SKTF -- Fall 1985

The Smith-Kettlewell Technical File

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

William Gerrey, Editor

Issue: [current-page:title]

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 Details regarding products, suppliers, and other contact information are original and may be outdated.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


From Paper to Project, Part III

The National LM3909 Flasher Chip

"Marklites" - A Burglary Deterrent System

The Editor's Anniversary

Merry Christmas

Part III


In this installment, we will consider building projects using 4-pin DIP (dual in-line package) chips. Specifically, a test amplifier will be described first; then, a test oscillator will be shown to accompany the amplifier on the same board. The resultant combination, encapsulated in a single box, will be a versatile instrument. Further, however, this same circuit board will be duplicated twice more, since the
combination of these two circuits is the heart of two more instruments, the resistance and capacitance bridges to follow. Therefore, this combination of circuits will provide ample practice, since three sets should be reproduced.

[It's only fair to tell you that a test amplifier (Radio Shack No. 277-1008) is available. Oscillators called "signal injectors" can be gotten commercially. "Then why build these?" you fairly ask. Because, not only will the layouts described be necessarily reproduced in the audible resistance and capacitance meters mentioned, but these are very common circuits which have occurred over and over again in many devices for the blind--barring minor variations that stem from whim of the designers. The 555 oscillator has become a ubiquitous component in everything from light probes to doorbells; the LM386 amplifier has found its way into items ranging from pocket radios to the ol' TSI "Speech Plus" calculator. You will be building variations on these circuits in your sleep, so get your practice now.]

Know Your Parts

The chips referred to come in the "dual in-line package" (DIP), which means that there are two parallel lines, or columns, of pins. In the case of these chips, the Signetics 555 "timer" and the LM386 amplifier, there are two columns of four pins (eight in all). The result is that they look like little square tiles with four legs extending downward from opposite sides.

No matter which size of DIP we're talking about, the pins are numbered in accordance with a simple system. At one end of each chip is a mark; this mark often occurs between "pin 1" and the highest-number pin. For example, a mark at the end of a 40pin chip denotes that pins "1" and "40" are at this end. From here on, the technician knows that the pins are counted in a clockwise fashion (viewed from the bottom side). (The style and exact position of the mark varies between manufacturers, but the counting system is rigidly adhered to.)

Counting pins on a chip is done from the bottom view, and it is always done with the "keyway" (the mark) positioned away from you. Using the 40-pin chip as an example: The chip is made to
lie on its back (the pins facing upward and arranged in two columns), with its "mark" on the case being between the top left pin and the top right one. The upper right-hand pin is "pin 1," while the upper left-hand pin is "pin 40." Counting down the right-hand column, the pins are "1" through "20." Counting the left column from the bottom upward (hence, clockwise), the pins are "21" through "40." If the chip has eight pins, the columns number "1" through "4" down the right side, and "5" through "8" up the left side. (Geezers like me will see this system as being familiar; "tubes" and "plugs" are always counted in a clockwise direction from their "keyway.")

The "mark" is often a notch at the upper end of the DIP IC; it may, however, be a very small indentation immediately above
pin "1" (on the top side of the package). If it is an
indentation, it may only be evident when explored with a braille stylus. (There will be other indentations in the top of the package, usually due to "releasing" the capsule from its mold.) The notch you're looking for is in the edge of the chip, centered at one end. The "dot," as you might call it, is an impression in the capsule, just over the "pin 1" corner.

Chip sockets are also marked. The variations here are frustrating; these marks may only be evident to the socket's inventor. Two common systems are: There may be a pronounced notch at one end between "pin 1" and the highest pin. Often, the corner to be considered "pin 1" is rounded off. You can take consolation in the fact that if the socket is turned around wrong, the project will still work, as long as you plug the chip in correctly. Some consideration should be given to the "next guy," since he may very well plug a replacement into a socket whose mark directs him falsely -- your reward for misdirection will be in the great beyond, I hope.

For the previous project, there were few variations on the style of parts you were likely to get. This is no longer true: capacitors are available in kinds that lie down, and kinds that stand up on end, and the size of capacitors depends on their manufacturer. Variations in parts' sizes are disappearing rapidly; in a couple of years, this paragraph will be obsolete.) Capacitors of the "radial-lead" type are meant to stand on end; capacitors of the "axial-lead type" are meant to lie down against the board -- much as resistors do. In this description, the electrolytic capacitors used are of the "axial lead" type, while the mylar units (0.1uF and 0.01uF) seem to be standing on end these days (they have "radial leads"). The ones you may buy
might be different. If your electrolytic units turn out to have radial leads, it's no big deal; brief mention will be given in the description which will direct your modified placement.

Many of the capacitors in these projects are "electrolytic"; they are therefore polarized, and require that you know which
lead is "plus" and which is "minus." Capacitors having axial leads (one coming out either end) are often identifiable by touch; their positive lead emerges from a rubber insulator, and there is often a ridge around this end which indicates that the outer can has been crimped around the rubber. On the other hand, capacitors with radial leads (both coming out one end) cannot be identified tactually, except in cases where the manufacturer has made the "plus" lead longer than the "minus" one. (Making the positive lead longer than the negative one is done to aid in production. With this system, people, or machines, do not have to look for the tiny visual markings that identify the polarity.)

You, dear student, having built a continuity tester, possess the key by which the polarity of electrolytic capacitors can be ferreted out, regardless of their dubious tactile markings.
(This system only works for capacitors whose voltage rating
exceeds that of the battery in your tester; i.e., units which are rated at 10 volts or higher.)

If you connect your tester across an electrolytic capacitor in reverse (with the positive tester lead on the minus lead of the capacitor), you will hear a descending pitch in the tone that
"levels off," as you might say--its descent settles down to produce a steady tone, or even one which begins to rise again. When this happens, turn your tester leads around and check the other polarity; the tone should descend to a low pitch, whereupon it will become intermittent.

This latter condition, an intermittent buzz, is what you're after; when you hear an intermittent buzz, the positive tester lead is on the positive capacitor lead. (Whenever you have put a charge on an electrolytic capacitor in the wrong direction, turn it around and charge it up the right way, since leaving it charged in the wrong direction can puncture the oxides that form the "dielectric.")

[You should know that there are very tiny electrolytic capacitors made using tantalum instead of aluminum. These tantalum capacitors, besides having the advantage of small size, have the reputation of being very high quality. Their quality is undisputed as far as leakage is concerned; however, their internal impedance is higher than that of standard aluminum units. Therefore, in critical bypassing applications, they are not as good as their larger cousins. In this amplifier, for example, I would feel comfortable about substituting tantalum units for the 10uF and 100uF coupling capacitors, but I would be nervous about buying them for the 22uF and 220uF bypasses.]

Another new gizmo you'll have to contend with is the
amplifier's volume control. This is a resistor which has a "wiper" or feeler that allows you to select any position along its length. You will see that it has three terminals. If you connect your continuity tester to the outer ones, a steady tone will be heard, and turning the shaft of the control will make no difference. However, if the tester is placed between the center terminal (which is usually the wiper) and either of the other terminals, the pitch of the tester will vary as the control is adjusted. [With the shaft of the control toward you, and the terminals near the bottom, the left-hand terminal will probably be the "bottom" end (the counterclockwise end), the center will probably be the wiper, and the right-hand terminal will probably be the "top" end (the clockwise end).]

You should picture the volume control (a "potentiometer") as a horseshoe-shaped resistor; the "wiper" traces the horseshoe
from the bottom left end over to the bottom right end. The

bottom right end of the resistor will be given a signal from somewhere (whatever you feed it); the bottom left end will be shorted to ground. The position of the wiper will determine just how much of this signal will be fed to your amplifier.

There are occasional variations in the layout of terminals on potentiometers; a good habit to cultivate is to check them. With the tester connected in various combinations, you can make sure of the terminals' configuration. For example, if they are

in the traditional order: the whole resistance will appear between the outer terminals, and the wiper's position will only affect the tester when one test lead is moved to the middle terminal. In the "turned-down" position, the wiper should appear to be shorted to the bottom end -- this "bottom end" being the left-hand lug when the shaft is facing you and the terminals are at the bottom.

Amplifier Parts

  • 2--0.1uF mylar or disc ceramic capacitors, low voltage (for low-level audio coupling, mylar units are less noisy; with this amplifier, it won't matter much). (The oscillator needs a 0.1uF too, so buy three of them.)
  • 1--10uF electrolytic capacitor, 10 volt (anything from 5uF on up will work). (The oscillator needs 10uF too, so buy two of them.)
  • 1--22uF electrolytic capacitor, 10 volt (10uF might work; anything from 15uF to 47uF, or higher, will certainly work).
  • 1--100uF electrolytic capacitor, 10 volt (anything from 100uF on up will work; if you increase this value by much,
    however, also increase the main 220uF bypass unit).
  • 1--220uF electrolytic capacitor, 10 volt (this can be from 200uF to as high as you want).
  • 1--470 ohm resistor, 1/4 watt
  • 1--10K ohm volume control (audio taper), 1/10 watt
  • 1--LM386 (or LM386N) amplifier chip

Oscillator Parts

  • 1--0.01uF mylar capacitor, low voltage (disc ceramic will do, if you don't mind a little frequency instability).
  • 1--0.1uF mylar or disc ceramic capacitor, low voltage (disc ceramic would be the preferred choice here, since it will provide the lowest impedance bypassing, although I used mylar).
  • 1--10uF electrolytic capacitor, 10 volts or higher (not used in the bridges of the next project).
  • 1--100K ohm resistor, 1/4 watt
  • 1--100 ohm resistor, 1/2 watt
  • 1--Signetics NE555 timer, or a host of equivalents

Miscellaneous Parts

  • 2--8-pin DIP sockets (high-priced ones will break less
  • 2--open-circuit mini-phone jacks (1/8 inch)
  • 1--piece of Vector Board (perforated board) such as Radio Shack No. 276-1396, or Vector No. 169P47
  • 1--2-1/4 inch Speaker, Radio Shack No. 40-246
  • 1--cabinet, Radio Shack No. 270-627 (6-1/4 by 3-3/4 by 2 inches)
  • 1--hat full of 9V battery clips
  • 1--switch (if just the amplifier/oscillator, SPST; if for the bridges, DPST)
  • 1--small spool of 22-gauge bare copper wire, no enamel or other insulation
  • 1--bundle of solid 24- or 26-gauge hookup wire with plastic insulation
  • 1--bundle of stranded 22- or 24-gauge hookup wire
  • 1--spool of rosin-core solder, approximately 0.03 inches in diameter
  • 4--3/4 inch 4-40 machine screws and nuts
  • 4--1/4 inch spacers, Kulka-Smith 4006 or Small Parts RSN-4/4 ($2.80 per 100)
  • 1--tube or small bottle of Goodyear Pliobond

Before we leave this section, do some experiments on the parts with your continuity tester. (Leave the chips out of these experiments, please. Although these chips are particularly

durable and would probably survive, many chips are not, and you should consider shocking them with testers to be a dangerous practice.) First, try the polarity identification tricks mentioned above on the capacitors. Note also that, as the capacitance value goes up, the "charging time" increases proportionally.

Now, assemble your resistors in a bundle and purposefully mix them up. You will notice that, as you test them, they cause two distinctly different pitches to be emitted (the third resistor, the 100-ohm unit of one-half watt size will not be so different from its 470-ohm neighbor -- but then, it is a different physical size.) The higher the resistance, the lower will be the tone; the 100K unit will cause the tester to emit a low buzzzzzz. Thus, without seeing the color codes, and without even measuring their ohmic values precisely, you can identify them with certainty -- using a process of elimination.

Just for fun, connect the tester across the speaker. This beep will be virtually the same pitch as that gotten by shorting the test leads together. Too bad; if you had a 10-ohm resistor, a 22-ohm resistor, and a shorted tubular ceramic capacitor, you would not be able to tell the difference with the continuity tester. When your resistance bridge is finished, you will be able to identify these.

In testing the smaller capacitors, the 0.1's and the 0.01, your tester may emit one rather inaudible click -- that's about all. Therefore, you'll have to identify these by size. When you finish your capacitance bridge, you can measure such items.

Presenting the Circuits
Irrespective of Layout

Once again, let me remind you that there are two ways a circuit must be viewed: The first is with an eye to circuit sense; the schematic diagram is a document by which the electronic doings of a circuit can be communicated and understood. The other aspect is viewing the circuit elements with an eye to putting the physical items together. The former was the designer's choice; the latter is the builder's opportunity to: make it compact, make it large with easy-to-use controls, or to keep its innards spacious so as to be easy to build.

The circuit diagram, whether a drawing or a verbal

description, is not arranged to present layout. By the same token, my descriptions of where to put parts and which leads to solder together (see the next section, "Step-by-Step
Construction") tell the electronics technician nothing about how the circuit works. Yet the schematic descriptions, or equivalent drawings, in this section are themselves incomplete because of the nature of integrated circuits.

Integrated circuits (IC's) are "black boxes" of pre-
assembled electronic components, and their innards are complicated. Two more levels of schematic presentation could be given:

One approach could show our outside components around a block diagram of the chips (showing the 555 full of flip-flops and comparators, and the LM386 as an operational amplifier with a built-in feedback network). The absurd extreme would be to draw our few parts around the incredible mass of actual "integrated circuit," wherein networks of junctions (making transistors and diodes of certain characteristics and biasings) become resistors and capacitors. This last approach is unusable, since any "circuit sense" has long-since been buried in a computer design model.

Using block diagrams is not a bad idea, but this can lead to confusion about which parts are external and which elements are "integrated" into the chip. The standard approach is to make you read the "theory of operation" of the chips elsewhere, and to
give the technician just enough information to permit educated guesses as to what the circuit is doing. Well, nothing's

The theory behind the LM386 can be found in "The LM386 HighGain Audio Amplifier" (SKTF, Fall 1980), and the 555 is best described in "Continuity Tester Circuits -- Old and New" (SKTF, Fall 1982).

There are statements that should accompany the circuit
diagrams of any well-documented designs. These are listed below:

The series-connected 470 ohm resistor and 10uF capacitor between pins 1 and 8 of the LM386 amplifier chip are, as shown by the manufacturer, a positive feedback network (counteracting the built-in negative feedback); this increases the amplifier gain. This gain will be at maximum, a voltage gain of 200, if the resistor is left out and the capacitor is connected directly between pins 1 and 8.

The 0.1uF capacitor between pins 4 and 5 of the 386 was found to be necessary to suppress high-frequency oscillations.

The NE555 is connected in an unusual configuration, with its output driving the "Trigger" and "Threshold" terminals, so as to obtain a squarewave, rather than a waveform other than that of
50% duty cycle.

Amplifier Circuit

The ground bus goes to the negative side of a 9V battery; the positive of this battery goes through an SPST switch to the VCC bus. [V, with the subscript CC, (VCC as presented here) is an old term that stems from the days of discrete transistors; VC was the "collector voltage" at any given time, and VCC was the supply voltage.] The VCC line is bypassed to ground by 220uF (with the negative side of this capacitor at ground).

Pin 6 of the LM386 goes to the VCC line. Pin 4 is grounded. Also grounded is pin 2, the inverting input. Pin 7, a decoupling point, is bypassed to ground by 22uF (negative side at ground).

Pin 3, the non-inverting input, goes through 0.1uF to the arm of a 10K volume control. The bottom of this control is grounded, while its top end goes to the "hot" terminal (the tip contact) of the input jack. The "cold" side of the input jack (the sleeve contact) is grounded.

Pin 8 goes to the negative end of a 10uF capacitor; the positive end goes through 470 ohms to pin 1. (For more amplifier gain, omit this resistor.)

To suppress oscillations, the output, pin 5, is bypassed to pin 4 by 0.1uF, located close to the chip. Pin 5 also goes to
the positive end of a 100uF capacitor, with the negative end
going through the speaker to ground.

Oscillator Circuit

Pin 1 of an NE555 "timer" (or 555 of any make) is grounded. Pin 8 goes to the VCC line. Between pins 1 and 8, located close to the chip, is a 0.1uF bypass capacitor. Also going to VCC is pin 4, the "Enable" pin.

Pin 2, the "Trigger," and pin 6, the "Threshold," are tied together and go through the 0.01uF timing capacitor to ground. Pins 2 and 6 also go through 100K to the output, pin 3.

Pin 3, the output, goes through 100 ohms (1/2 watt) to the positive side of a 10uF coupling capacitor; the negative side of this cap goes to the hot side of the output jack (the tip contact). The cold side of the output jack is grounded.

Step-by-Step Construction

Because you come from widely varying degrees of experience, no one level of discussion can satisfy everyone. Because I am long-winded, I tend to present lots of detail; some of you may rightfully resent my fastidious attention to small detail. Therefore, I am presenting this section in two levels:
statements which just describe the steps to be taken will be printed with standard margins; paragraphs of more parenthetical nature will be indented two spaces. Thus, those of you who have the reaction, "Yes, yes! I know all that," can look down the left margin and easily find the next construction step.

Preparing the Board

Prepare a piece of Vector Board which is about 4-1/2 inches long and about 1-5/8 inches wide. Specifically, using techniques described earlier in this series, your width of board should be gotten by breaking it off at the
16th row of holes (thus giving you 15 usable holes across its width), and break this strip off at the 45th hole along its length (thus giving you 44 usable holes in this direction).

Since I have picked the size of board, you have an opportunity to cheat here. Before you load it with components, you can use it to drill mounting holes in one of the side panels of the cabinet; it's much easier to do this now than later. Using two or three strips of masking tape across its width, tape the board to the outside of the box, centering it roughly by hand in the middle of one long side. Count three holes along and three holes in from each corner -- three holes in both directions -- and drill through both the board and the box, using these Vector Board holes as pilot holes. Do this with a No. 40 drill bit, which will give you clearance for No. 4 machine screws.

[I never get to do this -- drill the mounting holes before the board is loaded with obstructions. It would tickle you to know why. I never know the size of board I'll end up with until
the project is finished
. I build something, and if there is any extra room at one end, I break it off as the final step. Then, I stuff the finished board into my pocket and head for the parts store to see if I can find a container that will accommodate it. Planning ahead is an admirable habit, but too much of this just delays the project.]

We must now adorn the edges of this masterpiece with "bus bars" -- lengths of bare copper wire which will become "Ground" and "VCC." (There are neat tricks regarding bus wire in the last section, "Hints and Kinks"; don't miss them.) The idea is to build a pair of "railroad tracks" along the edges of the wiring side of the board (not the component side). Since there is nothing else on the board right now, installation of these buses will determine which side is the "wiring side" and which is the "component side." In other words, once installed, the components will be laid to rest on the side from which the railroad tracks cannot be seen.

To install a bus, a right-angle bend is put in the end of the wire -- about 3/8 of an inch from the end -- whereupon, it is put through one of the very edge holes near one end. Once through, crimp it over against the component side to hold this
end in place. Put the other end of the bus wire through its intended hole -- on the same edge and very near the other end of the board; pull the bus bar taut, crimp it over as well, and cut off the excess.

For this project, the ends of the bus wires should come through the board at the 5th hole from each end -- in the rows of holes that are closest to the long sides. The result will be "rails" on the edges of the board which are 36 holes long, and which leave 4 clear holes at either end.

Wiring the Amplifier

Seen from the bottom view (so we can tell more easily where the pins are), let pins 1 through 4 appear in the 10th column from the right end; and let pins 4, 3, 2 and 1 be in the 7th,
8th, 9th, and 10th holes from the edge nearest you, respectively. This would put pins 5, 6, 7 and 8 in the 13th column, being in the 7th through 10th holes away from the long edge nearest you. (This "long edge" near you will henceforth be called "the ground edge," while the top row will be called "the VCC edge.")

It was arbitrarily decided to orient the amplifier with its input side nearest one end of the board. Why? Because, in
looking through the circuit, it is evident that electrolytic capacitors grow like tumors off the other side of the socket (counting the VCC bypass, which, by good practice, should be near the chip). If we started with pins 5 through 8 nearest the end of the board, we would have to plan carefully in order to leave room at this end for those electrolytics. Therefore, putting the input side of the socket, pins 1 through 4, nearest the end, we can let the other side of the socket take up as much board acreage as it will, without concern.

We still, however, have to plan a little room on the input side for input coupling. Therefore, the socket will be mounted an inch in from one end. Since my bypass capacitors are rather long, I decided to put the socket nearer the VCC bus, rather than centering it.

With the socket lying on its back, and the board resting atop its pins, bend each row of pins straight out to the side;
i.e., bend pins 1 through 4 to the right, and bend pins 5 through
8 to the left (this is with the socket positioned so that 1 through 4 are in the right-hand column). If necessary, use a screwdriver so as to flatten them down against the wiring side as much as possible. There are two reasons for wanting them horizontal: this will keep the socket from falling out of its holes, and, with the pins lying down like this, you will have the maximum surface area to solder to.

Installing Power Connections

It is usually sensible to install power connections and jumper wires first, since these are small items and logically fit up close to the socket. Insulated solid wire, as opposed to stranded wire, is the most appropriate for on-board jumpers. Remember, all such connecting wires are to be inserted from the component side; treat these jumpers as components.

Let us first connect pin 6 to the VCC bus. This jumper will be installed in the 14th column -- counting from the end the
socket is on. In other words, this jumper is as close to the
socket as we can get it, and it reaches from the 8th hole to the 15th hole in its column -- counting from the ground bus's edge.

Strip insulation off the end of your 24- or 26-gauge wire to expose about 5/16 of an inch. Right up next to the socket (in
the 14th column from the end of the board), poke the end of this wire into the hole adjacent to pin 6 of the socket (which will be in the 8th hole up from the "ground edge"). Make sure that insertion is far enough so that the insulation is up against the board; make sure that the insulation doesn't come through the hole to interfere with soldering.

Once through up to the hilt, bend the 1/4-inch protruding end down in the direction of the socket pin; now wrap the insulated portion around the board a couple of times to hold it in place. Put the board in a clamp and carefully steer this bare end so that it is exactly resting on top of the pin. Note that it is too long, thus reaching over pin 6 and pointing at pin 3. That's good, because its extra length is how you'll be able to find it with the soldering iron (see the section called "Landmarking," in "Soldering, Part II," SKTF, Winter 1981).

The previous paragraph contains an important principle; you must devise ways by which the connection you're making is easy to find with a hot iron. I have practiced so much that all I
usually need is to have the lead I'm working on pro trude conspicuously into the space between the rows of DIP pins. However, you may wish to insert a sewing needle in a nearby hole, or to actually install another wire nearby.

You should give serious consideration to the solder guides -- "The JA3TBW Solder Guide," SKTF, Spring 1983, and "The G3JYT Solder Guide," SKTF, Fall 1984. In using these, however, you

must be aware of the possibility of moving the wire off the pin with your enthusiasm for holding the guide in place.

Whichever system you use to find the connection, you can freely rest the soldering iron on the board between pins and press it up against the side of the connection. Resting the iron on the board is okay, because these boards are made of wonderfully inert material nowadays; they really don't burn or contaminate the tip of your iron (unless you have a very cheap phenolic one).

With the wire lying flat down over the socket pin, solder it there. Then, cut off the spare 1/8 inch protruding end (trying
not to cut the pin); cutting this short after it is soldered will
make the next protruding wire unique and easy to find.

Unwrap the insulated length of wire from around the board and cut this short so that it just extends, by perhaps 3/8 of an inch, past the "VCC edge" of the board. Strip off 3/8 of an inch of insulation and plug it in beside the bus bar, in the 14th hole from the input end.

Try to arrange this newly inserted lead so as to come past the bus bar on the inside; i.e., you want the jumper to force the bus wire in the direction of the edge of the board. Once
through, bend the jumper's protruding end away from you, so that it encircles the bus wire and extends off the edge of the board. Following the edge of the board with your soldering iron, you will easily find the protruding end of the jumper. Solder it, and clip off its excess, just against the bus bar.

Connecting pin 4 to ground is done similarly: a solid jumper is installed in the 9th column from the input end -- up close to the socket; its lower end goes in the bottom hole of this column (coming through right up against the ground bus), and its upper end is in the 7th from the ground edge (immediately adjacent to pin 4).

As before, I recommend making one connection before cutting the wire to length; this will help you "anchor" the unconnected
end in a durable fashion (wrapping it around the board being a favorite way of mine). I always do the socket end first: I strip 5/16th of an inch bare, insert it adjacent to the pin, bend it over, and wrap the free length of wire around the board to hold everything still. Once stable, make sure that the 1/4 inch segment that exposes itself on the wiring side is bent firmly down against the board, and that it is straight -- not pointing at an angle, where it could bridge to the pin next door.

Now you can cut, measure, and strip the free end to length; the fact that one end is securely soldered means that no clever tricks need be employed to hold things still. Unwrap the free
end from around the board and measure off the length that can reach from the socket down beyond the ground edge of the board by 3/8 of an inch. Strip 3/8 of an inch bare, insert this in the edge hole -- 9th from the input end -- and fold it over the bus bar where it is to be soldered. After it has been soldered, clip off any extra pigtail right up close to the bus bar.

Before making any of these connections, you should (as discussed in Part I) manipulate the wire or component from the "component side" of the board to disturb the point of contact on the wiring side; then rebend the connection firmly before letting go on the component side. This assures that the leads to be connected "spring" into place, rather than hover around and touch each other every once in a while. For example, in soldering wire to a socket pin, before you reach for the soldering tools, you should roll the wire with your thumbnail toward the socket on the component side -- this will cause the protruding end to lift up and leave the socket pin -- and while holding things in this position, rebend the tip of the wire down against the socket pin.

The lengths "5/16" and "3/8" for strip ping the jumper wire are, of course, approximate. They are chosen to assure that you will have extra-length "tails" to find, thus aiding you in soldering. The critical parameters are: Can you find it? Does
it reach too far and short something out across the way? Just use your own good judgment and enjoy experimenting. Don't forget to trim these extra tails off after the connection has

been made.

You might wisely ask, "Aren't there more things to be soldered to pin 4? Why can't I wait and solder several items at once?" You can. However, I have found that I make fewer bridged connections when I don't have loose floppy things nearby to go astray while I'm not looking. Soldering each new lead holds it in place and out of the way of other connections. Also, by making two or three separate solder connections on a pin, you assure that all items are inter connected; there is less possibility of soldering the components' leads together without including the pin underneath, for example.

Installing the Input Capacitor

Now, the circuit diagram next mentions pin 2 as having to go to ground. We could plod along and do this next; however, doing so will tend to obscure
pin 3 and make it hard to solder when its time comes. You will

get so that falling into this trap won't be cause for tears. For now however, while we think of it, let's install the coupling capacitor on pin 3 first, thus taking these connections "in order of pins," not "in order of circuit presentation."

My capacitor happens to be a mylar type with "radial leads" (the leads coming out the bottom, parallel to one another). My unit comfortably fits with its leads in the 4th and 7th holes

from the end of the board, both being in the row adjacent to pin 3 (8 up from the ground edge). (These leads could have fit in the 5th and 8th, or even the 6th and 9th from the end, but there's plenty of room at this end, and I wanted to leave room for my fat fingers to put in the pin 2 jumper that comes next.)

Once the capacitor's leads have been inserted in the designated holes, pull on them in order to bring its body snugly against the component side. Then, bend its leads out to the sides to keep it from backing out of its holes. The lead intended for pin 3 can now be cut to about 1/2 inch, then laid over onto the pin. Before soldering this lead, crimp the other one over the end of the board, or hold it down with an alligator clip -- some technique to keep everything stable. Solder the pin 3 connection and cut off the excess tail.

The important aspects of placement are: The lead going to pin 3 should definitely be in the 8th row up -- adjacent to pin
3; if it is not, it will have to reach pin 3 from an odd angle. This sloppy practice will earn you bridges in Hades (lots of bridged connections). Leaving space between the capacitor and the socket is a luxury you can afford, so you might as well create a space. The other capacitor lead can go anywhere -- although rules 4 and 5 of the Wiring Game do say something about not scrunching components or their leads where they do not naturally fit. In general, one should be more critical about which holes the leads of radial-type components occupy -- more so than with axial leads -- so as to prevent kinking and lead breakage under the bodies of the radial units.

The Forgotten Jumper

Now that pin 3 has been soldered, we can ground pin 2. This can be done in two ways: We could use a little hairpin loop to tie pins 2 and 4 together; this is tricky, since it is hard to keep the little hairpin adequately stable
(see the "Hints and Kinks" section). For now, let us jumper pin 2 to ground in exactly the same way the power connections were made.

This jumper will be put in the 8th column from the input end of the board. Its upper end will go in to the 9th hole up from ground, and its lower end will be plugged in at the ground edge,
right beside the pin 4 ground connection.

Note that, because this jumper is two holes to the right of the socket, the stripped end of this jumper will have to be a little longer in order to reach the center of the socket. This time, 3/8 or 7/16 of an inch have to be made bare for the pin 2 connection. Everything else, however, is done in the usual way, and should go without saying. This "every thing else" would include: soldering one end with the excess wrapped around the
board to stabilize it, then measuring a length which is slightly longer than necessary, stripping and plugging in the second end, leaving a tail that will help you find it with the iron.

I would like to re-articulate a common boo-boo: In
inserting these wires, minimize or eliminate the exposure of bare wire on top of the component side -- you don't want a jumper to short out against your input capacitor some day. On the other hand, make sure, using your fingernail to investigate, that no insulation of a jumper comes through a hole in the board to the wiring side, where it can easily contaminate the tip of your soldering iron.

The Pin 7 Bypass

My axial-lead 22uF capacitor is 3/4 of an inch long; this means that it spans the full eight holes, from the 1st to the 9th, from ground up to pin 7. I put it in the 16th column (three holes to the left of the socket). Once lying comfortably against the component side, I cut its upper lead so as to protrude 1/2 inch, and I let the full length of its lower lead reach down off the lower edge of the board.

Before you lean its upper lead over against pin 7 and solder it there, check the polarity of this unit with your continuity tester to see that it has been installed the right way round (see "Know Your Parts"). The positive lead must be the one adjacent to pin 7.

If your 22uF unit has radial leads, a good move is to stand it up right beside the socket with its leads in this same 16th column. The distance between its leads should be governed by how they leave the capacitor body, but the upper one can go into the 9th hole up from ground, where it can reach straight over to pin 7. Its lower lead can then be run along the wiring surface of
the board to cross the ground bus. (If it is not long enough to reach ground directly, use a jumper in column 15, just to its right, to ground this end.)

Output Capacitors

The processes to be accomplished are: A 0.1uF bypass capacitor is positioned immediately below the chip; it can go in the 5th row up, and with its legs in the same columns as contain the chip pins (10 and 13 from the input end). One by one, this bypass's leads are bent upward, where they can be soldered to pins 4 and 5 as intended. Then, the output capacitor goes to the left of the 22uF bypass; its positive lead reaches to the right, where it contacts either the 0.1uF bypass's lead or the number 5 pin. Wherever the negative lead ends up (to be discussed), cut this lead so that it protrudes about 1/2 inch, then bend it in a direction of vacant space; it awaits a speaker lead.

If your capacitor is short enough (assuming, for the moment, that it has axial leads), you could put it in the 20th column with its positive lead in the 7th hole up, and its negative lead up near VCC, perhaps in the 14th up from ground. My antique capacitor is 7/8 of an inch long, so I had to make the compromise described as follows:

Since my capacitor is too long, its placement is not ideal -- according to the rules of the Wiring Game. I chose to put its negative lead in the hole next to the VCC edge (14th up from ground). Its positive end then had to go into the 5th hole up from ground; this is two holes below the 7th row, where it could have made a straight run over to pin 5.

My solution was to run its positive lead straight to the right for a while (in line with the 5th row); then, immediately below pin 5, I put a right- angle bend in the lead to carry it up to pin 5. If yours is short enough to place in rows 7 and 14, please do so.

By the time my lead got to pin 5, its position was very unstable. Therefore, I used the 0.1uF capacitor's lead to "staple" my electrolytic's lead in place. In other words:

I put my 0.1uF capacitor in the same place as described above and soldered its pin 4 lead. Then, I cut the left-hand lead of the bypass unit to protrude 1/4 of an inch, whereupon I crimped this down over the right-angle bend in my 100uF unit, thus sort of clamping everything in place. I took great care to see that my output capacitor's lead was long enough to cover pin 5, but stopped short of pin 6; then I soldered everything in a hurry before the earthquake could come along and mess it up. (I soldered both pin 5 and the junction of the two capacitors.)

It is worthy of note that, if I had had a radial-lead 22uF unit hooked up as I described earlier, I could not have run the output lead in this way without crossing the ground leg of the 22uF by pass. The choice then would have been to run the output capacitor's positive lead along the component side of the board so as to keep these items separated.

If you are lucky enough to have an output capacitor with radial leads, you can position this below the socket. Place the radial-lead output capacitor with its positive lead in the column shared by pin 5 (the 13th column) -- put its negative lead where it comfortably fits -- and solder the positive lead in the same manner as that of the bypass above it. Remember, keep these things just short of reaching too far; you don't want to hook onto pins 5 and 6 together, for example.

Supply Bypass

This one is easy. Somewhere beyond all this mess (I used the 24th column from the input end, which is also the 20th column from the other end), I inserted the negative end of my 220uF main bypass capacitor in the ground edge, and its positive end into the VCC edge. I made sure that these would cross over the bus bars, bent them outward to wave in the breeze, and soldered them (cutting off excess lead after I was done).

Gain Enhancement Network

Variations in size of the 10uF capacitor can be expected; what I've set down here as a procedure should take care of all of these. First, I've installed the 470 ohm resistor in the 7th column from the input end; one lead is directly in line with pin 1 (in the 7th column, the 10th hole up from ground). The resistor's other lead is in the 13th hole up from ground. Solder the lower end of the resistor to pin 1. Then, the idea is to connect a 10uF capacitor between the free end of the resistor and pin 8 (with the negative end going to pin 8).

Now, whatever your capacitor is shaped like, it can go in a row near the VCC edge. I laid mine down along the 14th row; its positive end went into the 8th column in this next-to-last row. If your unit is not very long, or if it has radial leads, you can arrange for the negative lead to run along the wiring surface of the board until it meets pin 8. I chose to "make this run over to pin 8" on the component side; I inserted the lead in a hole adjacent to pin 8 (in the 14th column and 10 up from ground).

Don't be overly ambitious -- attempting to make all these arrangements and soldering them later. Anchor the resistor's free lead and solder its lower end to pin 1, then insert the capacitor appropriately, arrange its positive lead to cross its intended resistor lead, and solder these. Finally, make the run to pin 8 and solder this last.

Because pin 8 is an "end pin" -- it is at the upper end of its column -- it is legal to approach this pin at an odd angle with the negative capacitor lead. This means, though, that you'll have to use your head; don't cut this lead long enough so that it can stray over to pin 7. If you cut this lead too short, use a jumper on the component side to connect it (placing this jumper much as you did the 470 ohm resistor).

Off-Board Items

All leads coming off the board -- speaker, battery, and volume control leads -- should be of stranded wire (22- or 24-gauge wire). Because of its inherent flexibility, stranded wire breaks much less often than does solid wire. Using stranded wire does mean, however, that you will have to tin its ends before it can be inserted into holes in the board and into solder lugs.

Cut a bunch of 8-inch lengths of stranded wire, strip off an inch of insulation at each end, bundle the strands and wrap them with solder, and tin them. (See "Soldering, Part III," SKTF, Spring 1981.) If the tinning job is good, the bundle of strands will completely lose its flexibility; it will feel like a piece of solid wire. Cut off any irregular or "frizzy" ends.

Attach these wires to the various off-board components. All of these items have solder lugs: pass each wire through its intended lug to a point just ahead of the insulation, crimp the
wire onto the lug (by bending it double and squeezing it firmly, for example), solder it and clip off the excess tail of tinned wire. The volume control gets three such wires, the speaker gets two; the switch has one of these wires on one terminal, and the red wire of the battery clip on its other terminal. One more of these leads actually goes to the board, it is the cold lead for eventual connection to the input jack, and it is soldered to the ground bus.

The procedures for making the connections to the board are: Find a clear hole that is near the intended connection point;
insert the tinned end into this hole from the component side of the board. (Make sure that the tinned end is in up to the hilt; no wire should be exposed on the component side.) Bend the protruding end down against the wiring side so that it crosses the desired component lead; then secure the insulated portion in some way, either by wrapping it around the board or a component, or by clamping it down on the edge of the board with an alligator clip. Solder the junction and cut off the excess that extends beyond the junction.

Wiring the Power Connections

Somewhere near the 220uF bypass capacitor: insert the negative (black) battery lead in an edge hole and solder it to the ground bus, then insert the
switched VCC lead in an edge hole and solder this to the VCC bus. In each case, steer the tinned end up along the inner-most side
of the bus bar, then loop the end over the bus and out beyond the edge of the board. After soldering a connection, cut off the
excess tail. If you have not already done so, attach the red
battery lead to the other side of the switch.

On the battery clip, how do you tell which wire is black and which is red? With your continuity tester, naturally. The black wire is the one that goes to the small snap; the red one goes to
the large snap.

Speaker Connections

Not only is there no reason to care which wire is which, but you may wish to twist the two speaker wires together into a cable. About 20 holes from the input end, insert one of the tinned leads into the ground edge of the board; loop the protruding end over the bus so that it extends off the edge of the board and solder it. Somewhere near the negative end of the 100uF coupling capacitor -- perhaps 19 holes from the

input end and 12 or 13 up from ground -- insert the other speaker lead, bend it across the capacitor lead and solder it.

Input Connections

The ground lead for the input jack can be installed first; put it through the 7th hole from the input

end in the row of the ground edge, mechanically secure it somehow (perhaps at the VCC edge with an alligator clip), and solder it.

Find out which wire on the volume control is "the counterclockwise, or bottom, end." (With the tester connected between the wiper and this end, the pitch should rise as the control is "turned down.") Connect this wire to ground at the 6th hole from the input end. The arm, or wiper, of the control -- usually the middle terminal -- goes to the free end of the 0.1uF input capacitor. Insert this lead about 7 holes up from ground, and 2 or 3 in from the input end. Bend this lead down to cross that of the input capacitor and solder it.

We now need to find out which is "hot" and which is "cold" on the input jack. Clip one tester lead on the sleeve, or mounting nut, of the jack and try the other tester lead against the solder lugs. The one that beeps is "cold"; connect this to the wire that comes from the ground bus on the board. Connect
the remaining free wire off the volume control to the other lug
on this jack.

Troubleshooting the Amplifier

You have two options: you can plug in the chip, connect a battery and try the thing; or, you can make preliminary tests

with your tester. These continuity tests won't take long, and
they will safeguard the chip to some extent.

Clip the negative tester lead onto the ground bus (just bite the ground edge of the board with the alligator). Then, with a piece of solid wire in the positive test clip to serve as a "test prod," contact the pins of the chip one by one.

First of all, pins 2 and 4 should sound like shorts to
ground -- they are, you know. Pin 3, between them, should give

you no signal because of the coupling capacitor. Neither pins 1 nor 8 should make the tester sound -- their network is off by itself, with the chip unplugged. Pin 7 should sound like a small electrolytic, not taking very long to charge. Pin 5 should sound like a larger electrolytic (you are testing the capacitor and the speaker in series; the speaker's resistance is insignificant). Pin 6 should sound like a large electrolytic, and touching the VCC bus should do the same as pin 6.

Suppose these tests don't pan out; you now have three options: You can get a sighted friend to look at your work, where he or she might say, "Do you really want pins 6 and 7 soldered together?" You can exercise your wire cutters, clear off the board, and start over (you can save the electrolytics if you wish, since splicing solid wire to their leads is sometimes worth the trouble). Or, you can connect pieces of 26-gauge wire to your tester; then use these to "prod into your socket" (from the component side) to see just which pins are unconnected or bridged.

If bridges have been made, there are a couple of ways to clear them: While the pins are being reheated, you can often plow between them with a screwdriver blade on edge. If the tip of your soldering iron is a thin one, it may be used to do the plowing; just place it between the pins and run it between them. The hard way to clear a bridge -- which is sometimes necessary -- is to reheat a connection and remove one of the items (by pulling on it from the component side with locking forceps). Removing components from the area does clear the forest, after all.

Trying the Amplifier

Plug in the chip with its marked end facing the VCC bus. (Very often, before I attempt to fit a chip into its socket, I find a blank section of the board that I can
use as a "pin straightener." Fit the chip into holes in the board first; then, wiggle it around with your thumb until it feels rather mobile in these relatively large holes. Now try plugging it in.)

First, find out which position of the switch is the "off" position; connect your tester to its terminals and select the position for which a beep is not heard. Turn the volume control all the way up. Connect a battery to the snaps.

Turn the switch on. A very slight hissing noise will probably be evident with your ear to the speaker. Now, touch the middle lug of the volume control with your finger; the amplifier should growl, squeal, or complain in some way. If so, it works. There are other things possibly wrong: the volume control could be wired up backwards, the input jack and/or the volume control might not be grounded, or other variations. However, the "chip part" of the project works, and good for you.

Wiring the Oscillator

In the instructions for the amplifier, I told you to wire the jumpers first -- this seems logical. However, in reviewing the circuit for the oscillator, we see that the power connections are all on "end pins;" wiring these first will trap you into soldering pins between previously done ones. We can let this happen and practice the slightly finer dexterity necessary, or avoid the problem by skipping the first paragraph of the circuit -- these instructions will do the latter.

Installing the Socket

For the sake of consistency, we'll mount this socket at the same vertical position as the other --
its columns of pins going in rows 7 through 10 up from the ground edge. It no longer is practical to count columns from the amplifier input end of the board; we'll count them from the left end of the board (the component side down and the amplifier circuit to the right). Suppose I tell you to position the socket about 3/4 inch from the end. Rounding off 0.75 inches to 0.8, this would put pins 5 through 8 in the 8th column.

Therefore, pins 1 through 4 are in the 11th column; pins 5 through 8 are in the 8th column, and pins 1 and 8 appear in the 10th row up from ground. Spread the pins out to the sides as

before and take a deep breath -- here we go.

Installing the 2-6 Jumper

The first logical step is to jumper pins 2 and 6 together. We could make the jumper, cutting off a 1-1/2 inch piece of solid wire, stripping its ends and installing it. I prefer to strip the end of my wire, install and solder it to pin 2, measure off the amount it takes to go around the socket to pin 6, then cut and attach this end. (Remember that you not only have to reach pin 6, but you need another 3/8 of an inch beyond it, so you can strip and install this end.)

The pin 2 end of the jumper goes in the 12th column from the end, 9 holes up from ground. The loop can go either way round
the socket. I have mine passing by pins 4 and 5, then running up to the 7th column, 8 up from ground.

Installing the Charging Resistor

Putting this resistor between pins 2 and 3 is logical, on the one hand, but it cannot be laid out comfortably in two holes this close together. Nevertheless, there is plenty of precedent for standing resistors on end -- one lead coming straight off axially with no bend in it, and the other being bent sharply double, where it can be inserted alongside the standing resistor.

Right at the resistor body, bend one lead over so that it parallels the resistor and its straight lead -- both leads will effectively be emerging from one end, much like radial leads. Put the "straight lead" in the hole immediately next to the jumper on pin 2 (the 13th column from the end and 9 up from ground); put the "doubled-over lead" in this 13th column, 8 up from ground (adjacent to pin 3).

Bend one lead away from the socket to stabilize the resistor. With the other cut so as to protrude 3/8 of an inch, bend it over against its pin; you can now stabilize the assembly even more by bending things so that the resistor body leans up against the socket. Soldering the first lead will really help to stabilize things; after which, the second solder connection will be easy. (Be advised that if you reheat the first while you solder the second, the whole mess will fall apart on you.)

I put the doubled-over lead on pin 3 for a reason: On the component side of the board, this exposed lead makes a nice test point. When installing a resistor or tubular capacitor on end, keep in mind the creation of a test point. For example, whenever I install feedback resistors next to dual or quad op-amps, I always put the doubled-over lead on the output.

A main instability encountered with components on-end is their freedom to turn, thus leading to bridging adjacent pins accidentally. While making the first connection, you may find

the need to stabilize the other lead by "stapling" it down with a hairpin loop of wire, or by stuffing the free lead into a blank hole.

If the component's body is large, its two leads won't fit into adjacent holes in a column or row. Where this is the case, there is nothing wrong with turning it so that the two holes are on a diagonal. In fact, I often do this with 1/4-watt resistors too; for example, by putting the doubled-over lead in column 12 instead of 13 (thus, closer to the socket), we would have potentially more room for other components going to pin 3.

Installing the Output Network

This 1/2-watt unit can lie down properly in the 14th column from the end (three spaces to
the right of the socket); one lead is adjacent to pin 3 (8 up from ground), and the other is the third hole up from ground in

this column. Once the pin 3 end has been soldered, cut the free

lead to about 1/2 inch and lay it down so as to point toward VCC
-- away from ground.

The 10uF output coupling capacitor goes in the 16th column. Its positive end goes in the 5th up from ground -- where it can
be bent over to cross the free end of the resistor -- and its negative lead goes in the 11th up from ground. After soldering the positive lead and cutting off its excess, trim the negative lead to 1/2 inch, then bend it over to await connection to an output jack.

Making Power Connections

Now, without fear of boxing ourselves in, the ground and VCC connections can be made. (We could have first installed the timing capacitor before doing these, but this would preclude an option of placement that I want to show you.)

Tying pin 1 to ground is done with a jumper in the 12th column from the left end, stretching from the 10th in this column down to the ground edge. As before, my inclination is to solder one end with the bulk of the wire wrapped around the board for stability, then cut the jumper about 3/8 of an inch too long and connect this end later.

Now, we have the option of tying pin 8 to VCC and separately tying pin 4 to VCC, or jumpering pin 4 to pin 8, and connecting either of these to VCC with a single wire. For no good reason, I usually put a jumper between pins 4 and 8, then tie pin 8 to VCC. (In fact, I usually do this immediately after installing the 2-6 jumper; my confidence about soldering pins between pins is pretty good after 15 years.) I will describe separate connections; if you're building more than one of these boards, as suggested in
the abstract, plan on doing it both ways.

The pin 4 jumper can go into a hole immediately adjacent to its pin, the 7th up in the 12th column. Since there are wires going to pins 3 and 2 -- rather close by -- pay attention to the elimination of any exposed (noninsulated) wire on the component side; push that baby all the way up to the insulation. After soldering this end, cut the wire so as to reach about 3/8 of an inch beyond the VCC bus; strip and solder this in the usual way.

The pin 8 jumper can go from the 10th to the 15th hole in its column. Which column it goes in depends on whether or not

you decided to jumper pins 4 and 8 together. In that case, the column would have to be 6 from the end. The way we did it here, the 7th column is available.

The job remains to bypass these power connections with a 0.1uF capacitor. This can be done above the socket in the 12th column up from ground.

Put the capacitor leads in the 8th and 11th holes in the 12th row up from ground -- just above the socket. For stability, bend one lead away and crimp it over the VCC bus for now. Cut the other lead, by trial and error, so that it comfortably reaches its pin, but does not span the space between this and the next pin. (In other words, each lead approaches its pin, either pin 1 or pin 8, in line with the column that all the pins share; your job is to see that this lead falls short of the neighboring pin. Given the hole placement mentioned, this will be accomplished if the lead protrudes exactly 0.2 inches.)

After this lead has been soldered, uncrimp the free lead from the edge of the board; cut and attach it in like manner.
Especially while soldering the second connection, forcing the
capacitor to lean up against the socket on the component side
will greatly enhance its mechanical stability.

Installing the Timing Capacitor

Electrically, there are three places to put this unit: It could reach from pin 2 to ground, if there weren't so much clutter already there. It can easily go from pin 6 to ground -- 2 and 6 are tied together, remember. A third possibility, and one I often use, is to place it between pins 2 and 1 (1 being "ground" at the socket), thus treating this capacitor like a hairpin jumper or the charging resistor. The only thing wrong with this last option is that, with all jumpers and components in adjacent holes, it will tend
to be mechanically unstable while you're working on it. There are plenty of times in other circuits, however, that a capacitor has to go between adjacent pins, and I want to encourage you to try this configuration on your second or third versions of this circuit.

If you put it between pin 6 and ground, it should go into the 6th column. Its pin 6 lead will be in the 8th hole up from ground, and the placement of its other lead should be determined by its size -- possibly 6 up from ground. Run the bottom end over the ground bus, crimp it over the edge, solder this ground point, and cut off the excess. Cut the pin 6 end to length (perhaps 3/8 inch), lay it atop pin 6 and solder this junction. (Mechanical stability will be improved by forcing the cap to lean against the socket.)

If pins 1 and 2 are used to connect it, place the capacitor in the 13th column, forcing its leads to conform to the 0.1 inch spacing; i.e., its leads should go in the 9th and 10th holes up from ground. Bend the pin 1 lead away and clamp it in some way, perhaps by poking its end through a hole from the wiring side and crimping it there. Cut the pin 2 lead to length (3/8 of an inch for starters, trimming off its excess after soldering); position this lead on pin 2, force the cap to lean against the socket for stability, and solder pin 2. Cut the free lead to length and do the same on pin 1.

Wiring the Output Jack

Install 8-inch lengths of stranded wire on the output jack. With your tester, identify which wire is the cold lead and put this in the ground edge -- in the 18th column from the oscillator end. The hot jack lead can go in the 18th column, say 12 up from ground, where it then crosses the output cap's negative lead.

Testing and Troubleshooting the Oscillator

Just as you did with the amplifier, you should trace the circuit with your continuity tester. Connect the tester's negative lead to the ground bus. Outfit the tester's positive lead with a piece of solid wire to serve as a test prod. (If this "test prod" is of 26-gauge wire, you can insert it into the socket from the component side, thus checking the solder bond to the socket pin; otherwise, confine your testing to the solder connections on the wiring side.)

Pin 1 should look like a short to ground. Both pins 4 and 8 should look like a very leaky electrolytic (your tester is trying
to power the amplifier); though exactly how they sound is rather meaningless, both 4 and 8 should make the tester respond the same. All other pins should look like open circuits to ground.

For the next tests, the continuity tester can be clipped onto components in either of two places; put one of its leads on either the junction of the 100 ohms and the 10uF cap, or on the "test point" formed by the exposed lead of the 100K resistor -- this also going to pin 3. With a "test prod" in the other clip, explore pins 2 and 6; both should make the tester emit a low buzz, signifying that you are viewing them through 100K ohms. Both 2 and 6 should look the same to your tester, and an open circuit should be seen to pins 1, 4, 5, 7, and 8.

Inserting the Chip

As before, the end of the package containing pins 1 and 8 will be nearest VCC. (This kind of consistency is something you should strive for -- if you don't, you'll curse yourself later when repairs have to be made.) Because of the wide variety of manufacturers of the 555 chip, marking it could consist of: a notch between pins 1 and 8, a small indentation above pin 1, a "mold-release mark" that is closer to pins 1 and 8 than it is to 4 and 5, or even a painted dot over pin 1 (which can often still be felt with the tip of a braille stylus).

The first job for the test amplifier is to check the
oscillator. Make sure that the amplifier's volume control is turned up. Connect the battery and turn the switch on. Touching the junction of the coupling capacitor and the wiper of the volume control (the farthest connection at the amplifier end), you should hear crackling and hum from stray AC fields. With this finger on the amplifier input, touch the oscillator output (say, the junction of the 100-ohm resistor and its output capacitor) with another finger. The squall of the 700 Hz tone produced by the oscillator should be unmistakable; it should put the neighborhood tomcat to shame.

Boxing the Project

I declared that the cover plate of the box be the "bottom." The largest plastic face, now the "top," has the speaker mounted right in the middle, with the on-off switch and the volume
control mounted near one end. The two jacks are mounted in the long side, opposite the circuit board.

Specifically: I first drilled the center hole of the speaker pattern at 3-1/8 inches from one end and 1-7/8 inches from one side; then I made a pattern which was 7 by 7, with the corner holes missing. (See "Hints and Kinks" of the previous installment.) For the controls, I made pilot holes at a position of 1-1/4 inches from both the end and the side; this puts them 11/4 inches apart, by the way. For the jacks, I drilled two pilot holes an inch down from the top and about
1-1/2 inches from each end.

Next, the jack holes are drilled to 1/4-inch to accommodate the jacks' bushings. The appropriate hole for the toggle switch depends on which kind you have -- usually 1/4-inch for miniature units and 1/2-inch for the "standard size;" let this be in the
hole nearest the board. The volume control will probably need a 3/8-inch hole; let this be the hole nearest the jacks.

Mounting-holes in the board and box must now be drilled to match (you may have done this already, back in the "Preparing the Board" section.) A number 40 drill is used.

I tape a piece of braille paper to the side of the box to keep my solder connections from scratching it. Then I enlarge one of the desired holes in the board -- the third over and the third down from one corner. I position the board against the side of the box and carefully drill through this chosen hole to make a matching one in the box; I put one of the screws through it (3/4-inch 4-40 machine screws) to hold this position. With the board still in place, I count three over and three up on the opposite corner and drill through both items; I insert another screw to hold the board by two anchoring points. Finally, I drill the remaining corner holes. Count carefully; otherwise, the board will not fit its screws when it is flipped over and put inside.

The "proper way" to mount a board like this is: Insert No. 4-40 3/4-inch screws from the outside and lay this side on the table so as to keep the screws from backing out. Put 1/4-inch
long spacers on all four screws and install the board so that it rests atop these spacers. Thread on the No. 4-40 nuts and
tighten the screws.

When I have no spacers, I install nuts on the inside of the box to hold the screws in place and tighten them down. Then, I install another set of nuts which I adjust to hold the board up off the panel by 1/4 of an inch or so; place the board atop these nuts and screw on a third set to hold the board firmly. This last set must be tightened down inside the box with a sawed-off 1/4-inch nut driver; turning the screws on the outside won't do any good, since they're already tight. For this trick to work, the holes must match exactly -- I usually end up enlarging the holes in the board to make the assembly fit together.

Your toggle switch may have a pair of nuts, a lockwasher of the "inside-star" type, and another washer-like thing which has a "locating lug" on it. Nobody drills holes to accommodate this "locating lug" of the latter washer; you can either throw it away or use it upside down. The point of having two nuts is to permit adjustments for different thickness panels; a nut on the back
side keeps the bushing from coming all the way through and

looking ugly from the outside.

In mounting the switch, turn the bottom nut down to expose perhaps 3/16 of an inch. Next, install the "locating plate" with its little protruding lug facing the switch -- out of the way. Next install the lockwasher. Put this assembly into the hole and try to fit the outside nut on what's left of the exposed bushing. If it doesn't have enough threads, adjust the bottom nut until it does.

The volume control comes with a nut, and maybe a couple of washers (one may be an inside-star lockwasher). Many controls come with a "locating lug" on the edge of the case; this is meant to fit a hole in the panel, where it would keep the control from turning. Most people bend this lug out of the way, so that they don't have to drill a hole for it. If you choose this approach, bend the lug at an angle outward, where it can be adjusted to press up against the panel and give the assembly some resistance to turning. If the control comes with a lockwasher, put this between the panel and the control, not between the panel and the nut (you don't want to keep the nut from turning, necessarily,
but you want the lockwasher to hold the control still).

Mounting the jacks is obvious. They often come with smooth washers that look pretty behind the nut. Put smooth washers on
the outside of the panel; any lockwashers should go inside.

As discussed in Part II of this series, mounting of the speaker is done by: Rubbing the speaker around on sandpaper
until the cardboard rim is flush with the metal frame. Then
spread glue around the rim, trying fastidiously to avoid the speaker cone, and plop it into position. I prefer Goodyear's Pliobond cement for this, but any household cement (not "glue," in the strict sense) is worth trying.

I wish there were a good and simple way of mounting the 9volt battery. I don't like any of the commercial holders; the geniuses who designed them left no room for nuts to hold them down. The best one can do with these commercial holders is to mount them with screws whose nuts are on the outside of the box, which looks terrible.

The most direct solution seems to be to cut from packing material an odd-shaped piece of Styrofoam that can be cemented in place to cradle the battery. My all-time favorite method is to install two 1-inch-long No. 6 screws 1-1/2 inches apart (threading nuts all the way down to hold them firmly); then making a metal strap to fit over these bolt ends, whereupon two more nuts can be provided to squeeze this strap down onto the battery underneath. Maybe one of you has something better to offer.

Before securing the bottom plate with screws, try to ascertain whether or not it is touching solder connections on the nearby bus bar when in place. If this happens to be the "ground" bus, touching will cause no harm. If the nearby bus is VCC, the possibility exists that the battery can be shorted out from the outside of the box. In this latter case, put a strip of Dymo Tape along this edge of the panel so that it will be insulated from the bus. (Putting tape along both edges would make it so that the next guy won't have to figure out why one strip is there and which way the panel must go.)

You will want to make at least one more pair of test leads out of a bisected alligator clip lead. Cut a clip lead in two; strip and tin its ends. Mark one of them with tubing or tape to denote the "hot lead;" this one will go to the short lug (the tip contact) of the plug. The unmarked lead goes to the sleeve (the long lug with the forked cable clamp on it). (Before connecting anything, however, remember to slide the plug cover down the
wires first.)

Notes on Using
the Amplifier/Oscillator

First of all, because of the levels involved, the amplifier should never be made to look directly at the oscillator output without some attenuation. A 1 megohm resistor inserted in the
hot lead between the two is sufficient. One could, for example, make this into a code-practice oscillator by putting the key in series with 1 megohm, this combination going between the hot oscillator output and the hot amplifier input. (No provision need be made for the cold sides, since these are wired internally.)

When tracing signals in digital projects with the amplifier, keep a 1 meg resistor in that hot test clip. (The amplifier's
cold test lead will typically be grounded to the project under test.) Any signal around a volt or so, although it will be seriously clipped with the volume up, can be tested with perhaps 47K or 100K in the hot test clip. You can monitor modern tape deck outputs and things directly, if you have the right patch cord.

The output of the oscillator is quite high, by most
standards; it would be well to keep 1 megohm in its hot test clip when tracing back through circuits to find a dead stage. Although its frequency will be loaded, it can drive a speaker directly, making it handy for testing speakers and earphones.

With a 1 meg resistor between the oscillator and the
amplifier -- so as to get a good audible signal whose volume is adjustable -- the sound from this instrument can be recorded on tape. Playing this back will give you an idea as to how much "wow" and "flutter" the tape machine has.

The output of this oscillator has so many high-frequency components that it can be used to check shortwave receiver circuits. Giving it an antenna of a couple of feet of wire is sufficient to hear it quite well on AM radios.

Hints and Kinks

Straightening Bus Wire

No matter how careful they were in rolling it--or you are in unrolling it -- solid wire always has kinks and bends in it. There is a neat way to eliminate these completely.

Cut off a length of wire to 2 feet or longer. Put one end in a bench vise; grab the other end in a pair of pliers. Give a sharp tug on the wire, so hard as to stretch it slightly. When
you next examine it, all small bends and ripples will be gone. (Be sure when you do this that there is room behind you; you don't want to break an elbow on something nearby.)

Making Jumpers

There are times when you must unavoidably jumper adjacent pins together; a hairpin loop inserted from the component side is the logical choice. The first mistake to avoid is making it so small that you cannot control it. Make an honest component out of it, with leads that are 1-1/2 inches long. Once through, the protruding leads can be properly stabilized by stuffing one into a hole from the wiring side, or by holding it with a clamp.

To make such a long-legged jumper: Strip an inch at a time off the end of your insulated wire; do this three times until you have a nice overall length of 3 inches. Next, before cutting the bare section off, use your strippers to take an additional 1/4- or 3/8-inch piece of insulation up to the middle of the bare section. In other words, you now have a bare section of wire with a little bead of insulation which has not been completely removed. Now cut it off the spool and bend it into a hairpin at the bead of insulation. You will find that it is easier to insert if one leg is slightly longer than the other.

Making "Locator Holes" for Controls That Need Them

For controls that turn, having an offset lug which fits into a matching hole in the panel is a nice idea. The trouble is that the position for the appropriate hole must be exact, and this is very hard to measure.

First make the right-sized hole for the bushing. Then, insert the control from the front side, put its locating lug about where you want it, and press and turn so as to make a scratch in the panel. Drill a pilot hole at this scratch. Pick a drill whose shank diameter compares reasonably with the width of the lug and finish the job.

If the knob intended for this control is small enough to
make this locating hole visible, scratching around on the panel
would not be appreciated. You can sometimes make a mark that isn't a scratch; if you put masking tape where you want the hole, a sufficient impression can sometimes be made in the tape, rather than the finish of the panel.

More Tricks of Measurement

If you haven't gotten all the different kinds of "rules" yet (namely, the AFB Rotomatic Rule
and the Stanley Combination Square), there are a couple of ways to put your project on a straight course. For example, I have seen some wonderfully complex drilling and mounting of hardware which was laid out using perforated board as a template. If you have a piece of Vector Board which is larger than the box, you can securely tape the box to it -- carefully lining up two edges -- and count holes to find various distances.

For those who have a knack for paper folding (called
"origami" in Japan), you can start by cutting "panels of paper;" then, by strategically folding the paper panel, you can make the intersection of creases tell you where to drill.

For example: There are three hole locations on the top panel of this project. The exact center, from which the speaker grille sprouts, can be found by folding a piece of paper (measuring 6-1/4 by 3-3/4 inches) into quarters, then noting the intersection of the two creases. To get locations for the controls, fold the paper in half the long way and fold the corners of one end over to meet the middle (sort of like making a paper airplane). Now at this end, after the paper has been unfolded and laid flat, crease a flap that's about an inch wide; it will intersect the diagonal creases at two perfectly symmetrical control locations. (They won't be exactly where I drilled them, but it's your project.) Tape the paper to the top of the box and drill pilot holes at the appropriate intersections.



The LM3909 is a monolithic power oscillator particularly designed to flash light-emitting diodes (LED's). The main reason for describing it here is that it is the heart of the "Marklites" of the next article. Other than this, it has application in the toy field, and it can be run at audio frequencies.

There are times when I've wanted a flasher of simple design. For example, I was once told about a Swedish device for calling
the waiter's attention to your dinner table; you would show him your little red light and tell him that, if you needed him, you would put it where he could see it. (I don't know if the circuits for LED's would catch his attention, but this little chip can drive an incandescent lamp, which ought to bring him.)

This chip will light an LED from a 1.5V battery; its timing capacitor provides a voltage boost, so as to deliver pulses of 2 or more volts to the lamp. The current-limiting resistor for the LED is contained inside the chip -- along with internal timing resistors -- and thus, a low parts count is possible.


  • Supply Voltage Range--1.15V to 6V
  • Operating Time (with single alkaline D cell--2.6 years
  • Flash Frequency (with 350uF cap)--1Hz Typ.
  • Flash Frequency (with 0.3uF cap)--1.1kHz Typ.
  • Peak LED Current (with 350uF cap)--45mA
  • Pulse Width (with 350uF cap)--6milliseconds at half amplitude
  • [The editor didn't quite understand that last one.]


The only internal circuit shown in the literature is the "actual integrated circuit," not a block diagram of which we could make sense. However, certain key elements are worth noticing. For example, between the positive supply terminal, pin 5, and the typical LED's anode connection, pin 6, is a 12-ohm resistor; pin 6 is called "R Lim." The internal timing resistor is split. Between pin 4, supply ground, and pin 1 (which is called "Fast Timing") is an internal 3K ohms; between pin 1 and pin 8 (called "Slow Timing") is 6K ohms. Thus, you have a choice of timing resistances -- 3K (using pin 1) or 9K (the sum of the two at pin 8).

Pin 2 is called the "Output;" yet, you will notice that the LED is never connected there. Inside the chip, this "Output" is the collector of a transistor that shorts the positive end of the timing capacitor to pin 4. The result is that the cathode of the LED is driven "below ground," and the voltage from pin 5 to the capacitor's negative end exceeds the supply.

Pin Connections

  • Pin 4--Minus V
  • Pin 5--Plus V
  • Pin 6--R Limiting Pin 2--Output
  • Pin 1--Fast RC, 3K Pin 8--Slow RC, 9K
  • Pins 3 and 7--No Connection

The General Hookup

Pin 4 goes to the negative of the battery; pin 5 goes to the positive of the battery (1.5V or 3V). The anode of an LED goes to pin 6; its cathode goes to pin 1, or 8, or both 1 and 8 tied together (I gather that tying 1 and 8 together is essentially the same as using pin 1 by itself). Pin 1, 8, or both together also goes to the negative end of the timing capacitor, the positive end of which goes to pin 2. Pins 3 and 7 have no connection.

A 3-Volt Flasher

Pin 4 is the negative supply; pin 5 is at plus 3V. The LED goes between pin 6 and pin 8 (cathode at pin 8). Pin 1 is left open. A 300uF 3V capacitor goes between pin 8 and pin 2 (positive at pin 2). The flash rate is 1Hz.

Fast Flasher

Pin 4 is the negative supply; pin 5 goes to plus 1.5V. Pin 1 is not used, but there is an external resistor of 1K connected across the whole internal string -- from pin 8 to pin 4. The LED goes between pins 6 and 8 (anode at pin 6). A timing capacitor of 100uF goes from pin 8 to pin 2 (positive at pin 2).

Incandescent Flasher

Pin 4 goes to the negative of the battery; pin 5 goes to plus 6V. The lamp goes between pins 5 and 2. A 400uF timing capacitor has its negative side going to pins 1 and 8 (which are tied together), and its positive side goes to pin 2.

"A Buzz Box"

This is said to be good for checking continuity of low-resistance paths, such as coils. The negative side of a 1.5V battery goes to pin 4. The positive side of the battery goes to a test probe; the other test probe (the negative one) goes to pin 5. The speaker (12 to 16 ohms) goes from pin 2 to pin 5. A 1K external resistor goes from pin 8 to pin 4; pin 1
is left open. A timing capacitor goes from pin 8 to pin 2 (positive at pin 2), with pin 2 going through 0.1uF to ground.

Note: The literature says this chip can drive an 8-ohm speaker; why they specify 12 to 16 ohms in the above, one can only guess -- and try an 8-ohm unit.

A 1 kHz Squarewave Oscillator

Pin 4 goes to the negative of the battery, which is taken as ground. Pin 5 goes to plus
1.5V. Between pins 1 and 8 is a 10K pot (symmetry adjust); the wiper of this pot goes through 0.2uF to pin 2. An external 2K resistor goes from pin 1 to ground. Pin 2 is the output. The waveform, which is a little ragged, goes from 1.2V down to 0.1V; this is shown with an output load of 10K, from pin 2 to pin 4.


by Mark Brown

Have you ever seen the need to deter unwanted intruders, be they kids, burglars, landlords or worse? Your first thought
might be to put up a sign, "DANGER! HIGH VOLTAGE!" By itself, though, this sign would not be very convincing. Anybody could have one of those.

It would seem more frightening if you put up a piece of junk electronics along with the sign. But then, somebody could smash that and break in anyway. What if the decoy electronics actually did something?

Putting in a burglar alarm is a lot of work. You have to install power lines and sensors, and you have to be wary not to
get caught in your own trap. I invented these "Marklites" as paste-on decoy indicators that there is probably a burglar alarm to overcome. They are merely flashing LED's which are inexpensive to put on a number of windows, and there is no complicated installation necessary.

A Marklite consists of a red LED driven by the National Semiconductor LM3909 flasher chip (Radio Shack No. 276-1705). The only other components needed are an electrolytic capacitor and a 3-volt battery pack. The LED may not be visible in direct sunlight, but I've been able to spot it from 50 or 100 feet away in the dark. It should be visible to anyone carefully examining the window.

Life Expectancy

The life of the batteries will be
affected by the efficiency of the LED and the flash rate. I am fairly happy with the LED listed here, and with my choice of capacitor that determines the rate. A 220uF capacitor produces a blink rate of about 1 to 1.5 cycles/second. The life expectancy using alkaline batteries is typically about 6 months. To help you keep track of when to change the batteries, put a stick-on label with the date of replacement on the battery pack.

Cost of Operation

All the parts for the Marklite may be purchased from Radio Shack. The total price for all of the parts listed here, except batteries, is $9.42. If you shop around, or already have some of the parts, the cost could be as low as $4. Batteries will cost about $2-$3 per Marklite per year.


The flasher should be located on the inside surface of the window, aimed outward so that it is visible from the outside, preferably near the window latch. Mount the battery pack so that it's not visible from outside the window. Attach
the battery pack to the underside of the window sill using a
small piece of stick-on Velcro so that it's easily removable. Some windows may require the leads from the battery pack to be longer than those that come on the snap connectors; if so, add an appropriate length of two-conductor wire to the battery snaps.

I hang all the components directly on the socket without a circuit board. I bend the leads of the LED so that it faces the same way as the top of the chip; then I attach the Marklite to the window using a 5/16th inch square piece of double-sided foam tape on top of the chip. The intruder sees the chip pins (from the top), an electrolytic capacitor, and the flashing red light.


The Marklite uses an LM3909 flasher chip, a Radio Shack 276-0412 LED, and is powered by two AA cells in series. The negative battery wire goes to pin 4. The positive battery goes to pin 5. The anode of the LED goes to pin 6. The cathode goes to pin 8. Also going to pin 8 is the negative side of the 220uF electrolytic; its positive side goes to pin 2.


First, cut off pins 1, 3, and 7 from the socket; this will give you room to crimp leads of components to the pins that are left. Position the electrolytic (mine having radial leads) above and in line with the socket and connect it between pins 8 and 2 (positive on pin 2). With the LED facing the table and its anode lead nearest you, bend both leads straight to the right and solder them to pins 6 and 8 (the anode to pin 6). Then, attach the two-conductor cable to pins 4 and 5.

Parts List


Shack #
LED Flasher IC276-17051.69
Red LED276-0412/.69
220uF Capacitor272-956.79
8-Pin DIP Socket276-19952/.59
9V Battery Snap270-3255/.99
Battery Holder for 2 AA270-382.79
2 AA Alkaline Batteries23-5524/2.99
2-Sided Tape Strips64-23448/1.89
Velcro Fastener Strips64-23454/1.89
6" 2-Conductor Wire- -.10



Five years, I've been doing this! With your help I have. Also with the backing of The Smith-Kettlewell Eye Research Foundation, and the National Institute of Handicapped Research. (The first issue was in the Fall, or thereabouts, of 1980.)

For those who might think that the "Paper to Project" series threatens to engulf their favorite aspects of this magazine, have no fear; the circuit board of this issue comprises most of the work in building the lab instruments to come. What is important to my responsibilities is that the text of that series is research. You are my research subjects (guinea pigs).

You are the ones who can determine if this is an effective way to transform our experimental training program into a consumer product. Please keep this in mind while you read the series, especially while you build the instruments described. A question to be asked is, would you recommend these articles to others who are joining the field of electronics?

More computer-related subject matter is coming. I can assure you, however, that it will primarily be in the vein of computers as components, with little emphasis put on evaluating computer systems as traditionally viewed (a keyboard and screen, an access system, and word-processing software). Computerrelated chips are being incorporated into devices just as vacuum tubes used to be. Where they can be an advantage to builders of special devices -- used as "dedicated components" -- they will appear here.

For those who use "personal computers," or wish to, I can
highly recommend the book by National Braille Press called, "The Second Beginner's Guide to Personal Computers for the Blind and Visually Impaired," compiled and edited by Ms. Diane Croft. (National Braille Press Inc., 88 St. Stephen Street, Boston, MA 02115, 617-266-6160.) The list of books and resources in the third volume is astounding; it would be foolish for me to duplicate it or try to improve on it.

As far as catalog indices are concerned, it's running two-to-one against reprinting them within these pages. (I got three replies to my query -- two said not to.) Well, there goes another one of my bright ideas to fill magazines and put us back on time. Thanks a lot. (Somehow, I'll pass out recordings of same.)

Speaking of putting us back on schedule, I have this pesky engineering career that is my first responsibility; the magazine can only share a small part of my time. It sure would help if those of you who are sitting on stacks of blank paper and pinned under hat-fulls of ideas would send me some material. Thanks to Bob Gunderson, and a couple of short papers from others, I have a head start on 1986, but I don't see a vacation in sight. (No impertinence meant to those who have contributed heavily, but please keep it up.) Come on there, gang -- bail out your ol' pal Bill.

Whatever happens, I'll promise to put out a sixth year, if you'll work toward having a happy and fulfilling New Year. May irons in the fire keep your gentle selves warm.


You will note the inclusion of a Christmas present. Thousands of these animals were donated to the Handicapped Users' Group of The Tulsa Computer Society, who passed a great box full onto me.

Donated by IBM, they are vacuum switches. (Although stamped with an IBM number, I think they are originally manufactured by Fairchild.) Pulling a vacuum of perhaps one-half inch of mercury closes a leaf switch inside. A nipple is provided for connection to the vacuum source (your lungs, for example).

Actually, conversion to a pressure switch can be done by cementing a fitting on the opposite side, over a hole that serves to let in the atmosphere. With nipples on both sides, this switch can detect a pressure differential. If you'd like more, say so and I'll send you some.