck B goes through 1K to position 10 of deck. Position 10 of deck A goes to position 9 of deck B. 9B goes through 1K to 9A. 9A goes to 8B. 8B goes through 1K to 8A. 8A goes to 7B.

7B goes through 1K to 7A. 7A goes to 6B. 6B goes through 1K to 6A. 6A goes to 5B. 5B goes through 1K to 5A. 5A goes to 4B.

4B goes through 1K to 4A. 4A goes to 3B. 3B goes through 1K to 3A. 3A goes to 2B. 2B goes through 1K to 2A. 2A goes to 1B. 1B goes through 1K to 1A. 1A goes through 50 ohms to ground.

Position 1 of S3A (toggle) is grounded. Position 2 of S3A is not used. The swinger of S3A goes to the junction between the 50-ohm resistor and position 1, deck A of S1. Position 2 of S3B goes to the junction between the 50-ohm resistor and position 10, deck B of S1. Position 1 of S3B is not used. The swinger of S3B goes to the junction between the 50-ohm resistor and the arm of the 10K pot.

The arm of S1, deck A, goes to position 1 of S2; the arm of S1, deck B, goes to position 11 of 82. Position 1 of 82 goes through 10K to position 2 of 82, then through 10k to position 3, etc. In other words, ten 10K resistors are placed between adjacent positions on S2. The arm of S2 is the take-off point of the network; it goes to position 5 of the mode switch. The midpoint of this network, which goes to position 3 of the mode switch, is taken from switch 1, deck A, position 6.

Calibrating the 10-Turn Pot

If the 10-turn pot is used instead of the switchable divider, a braille scale must be made for it. This isn't as bad as it sounds; after the unit is completed you can use the 10-turn pot to make its own dial.

Fit the 10-turn pot with a pointer knob which has approximately equal length ends (only one of which is pointed, of course) and place a piece of tape or other temporary marker at the location of the pointer when the pot is set to 0. Not only can you count full turn, but half-turns are indicated as well when the back of the knob is lined up with the marker.

Using the machine, you can now make a nice 20-division dial whose diameter is set to match the pointer on your 10-turn pot. When the 20-division dial is fitted to the 10-turn pot, the dial maker is equipped to make dials with marks at increments of 0.5%.

Using the Dial Maker

In principle, the steps required are:

  • Mount the plastic blank to its jig and slide this assembly onto the 1-turn pot, adjusting its height to be appropriate for the slate.
  • Move the slate so that a cell in the preferred column has its nearest dot (dot 3) at the desired radial distance for making the dial.
  • Set the "top-of-scale adjust" (mounted on the right panel) so that the full range of the calibrated divider covers the desired number of degrees in the eventual braille dial.
  • Finally, use the calibrated divider to "dial in" desired markings to be made on the blank; adjust the blank each time for a null and make your mark.

Those are the operations in principle. There are several fine points to be considered along the way:

Setting the Full-Scale Angle

First of all, a 5K rheostat in series with the 1-turn pot which holds the blank has been included so that you can assure that the calibrated divider can be made to represent 360 degrees. For example, if your 1-turn pot has a significant gap between its upper and lower ends, the above rheostat can be adjusted so that at least the angular relationship between the divider and the pot is preserved.

This relationship can be tested by setting the mode switch to the half-scale position and assuring that the instrument nulls with the 1-turn pot turned 180 degrees up from its lower end. This can either be measured by means of a protractor or by installing a square blank, and squaring this blank up with the edge of the slate, first for zero and then for 180 degrees. Adjust the SK rheostat as necessary to achieve this result. (For this test, the top-of-scale adjustment should be set at maximum.)

The full-scale angle for which the dial is to be made is perhaps best measured with a protractor. If the dial is to be made for a potentiometer, make sure to measure the full "electrical rotation" rather than the mechanical rotation; they are often different. This can be done with the aid of a protractor, a pointer knob, and a continuity tester.

Assuming the desired angle of the dial is known, the top-of-scale adjust can now be set so that the full excursion of the calibrated divider causes nulls to be obtainable as the 1-turn pot is rotated through this desired angle. First calculate what percentage of a 360-degree circle the desired angle covers, after which adjustment of the instrument can be made.

For example, if a 300-degree dial is needed, note that 300 over 360 is 0.833, or 83.3% of full scale. With the top-of-scale adjust at maximum, set the calibrated dial for 0.833 of its full scale. Turn the 1-turn pot to obtain a null and leave it there. Now set the mode switch to the top-of-scale position, position 4, and adjust the top-of-scale pot control for a null again. Now, the full range of the divider will cause the 1-turn pot to track through 300 degrees of rotation.

It is worth noting that if the 1-turn pot can be replaced with a potentiometer for >which you wish to make a dial, generating the dial can be done directly with the top-of-scale adjustment at maximum; problems such as measuring the "electrical rotation" and nonlinearities of the pot are accounted for automatically.

Squaring Up the Blank

If you are using a blank with square edges, the midscale position can be arranged so as to land in the center of the top edge of the blank. After the top-of-scale adjustment is made, and the blank is installed on the jig, slip the jig over the pot shaft, but do not tighten the thumbscrews. Align the top edge of the blank with the edge of the slate, then set the mode switch for the half-scale position (position 3) and adjust the 1-turn pot for a null while holding the blank very still. Tighten the thumbscrews vigorously and rotate the blank back and forth slightly to assure that the null occurs with the blank properly aligned.

General Comments

In setting the position of the slide that carries the braille slate, position a convenient cell in the preferred column so that its dot 3 is slightly farther from the pot shaft than the length of your pointer. This should be done so that the pointer does not cover up the braille markings against which it is read.

Linear dials can be made by advancing the calibrated divider by regular intervals and making markings of 1, 2, or 3 dots. Major divisions of the dial should have the largest number of dots.

Nonlinear dials can be made by calculating the positions of desired marks and converting them to percentages of the full-scale angle-adjust the calibrated divider to these percentages and make the marks.

A wide range of materials can be tried for your dials. A heavy grade of plastic is available from American Thermoform Company. Various clear plastic covers for stationery folders can often be found in office supply stores. Although very thin, plastic with adhesive backing is available, intended for "laminating" licenses and wallet items; this has the advantage of not needing cement or screws to hold the eventual dial in place. Scraps of plastic from unlikely sources such as lunch meat packaging are always worth keeping in mind.

When you mount the blank in its clamp (on the panel bushing), be sure that the mounting hole is just large enough to fit over the threaded panel bearing. Even with the clamp tight, a large hole will invariably result in slippage of the blank, creating a dial which is out of round.

As you familiarize yourself with the process of making dials, you will undoubtedly, as we have, make many mistakes. This most often happens when you forget which kind of mark you want to make next, lose track of the position of the divider, or use the wrong cell on the slate. In many cases, these erroneous dots can be sliced off using a scalpel or a razor blade (preferably single-edged). This creates much less damage than trying to flatten a dot as you would in erasing paper braille.

Although this device and the use of it may seem somewhat formidable, after lousing up a few nice pieces of plastic and spending a few sleepless nights, you will begin to get the hang of it. The ability to create your own braille dials suited to your individual haptorial pleasure is a rewarding experience well worth the trouble. Ah' Ah' Ah! Do touch that dial!

Parts List

Resistors (1/4 watt, 5%):

  • 1--10 ohm
  • 1--470 ohm
  • 1--2.2K
  • 1--47K
  • 2--100K


  • 1--5K trim pot
  • 1--10K audio taper volume control
  • 1--10K linear nonprecision panel mount
  • 1--10K 10-turn precision pot fitted with pointer knob and braille scale, Bourns No. 3540S-1-103
  • 1--Single-turn precision linear pot, preferably with a long shaft, Bourns No. 35358-744-103 (with 1/8-inch shaft) or Bourns 6657S-1-103 (with 1/4-inch shaft) or Beckman No. 6187R10K


  • 2--0.01uF disk
  • 1--0.1uF disk or Mylar
  • 1--0.22uF disk or Mylar
  • 1--25uF 12V electrolytic
  • 1--100uF 12V electrolytic
  • 1--250uF 12V electrolytic


  • 1--Zener diode, 1N5325 or Radio Shack 276561 (voltage not critical)
  • 1--GE H11F3 Opto-Isolator
  • 1--NE555
  • 1--LM386


  • 2-pole 5-position switch
  • Panel bushing, shaft collar, thumb-type set screws, and two large flat washers
  • 7-inch long piece of 1-inch aluminum angle stock
  • 1-inch square block of Plexiglas whose length is the width of the braille slate
  • 1-1/2 inch long 1/4 20 bolt with wing nut and flat washer for the above plastic block
  • Postcard braille slate with "locating pins" removed


  • The Bourns pots and the Opto-Isolator are available from Hamilton Avnet, 1175 Bordaux Drive, Sunnyvale, CA 94086; phone: (408) 743-3350.
  • Beckman is distributed by Wiley Distributing Group, 3000 Bowers Avenue, Santa Clara, CA 95051; phone: (408) 727-2590.
  • American Thermoform Company, 8640 East Slauson Avenue, Pico Rivera, CA 90660; phone: (213) 277-0516.

[Editor's Note: Yes, I know it's long; I'm sorry I wrote the original!]



Exercise bicycles often come with a booklet containing a formalized regimen of exercise; they have charts listing speed, time, etc., on which the user can plot his performance. If a blind user is inclined to follow the program in the book, he needs a speedometer he can read. This instrument has a switch that can be set to anyone of six speeds (in 5 MPH steps). Once set, a low tone indicates pedalling slowly, and a high-pitched tone indicates going too fast; the instrument remains silent as long as the desired speed is maintained (pedalling within 2.5 MPH of the speed selected). A tactile version is also described, presenting the fast and slow signals to buzzers on the handlebars. For those not interested in the specific instrument, it is worth noting that the device contains an elemental tachometer of the "integrator" type.

The Transducer

A very common low-cost "transducer" which can convert the speed information into an electrical signal is the little generator that cyclists use to power their running lights. These are actually alternators (producing AC), and they are manufactured by many concerns. (The local Schwinn dealer sold me two models of what they call "block generators," one made in France and the other of Asian manufacture.)

They are usually a bottle-shaped affair, with the "bottle cap" being a knurled drive wheel that runs against the side wall of the front tire. A clamp arrangement secures the unit to the "fork" (the structure that supports the bicycle wheel bearing). A strong spring and a pivot allows the bottle-shaped alternator to tip over against the tire and affords firm contact with it. A latch is available-for disengaging the alternator; when the cyclist wants his lights off, he tips the bottle-shaped unit away from the tire and allows the latch to hold it in check.

The frame of the usual exercise bicycle is much heavier and is shaped differently from that of a velocipede. Mounting the alternator on the exercycle will no doubt require modification to its clamp. Longer screws for the mounting clamp will be needed; this may also mean that the screw holes in the clamp will have to be enlarged to accommodate larger bolts. In addition, the "fork" of the exercycle is usually straight, not curved like those on well-designed bicycles. A swivel on the alternator's clamp, which permits adjustment so that the drive wheel is perpendicular to the tire, may have to be modified to extend its range of adjustment.

Given a modest tool chest--a reamer here, a small rat-tail file there--the above changes can be made by the home handyperson. If you trust them, let the folks at the bike shop mount the alternator for you: ignore the looks on their faces. (Tell them your dog needs the light to read by. If they believe you, go to a different bike shop.)

Very often, the lamp socket and reflector assembly is mounted with screws to the alternator; you may remove them if you wish. One of my Schwinn units has everything molded together so that the lamp socket cannot be removed. It is of no consequence whether the lamp is left intact, since the circuit depends on the frequency of the AC voltage for its speed information.

The units live seen have a binding post for connection of additional lights; this binding post is the "hot" output. The body of the alternator and its clamp is the "cold" output; a ground screw is usually supplied to assure good contact with the frame of the bicycle. A short machine screw can be substituted for the ground screw to provide you with a binding post for the cold output.

The output voltage, for some odd reason which defies my training in electromagnetism, is not a direct function of speed. (This is lucky for the light bulb, which would quickly burn out in a downhill run.) With the 1/2 amp bulb in place, the alternator acts much like a current source. With the lamp disconnected, the output is still not a linear function of speed, although it can reach 25 volts or so. (My first prototype used a rectified version of the output voltage compared with a selectable standard; the above nonlinearity prevented this system from working.)

Circuit Operation of the Tachometer

We chose to use an "integrator type" tachometer to convert the alternator?s frequency into an analog voltage--then compared this with a selectable standard. A wide variation in frequency vs. pedal speed can be expected, depending on the cycle and the alternator chosen. In fact, comparing the two Sears cycles with which I have had experience, there is considerable difference between pedal speed and speedometer reading. The circuit shown was designed around the Sears Model 374.28535 in combination with the Schwinn "block generator," Model 04250. Our measurements show that pedalling at 20 MPH gives us an output frequency of 600Hz. Of course, information on how to design the circuit to fit your particular cycle alternator combination will be given in "Design Considerations."

The output wave of the alternator is first squared up and brought to full logic levels by a comparator--an RCA CA3130. This op-amp was chosen because of its ability to operate with its FET input taken below the negative supply, thus eliminating the need far a split power supply and/or a more complicated input network. The CA3130 is then used to trigger a one-shot whose output pulses are "averaged" in a simple RC series circuit. The more often the one-shot is triggered (its on-time being fixed), the higher will be the average charge left en the integrator capacitor. Thus, the voltage on the capacitor is a function of the frequency at which the one-shot is triggered--and hence the frequency of the alternator's output.

The above one-shot (one-half of a 556 dual timer), along with the series RC circuit on its output (470K and 1uF), constitutes a classical tachometer circuit. It is fairly linear over a frequency range of perhaps 10 to-1, its precision dropping off rapidly at the lower end.

The output of the tachometer goes to one input each of two comparators (contained in an LM358 dual op-amp); the comparators' other inputs go to a switchable voltage divider which constitutes a "selectable window." When the tachometer output is within the "window," the polarity of the comparator inputs is such that both their outputs are low. The output of one comparator goes high when the tachometer falls below the window; the output of the other goes high when the tachometer exceeds the window.

In our auditory version, the comparator outputs go though separate charging resistors which drive the audio oscillator (the other half of the 556) at two different frequencies. A diode in series with each resistor is required so as to disengage the comparator whose output is low; these diodes perform an "or" function. In the tactile version, the output of each comparator goes to the control terminal of its respective star Micronics buzzer.

An automatic on-off switch is included using a VMOS power FET in conjunction with an NPN control transistor. Each positive half-cycle of the alternator turns on the transistor and refreshes the charge on a capacitor in the FET gate circuit. When the alternator comes to a stop, forward bias on the FET gate bleeds off and the switch opens.

Design Considerations

Given an available combination of exercise cycle and alternator, it is required that the tachometer be designed in accordance with the expected range of frequencies. The determination of frequency versus speed does not have to be exact, however, only approximate.

It is unlikely that a frequency counter could be used to measure the frequency, since good speed regulation while pedalling is impossible. We simply listened to the alternator's output through a loudspeaker; one of our staff members has perfect pitch and he simply told us what frequency we had when pedalling at 20 MPH. Another way is to pedal so as to match a known frequency and then read the speedometer. For example, 4th octave B on the piano is about 494 Hz (close enough to be considered 500 Hz); our speedometer would then have read about 17 mph.

The listening setup consisted of connecting a 4-inch loudspeaker in series with a 1-watt resistor of 82 ohms--this combination placed across the alternator's output. A test amplifier can also be used; remember that the unloaded generator voltage can go as high as 25V. (With the lamp in place, the generator voltage is clamped very effectively at 6V.)

Using the information gotten from the above arrangement, figure out what frequency would be generated at 40 MPH and consider this to be the maximum (in our case, this turned out to be 1200 Hz). Next, find the period of one cycle by taking the reciprocal of this frequency (1 over 1200 equals 833 microseconds). The components around the one-shot can now be chosen so that its on-time does not exceed this period.

The formula for finding the period for a 555-type one-shot is 1.1 x R x C, where R is the charging resistor and C is the timing capacitor. Given our choice of 13K and 0.05uF, the period of the one-shot is 715 microseconds (less than 833 microseconds, which is what we wanted).

If eventual calibration of the device proves troublesome, either or both R and C can be adjusted. For example, if the window is always too low and cannot be brought up to match the speed selected, the components of the one-shot can be increased in value.


Circuit for the Audible Version

The cold side of the alternator is grounded, while its hot terminal goes through 100K to the inverting input of an RCA CA3130 op-amp. This inverting input is clamped with diodes so as not to be damaged by voltages beyond the supply lines. In other words, pin 2 goes through a diode to ground (anode at ground); pin 2 also goes through a diode to VCC (cathode at VCC). The noninverting input, pin 3, is grounded. Pin 4 is grounded, while pin 7 goes to VCC.

Pin 6, the output of the 3130, goes through 0.001uF to pin 6 of the 556 (trigger of the first half). This trigger, pin 6, also goes through a 100K pull-up resistor to VCC. (Bringing the trigger pin down fires the one-shot.} Pins 1 and 2 of the 556 (discharge and threshold, respectively) go through "C" to ground (0.05uF in our prototype); pins 1 and 2 also go through "R" to pin 14 (13K in our unit).

Pin 7 of the 556 is grounded, while pins 14, 4, and 10 are tied together and go through 22 ohms to VCC. Pin 14 is bypassed to ground by 100uF (negative at ground). For further noise suppression, 0.1uF goes between the supply pins (7 and 14), located next to the chip.

The output of this half of the 556, pin 5, goes through 470K, then through 1uF to ground (a tantalum capacitor with its negative at ground). The junction of this resistor and capacitor is the output of the tachometer. This output point goes to pins 2 and 5 of an LM358 (the inverting input of A1 and the noninverting input of A2).

Pin 4 of the 358 is grounded, while pin 8 goes to VCC. Pin 3 (the noninverting input of A1) goes to the lower edge of the window, while pin 6 (the inverting input of A2) goes to the upper edge of the window. The selectable window is as follows:

A 2-pole 6-position switch is used.

On pole A:

  • Position 6 goes through 2K to position 5
  • Position 5 goes through 2K to position 4
  • Position 4 goes through 2K to position 3
  • Position 3 goes through 2K to position 2
  • Position 2 goes through 2K to position 1
  • Position 1 goes through 1K to ground

On pole B:

  • Position 1 goes through 2K to position 2
  • Position 2 goes through 2K to position 3
  • Position 3 goes through 2K to position 4
  • Position 4 goes through 2K to position 5
  • Position 5 goes through 2K to position 6
  • Position 6 goes through 1K in series with a 10K rheostat to VCC.

A 1K resistor goes from the arm of pole A to the arm of pole B; the arm of A is the lower edge of the window (it goes to pin 3 of the LM358), and the arm of B is the upper edge of the window (it goes to pin 6 of the 358).

The second half of the 556 is the audio oscillator; pin 9 goes through 47 ohms, then through the speaker to pin 14. Pins 8 and 12 are tied together and go through 0.015uF to ground. Pin 13 goes through 22K to pin 12; pin 13 also goes to two charging resistors as follows: Pin 13 goes through 100K to the cathode of a diode, with the anode of this diode going to pin 1 of the 358 (the output of A1). Pin 13 of the 556 also goes through 22K to the cathode of a diode, with the anode of this diode going to pin 7 of the 358 (the output of A2). (These diodes can be any general-purpose silicon units, 1N914, 1N4148, etc.)

The negative side of the 9V battery is grounded. The positive side of the battery goes to the source of a P-channel FET (Siliconix VP0300M). The drain of this FET goes to the VCC line. Between the gate and the source of the FET is the parallel combination of 100K and 1uF (positive of the capacitor toward the source). The gate of the FET also goes to the collector of an NPN transistor (2N2222), with the emitter being grounded. The transistor base goes through 100K to the hot output terminal of the alternator.

Circuit for the Tactile Version

(For those interested in the tachometer circuit only, this description is most appropriate.) Since most of the circuit is unchanged, only the timer and buzzer system will be described here. A 555 is used instead of a 556, and is connected as follows: Pin 1 of the 555 is grounded, while pins 4 and 8 go to VCC. Between pins 1 and 8, and located next to the chip, is 0.1uF. (Massive decoupling as in the previous circuit should not be necessary.)

Pins 6 and 7 (threshold and discharge) are tied together; pins 6 and 7 go through "C" to ground (0.05uF), with 6 and 7 also going through "R" to VCC (13K). Pin 2, the trigger, goes through a 100K pull-up resistor to VCC. Pin 2 also goes through 0.001uF back to pin 6 of the CA3130. The 555 output, pin 3, goes through 470K, then through 1uF to ground (tantalum capacitor with its negative at ground). The junction of this resistor and capacitor is the output of the tachometer; it goes to pins 2 and 5 of the LM358 as before.

Pin 1 of the LM358, the output of the low comparator, goes to the control terminal of one buzzer, while pin 7 of the 358, the high comparator, goes to the control terminal of the other buzzer. (Star Micronics CMB-12 buzzers were used.) The positive buzzer terminals go to VCC; their negative supply pins are grounded. Across each buzzer's supply pins (mounted on the buzzer itself) is the parallel combination of 10uF and 0.1uF (negative of the electrolytics at the negative pin).

These buzzers have four pins which are in the corner positions of a 14-pin DIP socket; they can be labeled 1, 7, 8, and 14 (the sound holes are nearest 7 and 8). Pin 7 has no connection; pin 8 is the control pin; pin 1 is the minus supply; while pin 14 is for the plus supply.

The buzzers can be loosely taped or hung from the handlebars (allowing them to vibrate freely). Their sound holes can be taped over to deaden the sound if desired.


As chosen by the circuit components, the selector switch can be set for speeds of 5, 10, 15, 20, 25, and 30 MPH. Get someone to pedal the cycle at one of these speeds and set the switch accordingly. Adjust the 10K rheostat associated with the 6-position switch so that both alarm signals are off. By slowing down and speeding up, the edges of the window can be found and "balanced" around the set speed. The window will cover a range of about 5 MPH.

Parts List

Resistors (1/4 watt, 5%):

  • 1--22 ohms
  • 1--47 ohms
  • 1--1K
  • 1--Value of "R" (we used 13K)
  • 2--22K
  • 5--100K
  • 1--470K

Precision Resistors (1%, very low wattage):

  • 2--1K
  • 10--2K


  • 1--0.001uF disk
  • 1--0.015uF disk
  • 1--0.05uF disk
  • 1--0.1uF disk
  • 2--1uF 10V electrolytic (one of these must be tantalum)
  • 1--100uF 10V electrolytic


  • 4--silicon diodes, 1N914, 1N4148, etc.
  • 1--2N2222
  • 1--VP0300M Siliconix VMOS Power FET
  • 1--00130
  • 1--LM358
  • 1--NE556


  • 1-10K trim pot as calibration rheostat
  • 1--2-pole 6-position switch
  • 1-"block generator" or whatever your dealer calls it

Tactile Version

(only the changes from the audible version are listed here):

  • 2--0.1uF disk capacitors
  • 2--10uF 10V electrolytics
  • 1-NE555
  • 2--Star Micronics CMB-12 buzzers
  • Delete two diodes, the 22 ohms, the 100uF electrolytic, two 100K resistors, and NE556


Abstract--This simple attachment for your exercise bicycle will deliver about l00mA at a regulated voltage of your choice. For a power source, it uses a small alternator which cyclists sometimes use to power their running lights. An LM317T is used to get a nice, clean regulated output voltage for your radio, cassette machine, etc.

[Note--A discussion of these alternators and of anticipated problems in mounting them is given in the previous article. Also, there seems to be a famous story about a man who ran his TV with his exercycle; the current available from this unit is not sufficient for doing so. There is no reason why this circuitry cannot be included in the speedometer box of the previous article.]

The output voltage of these alternators as a function of drive speed is an electromagnetic anomaly. The alternators seem to be somewhere between a voltage and current source, neither quantity being a linear function of load or drive speed. On the two Schwinn units I have tried (each made in a >different country), the lamp being powered is a 1/2 amp 6-volt unit. With the bulb in place, the voltage across it does not change much after a certain speed has been reached (perhaps varying over a range of 3/4V). With the bulb disconnected, the generated output goes up very rapidly as pedaling begins; for fast pedaling, this voltage may go as high as 25V.

Whatever comes out of the alternator, the voltage regulator takes care of it for most applications. I suppose the worst that can happen, if the load becomes too much for the system, is that valleys in the half-wave rectified alternator signal will dip below the "dropout voltage" of the regulator, and a high frequency hum will make itself known. If this happens, it is an indication that the voltage doubler circuit discussed here should be used.

The output frequency of these alternators is very high, at least a few hundred cycles, even when you're loafing. It was for this reason that a full-wave rectifier was deemed unnecessary.


I chose to remove the lamp from the unit to take advantage of the higher output voltage which resulted. This lamp could be left in place if the regulated output voltage were kept below 4.5 volts. A binding post on the back of these alternators is the "hot output," while the "cold output" is taken from the mounting clamp. (Unless you take deliberate precautions to prevent it, the bike frame will be common to one side of your circuit; remember this if you elect to try a bridge rectifier or if the device you are running has a positive ground.)

[Note--All electrolytic capacitors to follow are 50-volt units.]

Simple Half-Wave Rectifier Circuit

The cold alternator output is grounded. The hot alternator output goes to the anode of a 1N4002 diode. The cathode of this diode goes through 250uF to ground (negative of the capacitor at ground). This cathode is the unregulated DC output which will feed into the regulator; the cold output is ground.

Voltage Doubler Rectifier Circuit

The full speed alternator voltage must be limited so that the output of this circuit will not exceed the ratings of the voltage regulator. This is easily accomplished; a 27 ohm 20 watt resistor is put across the alternator.

The cold side of the alternator is grounded. Its hot terminal goes to the negative end of a 250uF capacitor; the positive end of this cap goes to the cathode of a diode (1N4002). The anode of this diode is grounded. The positive end of this cap also goes to the anode of another diode (1N4002); the cathode of this diode goes through 250uF to ground (negative of this cap at ground). The top of this latter capacitor is the unregulated output. The cold unregulated output is ground.

Regulator Circuit

An LM3l7T is used; it has a piece of aluminum as a small heat sink attached to it (2 square inches, perhaps). Its input pin goes to the unregulated output of one of the previous circuits. This input is also bypassed to ground by 0.1uF (located near the regulator). The regulator's output goes through 160 ohms to its "adjust" pin, with this adjust pin going through "R" to ground. The regulator's output is the positive output of the supply, while the negative output of the supply is ground.

The following is a table of values for "R," which determines the regulated voltage:

  • 3V--220 ohms
  • 4.5V-430 ohms
  • 6V--620 ohms
  • 7.5V--820 ohms
  • 9V--1K

[Note that the case of the regulator is not grounded. Note also that the voltage doubler may be necessary for outputs of 6 volts or higher. If the doubler is used, the generator should be loaded with a 27 ohm 20 watt resistor.]

Pin Connections for the LM317T

With the mounting surface toward you and the pins facing upward, the three leads are, from left to right: "adjust," "output," "input." The case is common to the output.

Parts List

  • 1--0.1uF disk
  • 1 or 2--250uF 50V electrolytic
  • 1 or 2--1N4002 rectifier diodes 1-LM317T
  • 1--160 ohms, 1/4 watt
  • 1--"R" (see table above)
  • 1--bike generator with the lamp removed
  • 1--velocipede or non-velocipede, as the case may be
  • 1--inspirational recording of an old song by the Happiness Boys, "Jump, Fritz, I Give You Liver"


by Albert E. Yeo of Walsall, England


If you are planning to put up brackets on your wall for shelves, speakers, etc., it is nice to be sure that you won't black out the whole house when you attack the wall with tools. This article describes a very simple form of metal detector that has served me well for this purpose. It is easy enough to build, and it may appeal to the beginner as a first project.

A single general-purpose NPN transistor is used in a standard "reactance oscillator" circuit. The inductor is wound on the outside of the casing, which in this unit is a large pillbox with a diameter of about 2.5 inches (not critical). (The housing can be either plastic or cardboard, 2.5 inches in diameter, and 1.5 inches deep.) The circuit, including the 9V battery, is housed inside this circular casing. A couple of feet of flexible hookup wire hang out of the casing to act as an antenna. (This antenna can be wrapped around the box for storage.)

When the inductor of this simple device is brought close to a metal object, the oscillator frequency will shift; you can listen for this on a transistor radio tuned to the 1MHz region.


The negative terminal of the 9V battery is grounded, while its plus terminal goes through an on-off switch to the plus 9V line. This on-off switch can be omitted if you wish; the current drain of the device is about 10mA.

The emitter of the transistor (BC108, 2N2222) goes through 68 ohms to ground. The base goes through the parallel combination of 3.3K and 10uF to ground (negative of the capacitor at ground). This base also goes through 2.2K to plus 9V.

The collector goes through the inductor to plus 9V. This collector also goes through 0.01uF to the emitter. Finally, the emitter goes through 0.1uF to plus 9V. That's all there is--easy, no?

The inductance is 100 turns of something like 24-gauge enameled wire "close wound" on the outside of the casing. The antenna can come from just about anywhere in the circuit; five or six turns of the antenna wire can be wound on the outside of the inductor if you wish. You can cover the whole winding with Scotch self-adhesive tape and give it a "finished" appearance.


Install the battery and tune in the signal on the medium-wave band. Place the unit against the wall and move it around on the surface. When the pitch of the signal shifts, this means that you have located a cable or some plumbing inside the wall. Do not use your power drill just there. Incidentally, this device is handy if you drop a small screw or washer. Now get cracking, and good luck!

Parts List

  • 1--68 ohms, 1/4W, 5%
  • 1--2.2K, 1/4W, 5%
  • 1--3.3K, 1/4W, 5%
  • 1--10uF, 10V electrolytic
  • 1--0.1uF, disk (ceramic)
  • 1--0.1uF, disk or mica
  • 1--NPN transistor: BC108, 2N2222
  • 100 turns of 24-gauge wire
  • Round container, 2.5'inches diameter, 1.5 inches long (or longer)
  • 9V battery
  • Medium-wave receiver


By Albert E. Yea of Walsall, England

Abstract--This device is an extremely easy-to-build measuring instrument for identifying capacitors. Its output is of the "null" type; a braille dial is rotated to achieve a null in an audible tone.

There are as many ways of measuring capacitance as there are people wishing to do so. One popular current method is to integrate pulses from a 555 and to measure the resultant charge on the capacitor with a meter. (This is similar to integrator-type tachometers.) Another way of measuring them is to build a tuned circuit around them and measure the resultant resonant frequency (using a grid-dip oscillator).

The system to be described consists of a good old "balancing" circuit (a bridge) in which the four elements are as follows: The "unknown" capacitor is connected via two binding posts. In series with this is a set of "standard" capacitors which are selected by means of a 6-position rotary switch. A precision linear pot fitted with a braille scale forms another arm of the bridge; this goes through a calibration rheostat which is the fourth arm.

The accuracy of this instrument depends on the quality of the precision pot and, more importantly, the precision of the capacitors. Where possible, capacitors with a tolerance of 1% should be chosen as standards. However, the higher ranges use electrolytic units (1uF tantalum and 10uF tantalum), and the precision of these ranges should always be stated with caution. On the other hand, distributed capacitance (in the selector switch and the circuit wiring) may be significant at the lowest range (the standard being 100pF), and you should be aware of this possibility for error.

The six ranges are: 0 to 0.0001uF (0 to 100pF), 0 to 0.001uF (0 to 1000pF), 0 to 0.01uF, 0 to 0.1uF, 0 to 1uF, and 0 to 10uF.

Controls on the instrument are: The braille-calibrated pot from which measurements are read, the 6-position range switch, a volume control, an on-off switch, and two terminals for connecting unknown capacitors.

Some Notes on Construction

Conventional binding posts could be used as test terminals. However, whenever I build instruments for measuring components which have stiff wire leads, I adopt the following procedure:

Prepare two holes in the back of the box to accommodate the test terminals. If the box is metal, cut a hole in the back of the box which is rectangular and measures 1-3/4 inches long and 7/8 inch high. Cut a piece of plastic board (not copper-clad) large enough to be bolted to the box so as to cover the rectangular aperture. The two holes can then be drilled through the plastic board in such positions as to be insulated from the box. If a plastic or Bakelite cabinet is used for the instrument, no plastic cover plate will be necessary and the holes can be drilled straight through.

The holes for the terminals should be 4.8 millimeters in diameter (3/16 inch) and with their centers 1-1/4 inches apart. Obtain two "2-B" (3/16 inch) machine screws, washers, and matching nuts. Put these screws through the holes from the inside of the box; thread on the washers and nuts. Screw down the nuts firmly. Then, using a coping saw, cut a horizontal slot down the length of each screw--stopping the cut just short of the nuts.

You can now use a screwdriver to open out the slots just a little. You will end up with tapered slots into which the leads of the unknown capacitors can be fit easily, and yet make good contact.

The calibrated control and braille scale is mounted in the center of the main panel. The range switch, on-off switch, and volume control can be mounted off to the side on this same panel.

Circuit Operation

An oscillator made from an NE555 timer is used to drive the bridge with a square wave signal. One half of the bridge consists of the braille-calibrated pot and its associated calibration rheostat. The other side of the bridge consists of the switchable standard and the unknown capacitor. The bridge "centers" from which the condition of balance is detected are connected across the input of an LM386 audio amplifier. Because these bridge centers are not common to ground, separate supply batteries are used--one powers the 555 and the other powers the audio amplifier.


RV1 is a 50K rheostat which is used for calibration; one end of RV1 goes to minus 9V for the 555. The arm of this rheostat goes to the bottom end of RV2, the braille-calibrated precision pot (20K). The arm of RV2 goes to the arm of S2, the 6-position single-pole range switch.

Position 1 of S2 goes to one end of C1. Position 2 of S2 goes to one end of C2. Position 3 goes to one end of C3. Position 4 goes to one end of C4. Position 5 goes to the positive end of C5. Position 6 goes to the positive end of CG. The far ends of these capacitors are tied together and go to the positive test terminal (one side of the unknown). The negative test terminal (the other side of the unknown) goes to minus 9V for the 555.

Pin 1 of the 555 goes to its minus 9V. Pins 4 and 8 are tied together and go through one pole of the on-off switch (S1, DPST toggle) to the positive terminal of the 555's 9-volt battery. Between pins 1 and 8, located near the chip, is 0.1uF. Pins 2 and 6 of the 555 are tied together and go through 0.01uF to ground. Pins 2 and 6 also go through 100K to the output, pin 3. Pin 3 goes to the junction of the arm of the 6position selector switch and the arm of RV2.

An LM386 is powered from its own 9V battery. Pins 2 and 4 go to the negative terminal of this battery, while pin 6 goes through the other pole of the DP8T on-off switch (S1) to the positive terminal of this battery. Pin 6 is also bypassed to pin 4 by 250uF (negative at pin 4). Pin 7 is bypassed to pin 4 by 25uF (negative at pin 4). Pin 8 goes through 470 ohms, then through 10uF to pin 1 (positive of the capacitor at pin 1).

Pin 5, the output, is bypassed to pin 4 by 0.22uF. Pin 5 also goes to the positive end of a 100uF capacitor; the negative end of this capacitor goes through the speaker to pin 4.

Pin 4 of the LM386 and the negative side of its supply goes to the junction of RV1 and RV2 (the arm of the calibration rheostat and the bottom of the precision pot). Pin 3 of the 386 goes to the arm of RV3, a 10K volume control; the bottom of this control goes to pin 4. The top of RV3 goes through 39K to the positive test terminal (the junction of the standard capacitor and the unknown).


To calibrate the instrument, connect a good grade capacitor of known value between the test terminals. Set S2 to the appropriate range. Set RV2 to the correct reading on the braille scale, and adjust RV1 for a null. The bridge is now calibrated on all six ranges.

Miscellaneous Comments

The null on the 10uF range is not as deep as that which can be achieved on the other ranges; you may wish to omit this range. In another article, we might deal with some ways of measuring electrolytics and other large-capacitance units.

As can be seen from the circuit, this connection of the 555 is the type which generates a square wave the charging resistor is driven from the 555's output. The formula for calculating the frequency is: 1/f = 1.4 x R x C

In this circuit, R is the 100K resistor from pins 2 and 6 to pin 3, and C is 0.01uF. The calculated frequency is about 700 Hz. If R1 were made up of a fixed resistor and a rheostat in series (say, 82K in series with a 100K rheostat), the frequency could be varied to suit the user's taste. This oscillator connection is susceptible to loading. You will notice a slight drop infrequency on the higher ranges.

Parts List

Resistors (1/4 watt, 5%):

  • 1--470 ohms
  • 1--39K
  • 1--100K


(since these are circuit elements, precision is not critical):

  • 1--25f3uF, 10V
  • 1-100uF, 10V
  • 1--25uF, 10V
  • 1-10uF, 10V
  • 1--0.22uF, disk or Mylar
  • 1--0.1uF, disk
  • 1--0.01uF, disk or mica

Capacitors of the Standard

(high precision where possible):

  • C1--0.0001uF (100pF), silver mica
  • C2--0.001uF (1000pF), silver mica
  • C3--0.01uF, silver mica
  • C4--0.1uF, Mylar
  • C5--1uF, tantalum, 10V
  • C6--10uF, tantalum, 10V


  • 1--NE555
  • 1--LM386, Radio Shack 276-1731

Switches and Pots:

  • S1--DPST toggle
  • S2--Single-pole 6-position good quality (ceramic) rotary
  • RV1--50K rheostat, can be PC-mount
  • RV2--20K precision linear, suitable for fitting with braille scale
  • RV3--10K audio-taper volume control without switch


  • Suitable binding posts
  • 8-ohm PM speaker
  • 2--9V batteries
  • Suitable cabinet, preferably plastic or Bakelite

[Editor's Note--I have had astoundingly good results in building this instrument. What I mainly want a capacitor meter for is identifying units laying around in my junk box; an accuracy of 10% or so would be sufficient. However, just pulling "postage stamp mica" units out of our parts drawers (their values were labelled), Lady Luck gave me a bridge which is accurate to within 2% or so. A greater accuracy could be had by measuring your assortment on a commercial bridge; this is a lot of work and is perhaps not worth delaying the project for. Good luck with it, and thank you, Albert Yea.]


By J.C. Swail Medical Engineering Section Electrical Engineering Division National Research Council Ottawa, Canada, K1A 0R8


I have been in charge of a rather modest research program devoted to the development of vocational aids for the blind and deaf blind for the past eighteen years. Although I have never taken the time to count the number of devices my technician and I have worked on, someone came up with a total of over a hundred. Many of these are so specialized, having been designed for one very particular job, that they would be of little or no interest to others. However, a number of these do have a more general appeal. The following list gives a very brief description of some of these.

In the following list of devices you will find a number, prefixed with "A0" after the title; in some cases there are several of these numbers. These denote the drawing numbers given to the schematic diagrams for the device. The numbers are assigned by our drawing office. Thus, if you wish to have the circuit of any of the listed instruments, you need only write to the address given at the beginning of this article requesting the appropriate number.

In a very few cases we also have mechanical drawings. I have not listed these as they tend to be specifically designed around available materials and most constructors will have their own ideas and materials.

I would like to express my sincere thanks to the SKTF and its editor for agreeing to print this information. Although we can and do publish this data, it is entirely in ink print and not available to those who most need it. For those who were readers of the Braille Technical Press, many of the devices will seem like old friends.

Many of the circuits represent old technology. However, whenever we have changed from our original circuits using individual transistors to the more modern integrated circuits, we have found a marked increase in power consumption. This can be important when batteries are the source of power. Thus I have tended to leave the circuits as originally designed.

List of Descriptions and Diagram Numbers

Adapted Volt-OhmMilliammeter (A0-11-22C)

This circuit arrangement substitutes a tactile scale and an audible nulling system for the visual meter in a commercially available VOM. Our design is available, using the Simpson model 260, from SFB Products.

Meter Readers (Chopper A0-11-116B, Tone Matching A0-11-92B)

Two types of meter reader circuits have been designed. Both are intended to be connected across the terminals of an existing visual meter and are read by a tactile scale. The correct pointer position is determined either by a null (the first circuit) or by matching tones (the second circuit). The chopper is best where absolute readings are required, the tone matching where adjustment is being made for a maximum or minimum, as in tuning a transmitter.

Automotive Tune-Up Meter (A0-11-150A)

This device is based on a Heath model CM2045 small engine tune-up meter kit. A tactile scale and audio nulling circuit enable the blind mechanic to read the meter. The device measures voltages to 20 volts, resistance to 100K, two ranges of tachometer to 15,000 RPM, and dwell. Used on 1-4 cylinder engines, either 2 or 4 cycle.

Antenna Direction Indicator (A0-11-117C)

A tactile compass scale and nulling audio circuit enable a blind operator to accurately read antenna position using either the Cornell Dubilier model CD44 or HAM-M rotators.

Impedance Bridge (A0-11-154A, or A0-11-12C)

These circuits each give details of audible and tactile resistance-capacitance-inductance bridges. The first is a conversion of the Heath model IB2A kit, and the second a complete circuit for a scratch-built unit.

Light Probes (A0-11-93B, A0-11-114B)

Two versions of light probes are diagrammed. The first incorporates a light source, making it useful for detecting print, while the second is a passive device depending on ambient light.

Tactile Digital Display (A0-11-141C, A0-11-144C, A0-11-145C, A0-11-146C)

Also described on page 11 of QST, March 1982. This system is designed to read out the digital displays on most modern amateur transceivers, as well as frequency counters and voltmeters. The system employs a combination of touch and sound to read out the information.

Braille Calculator (A0-11-136C and A0-11-146C)

This is a similar digital read-out system to the above, but shown adapted to commercially available pocket calculators.

Line Voltage Indicator (A0-11-120B)

This device compares the line voltage to a reference. If the line varies more than a few volts from its nominal level, an audio oscillator is activated, alerting the user.

A very useful device where an adjustable line transformer is available to correct the voltage.

Line and Load Tester (A0-11-119B)

A very simple device which yields an audible indication of the presence or absence of line voltage, as well as indicating the presence of a load. Useful in testing household circuits and appliances.

Simple Sound Beacon (A0-11-104B)

A repetitive sound beacon which gives a short beep every five seconds or so. Useful for swimmers, or for relocating the lawn mower, etc.

IBM Punched Card Reader (A0-13-35B; also mechanical drawings available)

A line of twelve photo cells read the information on the card, column by column, as a carriage is moved along it. Solenoid operated pins tactually present the data to the user. The column number is read from a raised scale on the carriage. This unit has also been arranged to read out in synthetic speech.

Auditory Thermometer (A0-11-74B)

This unit contains the same type of nulling circuit and raised scale as in the meter reader. A temperature probe is arranged in a resistance bridge. Almost any scale may be devised; the one shown in our diagram is for a darkroom thermometer.

Relative Power Reader (A0-11-100B)

A simple tuning aid for the radio amateur. The highest frequency tone indicates maximum power output.

Program Level Meter (A0-11-78B)

An accurate program or recording level meter which gives a low frequency tone when level exceeds preset value. Successively higher pitched tones indicate amount that peaks may have exceeded this level. Uses include high quality tape recording and program monitoring in broadcast and tape duplicating studios.

Machinist's Level (A0-11-89A)

A very sensitive level using a commercially available gravity sensor. A tone indicates when surface is not level.

Continuity Tester (A0-11-88A)

A simple circuit for checking circuit continuity. Pitch of tone gives an idea of resistance. Can be used for checking fuses, switches, diodes, and polarity of electrolytic capacitors.

Liquid Level Indicator (A0-11-137A)

A circuit somewhat similar to the continuity checker is built into a small cylindrical case. A pair of prongs extends from this into the container which is being filled with liquid. When liquid comes up to the tips of the prongs, continuity is formed through the liquid, and a warning tone sounds.

Tape Tone Indexer (A0-11-149B)

This device is connected in the microphone line of a tape recorder. A push-button inserts low frequency tones on the tape. These are inaudible at normal playing speed, but appear as high pitched beeps at fast forward or rewind. Used for marking pages or chapters on tape.

Digital Synthetic Speech Interface (M-11-138B)

A system is shown for adapting a frequency counter to have synthetic speech read-out. May be adapted to most instruments having digital displays.

Synthetic Speech Computer Terminal Attachment (A0-13-40B through A0-13-57C, inclusive)

A spelled speech unit designed to be connected in parallel with a video display terminal. Has an 8K buffer, variable speech rate, and search facilities. For use where full word speech is not required. Useful for programming, charge card verification, etc.

Deaf-Blind Signalling Device (A0-11-153A)

Built around the Radio Shack Mobile Alert. A vibrator is added to the belt-worn receiver, and the transmitter is connected to the doorbell or any other device which the deaf-blind user may choose. When a signal activates the transmitter, the receiver vibrates until switched off, thus alerting the user.

Deaf-Blind Code Receiver (Am-11-151B)

This device, when connected to a receiver, divides the usual 800Hz tone used in code reception to 200Hz. This, in turn, drives a loudspeaker converted to act as a vibrator. This --allows a deaf-blind person to feel the incoming Morse Code signal.


Dr. Lawrence A. Chang, a blind mathematics professor at Lawrence Livermore National Laboratory, has written a book called "Larry's Speakeasy." A print pamphlet, it describes procedures by which mathematics can be "spoken."

It covers subjects such as trigonometry, analytic geometry and the calculus, topology and abstract spaces, business mathematics, diagrams and graphs.

It is of particular interest to the following users: Those who read mathematics orally and blind people who use readers, those interested in spoken output from computers, and transcribers of mathematics.

Contact the LLNL Office of Equal Opportunity, P.O. Box 808, Livermore, CA 94550, telephone (415) 422-9541.

Ziegler Items

  • In the December 1983 issue is the following announcement in "Special Notices."
  • "Compuserve Kit. Tandy Way, 8989 Peppermill Court, Arampa, IL 33614, offers a cassette recording of the 'Start-up Kit' for the Compuserve Information System. The cassette includes hints on getting around the system. Write for cost information."
  • The following announcement appeared in "Special Notices" of the Mid-Winter 1984 Ziegler.
  • "Cassette Course on Computers. VOICE'S, a free tape-lending library, Box 837, Bethel, ME 04217, will lend you Keys to a Better Life, a cassette covering the following subjects: Use of home computers, availability of large print computers, a glossary and a bibliography, and other reference" materials. This cassette may be borrowed for thirty days or purchased for $5." [They say to establish your eligibility for this service, mention the Ziegler.]
  • Also in this Mid-Winter issue is an 8-page article by Harvey Lauer (head of a project at the Hines VA Center) entitled "How to Select a Computer: What Kind of Computer Should I Buy?"
  • A new newsletter:
    • "I am coordinating an informal tape newsletter for blind microcomputer users. Cassettes will be 90 minutes in length and will be issued quarterly. Subscribers will be asked to provide their own cassettes, but the newsletter will be free of charge. Its purpose is to share ideas, programs and friendships.
    • "I intend to make the tapes informal so as to reflect the atmosphere of a users' group meeting. Currently, enough material has been collected to fill one tape, most of which is related to the Votrax Personal Speech System. If there is enough interest, the newsletter will be 'Votrax-specific.' It will not, however, be specifically for users of one brand of microcomputer.
    • "I encourage contributions and suggestions in typing, on tape, or in braille (no handwriting, please). For the first issue, send a C-90 cassette of reasonable quality to: Debee Norling, P.O. Box 5702, Berkeley, CA 94705."