Smith-Kettlewell TECHNICAL FILE

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

Bill Gerrey, Editor

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

Produced by

TABLE OF CONTENTS

From Paper to Project, Part II

A Transistorized Wide-Range Crystal-Controlled Test Oscillator

A Great Gift Idea for Someone Else's Kid

The In's and Out's of the Jameco Catalog

Editor's Corner

FROM PAPER TO PROJECT
Part II

Abstract

In this installment, building of a "continuity tester" will be described. It is the first instrument in this series which will be fitted into a box. Methods of drilling speaker holes and securing things will be entertained.

Introduction

This circuit has appeared in SKTF before. If you do not have such an instrument yet, shame on you! Without it, you cannot tell which way diodes are facing (or many electrolytic capacitors, either). Without it, you will have to ask a sighted friend which lead is which on battery clips (they being color-coded black and red). Without it, you will not be able to discriminate between the various resistors that you have bought for a project. (You cannot identify resistors, exactly, with this instrument; but you can, knowing what you have in your collection, figure out which ones are which.) Without it, you cannot identify the terminals on mini phone jacks. Without it, checking your wiring of other projects is much less certain. Without it, you cannot test cables and patch cords. Build this one thing, at least!

(It's only fair to tell you that this identical device is commercially available from Science Products (formerly Science for the Blind), 1043 Lancaster Ave., Berwyn, PA 19312; phone, (215) 296-2111. It is sold under the name of "Audicator.")

Future installments of this series will rely heavily on use of this particular continuity tester (its only partial substitute being pretty good vision). In the last installment, you will perhaps recall the pointed question, "How can you read the color codes on the parts?" "This instrument, that's how," is the answer.

There are other commercial audible continuity testers on the market -- Radio Shack sells one. However, only one I have seen (and it was quite expensive) gives any indication about the resistance being tested. Without this feature, all you can check with them is the polarity of garden-variety diodes, battery clips, and open circuits -- only half the job. [Our deaf-blind readers should not feel left out; the article "Continuity Testers -- Old and New" (SKTF, Fall 1982) describes a circuit using a 555 "timer chip" which can deliver enough power for good tactual indication, provided the frequency range is lowered.]

Getting Started

Let's review the rules of the "point-to-point wiring game" -- the abbreviated set:

  1. We have to choose the size of board. (In this case, this will be given, since the size of the box that contains the project fixes this size.)
  2. All components are inserted from, and rest on, the "component side" of the board. Even wires from the battery are inserted from this side-the top side. Interconnection of the parts is all done on the "wiring side" of the board-the bottom side.
  3. Very little, if any, extra wire should be necessary (with the exception of supply busses). Placement of the items should be done so that components' leads will be the main source of wire.
  4. Components should not have to reach over each other; give them a space of their own, where they can rest in comfort.
  5. Where possible, components shall not be placed at odd angles; they should be fitted into a straight row or column of holes.
  6. All of these rules are to be broken "for cause"; none should be held binding where compromise would make more sense, or wherever critical electrical place ment makes compromise necessary.

You get the advantage of having a physical design that was arrived at through several months of work in our training program. A good cabinet has been chosen -- the Radio Shack No. 270-221, measuring 4.4 inches by 2.44 inches by 1.06 inches. A board that is 2.2 inches wide and 3 inches long covers most of the bottom of the box, leaving room beyond one end for the 9-volt battery to stand up on edge. All of the components (with the exception of the test lead jack) are mounted on the component side of the board; even the speaker is cemented face down on the board, where it speaks through the perforations. Yet, with all its refinement, this design is not completely free of rule breaking, as you will see.

Know Your Parts

The following items are to be procured:

Which Part is Which

Now is when you could really use some background in "electronics theory." In the training program, it is our experience that, if students tend to forget (or loosely associate) terms like "transformer," "transistor," "resistor," and the like, instructions on putting them together come hard. Take an hour or two of drill, review your tape (or notes) on seeing electronic parts, and become familiar with the following:

A jack is a thing that you plug a plug into; the plug is the male item that plugs into the jack. The resistor is a little round capsule with a lead coming out each end. A transistor is a device with three leads (think three!); these leads are called the emitter, the base, and the collector. A transformer is large, even in the best of cases; it has "windings" around some sort of magnetic core. The capacitor in this project may take several shapes. If it is disc ceramic, it will look like a disc with leads emerging from one edge. If it is mylar, it may either have "radial leads" or "axial leads." If it is mylar with radial leads, it will look like a piece of candy with leads out the bottom. If it has axial leads, it will look like the resistor, except that its diameter will be larger.

Plugs, jacks, and speakers usually have terminals called "solder lugs" to which hookup wire is attached. Most often, solder lugs are small metal tabs that have holes into which wire can be inserted -- then bent double and squeezed tightly about the lug with needle-nosed pliers. (In the old days, the practice was to insert the wire, and then wrap the wire around the lug. This practice caused more trouble than it was worth, especially in repair.)

The transistor you buy may be in one of a few different configurations. For example, an odd-ball package might be a half-round plastic unit with the three leads "in line" (in a row adjacent to the flat side); usually, with the leads pointing upward and the flat side away from you, the three leads -- from left to right -- are: emitter, base, collector. The more usual transistor has the leads arranged in an isosceles triangle, as described below.

Most commonly, the leads of a small-signal transistor are at the vertices of a triangle, this triangle having a long base and two short sides. With the leads pointing upward and the long side of the triangle toward you, the lower left corner lead is the emitter, the one in the middle at the top is the base, and the one at the lower right corner is the collector. Also, quite commonly the emitter lead is marked on the case -- there will be a small protruding tab (or, in some cases, a flat side on the package) immediately below and adjacent to the emitter.

The transformer has a bobbin of windings around the center post of an iron frame, this frame encircling the bobbin on two sides as well. Two leads emerge out from under one edge of the bobbin on one side of the iron frame; these are the leads to the "secondary winding." On the other side of the frame, three leads emerge; these are leads to the ends of the "primary winding," plus a "center tap" on this winding. Usually, the center tap is the middle lead (although this needn't be the case, since they are color-coded as well). Sometimes, the secondary has a tap or two for different impedances; an ohmmeter is required if the blind person is to identify which winding is which, in that case.

Usually, on the corners of the frame adjacent to these two sets of leads, there are metal mounting tabs by which the heavy transformer is secured to its mounting surface. For large transformers, these tabs contain screw holes; on this miniature unit, the tabs are meant to solder into a PC board. On perforated board, holes are enlarged to accept the tabs, whereupon they are bent outward against the "wiring side" of the board.

Presenting the Circuit, Irrespective of Layout

(Note: For the theory of operation of this circuit, see "Continuity Tester Circuits, Old and New," SKTF, Fall 1982 -- the particular circuit being called "Transistor Blocking Oscillator.")

In Part I of this series, I did something I will rarely, if ever, do again; I described the layout of a circuit in pictorial terms, and followed this with the circuit. No more free rides for you; this time, the presentation will be in the proper order -- circuit diagram first! Someday soon, the "circuit" will interest you most. Any palaver about where some guy placed the parts will be of more casual interest -- "Hum, that's how he did it?"

As you read this -- two or three times -- try to arrive at your own ideas of placement. You know the shapes of the parts: the transformer has three leads of immediate interest out one side, the transistor is a three-legged stool whose "legs" are in an order you have memorized (emitter, base, collector), and the capacitor is a "coupling device" which would logically fit in between its appointed connections. Forget about the size constraints of the board, for the moment; picture these items lying on their backs on the table in front of you, and how they might be connected together in the middle of a circuit board as big as your hat.

Circuit Description

The centertap of the transformer's primary goes to the emitter of a transistor. The collector of this transistor goes to the negative side of the 9-volt battery and is taken as ground. One end of the primary winding goes through 0.1uF to the base of the transistor, while the other end of the primary goes to plus 9 volts.

The sleeve of the test-lead jack (and hence, the negative test lead) is grounded. The tip contact of this jack (and hence, the positive test lead) goes through 4.7K to the transistor base.

The transformer secondary goes to the speaker. This completes the circuit.

I don't care, at this stage, what your mental image of the physical arrangement is. I do, however, want you to refine this image, based on the very specific construction procedures given in the next section. When you've finished building the instrument, come back and re-read this circuit to see if it doesn't play a familiar tune.

Step-by-Step Construction

It's only fair to mention here that the "Hints and Kinks" section contains ideas you will want. Rather than mindlessly following these instructions, only to read on and scream "Now you tell me!," I recommend that you read this section, then go on to "Hints and Kinks," and then come back to the construction business. Also, this paper ends with a discourse called "Commiserations" which may cheer you up as frustrations mount; you have my permission to read this first, as well.

First, you need a piece of perforated board of the prescribed dimensions -- 2.2 inches by 3 inches. As per the procedure in the "Hints and Kinks" section of Part I, "knock off" such a piece. (I counted 21 holes in from one edge of my large piece, lined this up with the edge of my clamp, then scribed and broke the board along its full width. Then, I counted 29 holes away from one end of the long strip I had made, and made another break here.)

If you have the recommended piece of Radio Shack board, it will be too long for many clamping arrangements. It is possible, with care, to make a partial break -- as long as the clamp is wide -- then move the board along and finish the job. Another way, of course, is to make a saw cut into the board so that a tongue-like section the width of the clamp is available. Don't, for heaven's sake, worry about having oddly shaped pieces left over; you'll use these in oddly conceived projects.

Mounting the Transformer

In order to fit the mounting tabs through the board, it is necessary to enlarge holes to accommodate them. For this purpose, I used a No. 44 drill bit in a hand drill (electric is fine -- doesn't take very long, that's for sure). In the case of this transformer, the tabs are 0.8 inches apart. Therefore, I drilled out the following holes:

Picture the board from the component side with the short dimension toward you (its length being perpendicular to the edge of the table). The nearest two-thirds of the board will eventually have the speaker glued on it, and will be left blank, as far as circuitry is concerned. The transformer is placed near the upper right corner -- its three primary leads pointing to the left. In a column of holes, four in from the right-hand long edge, I drilled out the second hole down from the top, and the tenth one down from the top. Fit the transformer's tabs into these holes, being careful not to trap any leads under it; turn the board over and bend the tabs outward with a screwdriver.

What would you do if your transformer were 0.75 inches wide and didn't match the holes? Well then, feeling for holes with its tabs, turn it until you find a likely match. You heard me, mount it askew, thus violating Rule 5.

Inserting the Transistor

You are indeed lucky if your transistor is a 2N2905; the leads emerging from this unit (a TO5 package) match up with holes in this pattern of board. Those of us with 2N2907's and 276-2023's have a problem; these leads are too close together. Leads of smaller transistors have to be fanned out into a slightly larger triangular pattern. This also means that the small units cannot be pulled down properly against the component side of the board. (Any attempt to force them down against the board will cause leads to short out against the rim of their metal can.) Students of the Wiring Game should take note that this must constitute a minor violation of some kind -- perhaps bending Rules 2, 4, and 5.

There seems to be a certain logic in putting the transistor adjacent to the leads of the primary winding; we'll orient it so that its emitter pretty much lines up with the center tap. Therefore, the collector and emitter leads will go into holes in the fifth row from the top (the board being oriented as before, with the transformer at the upper-right corner).

Note that, with the leads of the transistor pointing straight down (and the tab being at the lower right), the order is, from left to right: collector, base, emitter. The collector will go into the fifth from the top and sixth from the left; the base will be in the fourth from the top and seventh from the left; the emitter will go in the fifth from the top and the eighth from the left.

Once through, bend the emitter lead in the direction of the transformer. For reasons that will soon become apparent, bend the base lead straight up toward its edge of the board. Bend the collector lead toward you. Rocking the transistor in various directions (so as to encourage each lead to lift off the wiring side of the board), bend them all firmly down against the wiring side.

Without cutting it short, poke the center tap lead down near the emitter lead -- say, the fourth row down and ten from the left. Bend its tinned end down over the emitter lead and solder it there.

Students of the Wiring Game might think that this lead should be cut short so as not to leave a big loop on top of the board. Not so. In the first place, this lead has been nicely tinned by the factory; why mess it up. In the second place, if you ever salvage parts from this project (or repair it), you'll thank yourself for retaining this lead's length; such a loop is actually called a "courtesy lead." Thirdly, the connection will stay in place while you're soldering if you put the loop to good use; clip it down against the edge of the board with an alligator clip, thus keeping the elusive little blighter in a stable position.

Installing the Capacitor

It makes good sense to put the 0.1uF capacitor near the top edge of the board, between the transformer and the transistor's base lead. Let's put it into holes in the second row from the top. The distance between its leads depends on what kind you've got. Mine happens to be a fairly large one, so its leads end up in the eighth and thirteenth from the left.

Bend the two capacitor leads straight to the sides. One will cross the base lead, to which it is then soldered. The other awaits a transformer lead. You can plug the transformer lead into any hole to the right of the capacitor -- in the third row down. (Mine is in the third down and the fourteenth from the left.) Bend its tinned end down so as to cross the capacitor lead and solder it there.

As soon as this latter transformer lead has been soldered, cut off any excess capacitor lead that reaches under the transformer. Do not, for a reason to be seen, cut the capacitor's other lead after it crosses the base.

Placing the Resistor

I placed a nice large 1/2-watt unit in a column near the edge of the board. First, bend the resistor's leads at right angles to its body; its body will ly flat on the board when it is installed. In the case of my 1/2-watt unit, the top lead goes into the top row and third from the left; its bottom is in the sixth down and third from the left. Use of a 1/4-watt resistor would put its lower lead in the fifth hole down.

Cut the upper resistor lead so that it protrudes only about one-fourth inch; bend it over across the capacitor lead and solder it there. Why cut the lead? Well, if you don't, you'll stand a fair chance of soldering it to its other lead farther down in the column, that's why.

Installing the Battery Snaps

Now, if you had a continuity tester, you could tell which lead is the positive and which is negative. This is done by clipping one of your future tester's leads to a snap -- say, for example, the small snap, which is the negative -- then touching the wires with the other test lead until you find the one that beeps. For now, however, I guess you're stuck with getting sighted help.

Poke the black lead (the negative) into the board on the transformer side of the transistor's collector lead -- say, six down and seven to the right. Bend it over to cross the collector and solder it there. Somewhere near the lower portion of the transformer's frame, poke the remaining primary lead and the red battery lead into holes next to each other; solder them together on the wiring side of the board. (Do this high enough up so as to leave room for the speaker.)

The above two connections sound easy. Actually, because of the wire sizes (smaller than I care to mess with), these procedures are tricky. You may need to strip additional insulation and re-tin the wires, especially on the battery leads. (Try to leave the transformer lead alone; if you break it, you'll have to replace this item.) When you do get the tinned ends to meet, their ephemeral ends will not stay in contact unless their insulated portions are held for stability. They can be wrapped around the capacitor, for example, or they can be clipped to the edge of the board in the jaws of an alligator clip.

Test-Lead Connections

Eventually, the test lead jack will be tethered to the board by two 5-inch lengths of hookup wire. For the moment, however, I have a trick for you. Let us install a 10-inch loop of wire between the desired connection points on the board. We'll cut this loop and insert the jack later, after we know that the tester works. In this way, our continuity tester will test itself with this "short circuit" we have provided.

Strip and tin the ends of a 10- or 12-inch piece of stranded hookup wire. Plug one end into a hole next to the collector, perhaps seven down and five from the left; bend this over so as to cross the collector and solder it. Plug in the other end near the edge of the board, so that it can be attached to the free end of the resistor -- say, seven down and the second column from the left.

Installing the Speaker

Eventually, the speaker will be glued to the board on the component side. It may seem like a good idea to do this first, then connect it up; this would tend to establish a stable situation. On the other hand, I prefer to make these connections first; I find it appealing to have the freedom of orienting the speaker so that each terminal, in turn, faces me.

However you decide to do it, the idea is to put the tinned end of each wire through the hole in its respective lug, bend it double -- firmly -- and solder it. Naturally, the two secondary leads go to the two lugs on the speaker.

Preliminary Testing

At this point, we should be able to find out if this project works. Connect a 9-volt battery to the snaps; the instrument should emit a high steady beep. If this beep is not steady, you have a bad solder connection. If it doesn't beep at all, there could be one of the following reasons:

The wiring could be correct, but with a connection or two not being completed. Wiggle and jiggle each connection and listen carefully.

As far as incorrect wiring is concerned, the most likely suspects are: reversal of the battery wires, having the center tap exchanged with one end of the primary, or having inserted the transistor incorrectly (turned 60 degrees).

In order to change anything, reheat the offending connection and pull on one of the parts. Whatever you pull on can be extracted from the board in this way. Before frustration overtakes you, go have a cup of tea, and read the final section here called "Commiserations."

Wiring the Jack

After the oscillator has proven itself, you can use it to show you how the jack is to be wired. First, it would be good to discuss the plug and the jack, and to brace yourself for the variations you might encounter.

If the jack you have is of the "open-circuit type," it will have two terminal lugs -- plus the springy contact that meets the tip of the plug; once you have identified either one of these, belonging either to the tip contact or the sleeve, you are set. The plot gets thicker when you have a "closed-circuit" version; an extra contact senses when the tip contact has been forced aside by the plug. With closed-circuit jacks, the sleeve is easy to identify with your tester, but the other two lugs look the same electrically until a plug has been inserted. The way to distinguish between the "tip" and the "switch" is via the plug, which is described below:

The "handle" of the plug, called the "connector shell," can be unscrewed. Inside, you will see a short stubby solder lug (which goes to the tip) and a longer stave-like metal strip with a solder-lug hole halfway along it and a cable clamp on the end. Given the cable clamp, it makes sense that this latter item is a connection to the sleeve of the plug (this sleeve usually going to the shield of a cable).

With the plug inserted into the jack, the tester will indicate when there is "continuity" between the cable clamp and one of the lugs on the jack (the lug intended for connection to the sleeve). By the same token, continuity will be seen between the plug's and jack's tip connections. Of course, this procedure will work with the open-circuit jack, but it is not necessary to utilize the plug; you need only touch one test lead to the body (actual sleeve) of the jack, fishing around with the other until a beep indicates that you've found its lug; the remaining lug is for the tip contact.

Going back to the circuit board, cut the 10-inch loop of test-lead wire in half. Strip and tin these ends. Connect the battery, and your tester is ready for business (it should not beep when these leads are separated).

Hold one of your tester's leads against the body of the jack; fish for the lug that beeps with the other tester wire. This lug is the sleeve connection; solder it to the wire that comes from the transistor collector. With your alligator clip lead, short out the plug -- clipping one end onto the short lug, and the other onto the cable clamp. Insert this shorted plug. With the remaining tester lead, fish for the lug on the jack that makes the tester beep. Solder this wire, which should be the one coming from the 4.7K resistor, to this lug.

Preparing the Test Leads

While the soldering iron is hot, we might as well do this now, even though it could very well be left 'til later. Cut your alligator clip lead in two; strip and tin the ends. If you have a piece of spaghetti tubing, pass one of the leads through it (to mark it as "positive"); if not spaghetti, wrap it in a small bit of tape directly behind the alligator clip. Pass these leads down through the cover of the plug, before you forget to do so.

Insert the marked "positive" lead through the hole in the "tip" lug and solder it. Pass the "negative" (unmarked) test lead through the hole in the longer sleeve terminal; make this insertion from inside out -- inserting just above the positive lug. You'll have to make three-quarters of a turn around this member with the tinned end because of the "stave-like" shape of it. Solder this lead as well.

Cut off any excess wire, and tentatively screw down the cover over these connections. With the clips separated, plug this assembly into the jack; it should not make the tester beep (if it does, the plug is shorted). Touching the clips together should make the tester beep. If so, tighten the cover as best you can with your fingers, and congratulations, you've done everything right.

It would be nice if the cable clamp could be used to take stress off these connections. You can try wrapping the wires in a bit of tape immediately above the solder connections -- then squeezing the cable clamp around this assembly with needle-nosed pliers. Usually, however, the cable clamp is far too small to do any good; students here usually just forget about it.

Mechanical Assembly

If you haven't already done so, glue the speaker in place. Since the board is porous and the rim around the speaker cone is cardboard, white glue or animal glue will be sufficient in this case (this is usually not true). Make sure that the speaker is confined to the boundaries of the board -- otherwise, the board will not rest properly in the box, or there will be insufficient room for the battery.

Now it is time to ravage the box. Whether done randomly or by one of the methods described in "Hints and Kinks," drill some speaker holes in the bottom of the box, centering them at a distance of about 2-1/8 inches from the "battery end." Next, lay the board in the box with the speaker nearest the battery end, and find a nice clear spot for the jack. (I put mine on the opposite side from the transformer, drilling its hole 2-5/8 inches from the battery end, and 5/8 of an inch up from the bottom.) There are lots of places where you cannot put the jack; you cannot place it where it will interfere with installation of the battery, or where it will run into the transformer.

As far as drill sizes are concerned, I use a No. 43 for the speaker holes. In drilling for the jack, which requires a 1/4inch hole, I first use the No. 43 to make a pilot hole, then go to the No. 1 bit for enlarging it.

In drilling for the jack, your intention should also be to miss those pesky ribs inside the box (which are intended for holding circuit boards on edge). If your hole came too close to one of these ribs, so that the jack cannot rest comfortably as the nut is tightened, it is easy to remove the offending ribs. The Lexan out of which the cabinet is made is very soft. With a sharp knife, for example, you can calmly "pare" the ribs completely off. Another way is to nip them off with diagonal cutters.

Remove the nut and washer from the jack and put the bushing through the hole from the inside. Now, you have a decision to make. With the washer underneath the nut, there may not be room for the nut to screw down past the end of the bushing, thus interfering with the fit of the plug. Try putting the nut on with the washer first; if there is any doubt, leave the washer off. (Don't throw the washer out; save all hardware for future use.)

There are many ways to hold the board in place. Screwing it down is certainly possible; for doing so, you will need to get spacers to hold the board away from the box slightly, so as not to mash circuit connections against the box and, probably, bend the board. You can also obtain doublesided foam tape from a neighborhood hardware store, securing the board under the transformer and at the corners. I think it is sufficient to rely on a block of styrofoam which is glued to the lid of the box.

By gluing a piece of styrofoam to the lid of the box, you can solve two problems: If you position it right, this will keep the battery standing up on edge. By doing a little creative carving, you can accommodate the magnet of the speaker, and thus hold down the board. The best source of styrofoam of different shapes comes from packing material in all the little electronic gadgets we buy. (I mean the funny-shaped block of styrofoam, not the peanut-shaped, general-purpose packing material.) The best thing to cement it down is Barge Leather Cement, gotten from leather-goods stores or shoe repair shops.

Hints and Kinks

Speaker Talk

It is indeed a shame that 95% of my projects have loudspeakers in them, since making sound holes for them and mounting them is borrrrrring! Nonetheless, my methods are successful, and the projects look very nice.

The quickest way of getting sound holes is to punch a large round hole with a "chassis punch," then cement plastic screen or perforated metal stock behind this hole. However, a selection of large-sized chassis punches will mortgage the farm, and the resulting package is not nearly as attractive as a method I learned from Gene Merritt, a fellow subscriber to the ol' Braille Technical Press.

You can use Vectorbord as a template for drilling nice regular patterns of holes; the "A" pattern (available from Jameco as 64AA18) is well-suited for this purpose. This pattern has holes spaced at 0.265 inch intervals. You lay a piece on the panel to be drilled, and drill through the holes with a No. 43 bit (this enlarges the holes in the Vectorbord, but does not ruin it for use the next time).

Of course, the pattern I choose to drill depends on the size and shape of the speaker; I decide on this by laying the Vectorbord on the speaker first, counting the speaker's dimensions in "Vector holes" (with the aid of a braille stylus).

Most often, for round speakers, I decide on a diameter first; then, when I drill the pattern, I leave out corner holes as appropriate. I keep three No. 4-40 machine screws and nuts handy; I drill a center hole and bolt the board down, then I drill two edge holes and secure these. Then, I go after the project until my wrist gives out and the neighbors complain.

The pattern I drilled for the continuity tester is basically 5 by 5, but with the corner holes omitted (thus having three holes in the center of each side of the pattern). A square pattern of 4 by 4 would be perfectly OK -- this represents 16 holes already.

When I don't have any such Vectorbord (it does wear out, eventually), I put Dymo Tape in a braille slate upside down (so that the dots become dimples), whereupon I write a generous supply of "K's." Laying down a few lines of K's works OK, too. You then drill through the dimples; then throw away the tape.

No less tedious than drilling is cleaning up the holes afterward. I use a 1/2-inch drill bit for this purpose, spinning it in each hole to clear away the burs. If the box is a solid color, you can do this on the front side as well. However, if the box is painted, sighted folks don't like the results of deburring the holes on the outside. Even though they feel raggedy, leave them alone in this case.

Very often, the speakers you get nowadays have no mounting flanges. They are intended to fit cases which have been molded for them (which is a real great favor to us). I have had good results from gluing them to the panel if I use the right cement and do a little preparation first.

The best thing I have found for gluing speakers is a Goodyear product called Pliobond Cement. (There are good equivalents; for example, Plibond by General Cement.) This cement is rubbery stuff; if you slop a little over onto the speaker cone by mistake, it doesn't "stop the show," as you might say (it is meant to be pliable).

What generally fails in bonding the speaker is the cardboard rim. More often than not, when you open a project whose speaker is rattling around inside, portions of the cardboard are still stuck fast to the panel. I have greatly improved the reliability of my speaker-gluing jobs by sanding the cardboard rim down flush with the metal rim. I lay a piece of coarse or medium-grade sandpaper on a flat surface; then I rub the speaker vigorously, face down, on the sandpaper until I hit the metal rim. After this, the cement can bond to the cardboard and the metal, both. It is also a good idea to sand around the pattern of holes on the back side of the panel, thereby roughening the surface and improving the bond.

In gluing, I first clean the back of the panel and the rim of the speaker with isopropyl alcohol. Then, I get the rim of the speaker gooey with the cement. I never bother to apply glue to the panel, even though they always say to wet both surfaces.

Make sure you know exactly where you want the speaker before you set it down, and turn it so that the terminals face the way you prefer. If you set it in the wrong place, you will have to apply fresh glue before setting it down again. Once done, turn the project over on its face, letting the speaker's weight keep the pressure on, and split the scene for a day.

Applying Glues and Cements

I learned long ago that the finger is the best applicator of glue (which is why I stay away from "SuperGlue;" this can brutally stick yourself to yourself). I have heard respected sighted people in the piano business advocate using the finger; it is even more sensible for a blind person. When you use your finger, you know just how much glue is being applied, and you have a pretty good idea as to where it's going.

The trick is to keep appropriate solvents around to clean yourself up. For white glue and animal glue, a handy bowl of warm water is marvelous, along with a couple of heavy paper napkins. For PlioBond and Barge Leather Cement, paint thinner is OK. (PlioBond can be reconstituted with methyl-ethyl ketone.) When working with epoxy, a bowl of isopropyl alcohol is appropriate. Keep everything ventilated when you use these organic solvents, since they're rough on your endocrines.

You will be less annoyed with gooey fingers if you use a non-dominant finger to apply glue. I usually use the middle or ring finger of my left hand. Wrap any rings you have with masking tape (which is unnecessary when using white glue).

Before I glue anything, I always clean the surfaces with isopropyl alcohol. The enemy of all glue bonds is oil, either from your hands or from industrial processes in manufacturing. Alcohol is a moderately good solvent for these oils, and it is not fatal to breathe the fumes (within reason). In the process of cleaning the surfaces, you will also clean your fingers, which will prevent you from contaminating the glue as you apply it.

Miscellaneous

A person cannot say everything in a paper like this (although there are readers among you who accuse me of trying). There are things that a shop-worn person does that "do-it-yourself" books couldn't cover, nor would the reader remember them with screwdriver in hand. However, here are some "big ones."

The project cabinets you buy are held together with self-tapping screws (unless they are of the Hammond brand from Canada, these being of superior quality). The material in the box described here is pitiously soft, and the screws brutal; without realizing it, you can tap new threads every time you insert a screw. Therefore, it is essential that you learn to first turn each screw backwards, feeling for it to "drop" into its threads. Then, try turning the screw forward using the shank of the screwdriver, not the handle. If the screw resists even slightly, say "excuse me," back up, and try again.

In drilling holes in precise locations, you usually cannot get away with just placing the bit on the surface and pulling the trigger -- the bit will wander before it begins to cut. In metal, you first place a "center punch" in the desired location, whereupon you bash it with a hammer two or three times. In soft plastic like this, I keep a stray drill bit, perhaps No. 45, handy; I place this in the desired spot and spin it between my fingers, thus drilling a "center mark."

Also, in order for a drilled hole to "be where you put it," it is necessary to drill a pilot hole first -- a hole much smaller than the eventual size. Then, if it is critical, or if the material is hard, drill several times with larger bits, doubling the size of bit each time.

Keep scrap slabs of wood handy for drilling on, rather than boring into your workbench when the bit goes through the work. The slab of wood will also catch most of the "chip" (the shavings you get when you drill); you can then just up-end the drilling board into the wastebasket.

Deburring of drilled holes is always necessary, especially on the back side. A bit much larger than the hole is very handy for this. If you cannot reach the back side of the hole with a drill bit, the burr can be cleared away with a small knife blade that you don't particularly care about.

There are some nifty measuring tools for blind folks. AFB (15 W. 16th St., NY, NY 10011) is reviving their "Rotomatic Rule;" this is a calibrated screw, having sixteen turns per inch. Almost exclusively, I use two measuring devices: the "Stanley Combination Square" (also from AFB) and the "Rotomatic Rule."

I put the Stanley Square up against one side of the cabinet, holding it with my ever-so-adjustable tummy, and bring the Rotomatic Rule to meet it from another side. Then, I place my center punch or finger drill at their intersection, let the rulers fall as I turn my attention to holding the centering device, and make my mark.

By now, it will have become obvious to you that "holding clamps" and vises are a key to happiness. If you're stuck with the kitchen table as a work space, at least get a small "clamp-on-the-edge" vise from a hardware store, as well as some sort of board clamp (see "Soldering, Part II," SKTF, Winter 1981).

Speaking of holding things, the hardest solder connections to make are those in which the parts are poorly stabilized. For example, soldering hookup wire is nearly impossible, if everything you do knocks it out of position. Get good at figuring ways of stabilizing such wires. As mentioned earlier, you may have room to clamp the insulated portion against an edge of the board with an alligator clip. Wrapping the wire around the board, or around an item on the component side, is something I do most often. Wrapping the wire around the board clamp, or trapping it underneath, works too.

In principle, desoldering something is quite simple; you merely reheat the items and pull them apart. Since the parts you're working with will get hot, pulling on them with your fingers can be uncomfortable -- and awkward. The best pulling tool made is a set of "locking forceps" -- something you can get from your doctor, or from tool suppliers (see "Soldering, Part II"). An alligator clip sometimes works, as well as needle-nosed pliers with a rubber band around the handles to squeeze them closed.

When you pull a wire free of a solder lug (such as those on the speaker and the jack), solder often clogs the hole, preventing you from inserting a wire next time. There are three ways to handle this. One is to ignore it; wrap the lug around the outside next time and solder it that way. Another is to clear the hole with a pointed instrument; an old dental pick is the best thing for doing this. The third way is to heat the lug, then slam or toss the part down on the table.

Commiserations

Suppose your project doesn't work, and the meager suggestions given here don't point the way toward fixing it. There are three things you can do. You can ask a sighted pal to help you trace through the steps again. You can quietly analyze which are your weak points, get some junk parts and practice these operations on a scrap of board. Or, which is what I recommend, you can gleefully take wire cutters to the thing, clean off the board with isopropyl alcohol and an old toothbrush, buy new parts and do it again.

Do you realize how little money this first attempt cost you? First of all, the board and the box are still OK, and you didn't have a chance to ruin the jack yet (you'll have a turn at this, too; I've sent many to the great beyond). Therefore, you've had hours of beginning lessons for under five bucks. (Your kid sister could be dropping $10,000 a year on her Bachelor's degree.) Truly, no matter how it's turned out, your initial experience has been quite economical.

Any negative interpretation of your experience is artificial. Realistically, let us take stock of your individual accomplishments: you have installed four parts into holes that are so small and close together that they can only be clearly felt with your tongue (don't do much of that, by the way, since you'll get tiny fiberglass slivers in your tongue). You've made ten solder connections, and have tinned possibly four wires. Except for chips and diodes, you are now intimately familiar with the most common parts. I'd say, for five bucks, you got a pretty good deal.

In your next trip to the store, buy two sets of parts. Any you don't use here will not go to waste; variations on this circuit abound in audible devices. So what if you have to start over three times. When you're done, you will have a vital instrument. Your sighted counterpart, who builds a radio circuit or a photo timer, only ends up with that toy, and not an instrument of daily use. Then, too, the person whose project worked the first time has less experience at troubleshooting, and he is behind you in the practice of manipulating parts.

What's Next

First, now what you have one, you should read "Continuity Tester Uses," SKTF, Fall 1983. Just for fun, build a couple of the "attachments" for it that are mentioned there. Clip it across the resistors in the "attenuator" of Part I in this series; the two resistors will give you very different pitches.

What's happening next time is a test amplifier, such as the one described in "Point-to-Point Wiring on Perforated Board," SKTF, Fall 1980. If you want to get a head start, go ahead and review that article.

The one to be presented will have three main circuit changes: The regulated power supply will be omitted; a 9-volt battery will take its place, this battery being bypassed by 250uF (negative of the capacitor at ground). The output capacitor to the speaker will be changed to 100uF, thus making the project a little more roomy. Finally, a volume control will be included; its bottom end is grounded, its arm goes through the 0.1uF input capacitor to pin 3 of the LM386, and the top of the control will accept the input signal (via another pesky jack).

A TRANSISTORIZED WIDE-RANGE, CRYSTAL-CONTROLLED TEST OSCILLATOR

By Bob Gunderson, W2JIO

Abstract

The test oscillator described in this article is one of the handiest gadgets for use in the ham shack or laboratory. It will oscillate from very low frequencies -- below 50 kilocycles to well over 100 megacycles. I use it as a calibrator for my frequency counter, as a signal source for receiver alignment, as a means for checking crystal frequency and crystal activity, and even as a marker generator.

If you plan to build this gadget, you'll surely want to begin collecting surplus crystals; you can easily attach Dymo Tape labels so that you will know each crystal frequency. My collection includes a low frequency of 50 kc, and many 20 mc units, plus a number of overtone frequencies. Since the test oscillator is a Pierce circuit, these overtone crystals produce their fundamental output which is some value around one-third the marked frequency, though not an exact third.

The oscillator is fitted with an auditory gimmick circuit, so that the operator can tell if the crystal is oscillating. A 2-pole 3-position switch makes it possible to disable the output indicator when not in use. In addition, I have used a rechargable 9-volt Nicad battery and a built-in charger. An 18-volt transformer feeds into a bridge rectifier with a limiting resistance to hold the charging current down to 10 mA. The transformer is the type used with the Princess Telephones, and is outboard -- attached to the oscillator by means of an RCA type phono connector. It simply plugs into the AC line.

The oscillator is a simple Pierce Circuit, using an insulated-gate MOS Field-Effect transistor, type 3N139. However, any good FET should work quite well.

I should mention that there are numerous crystal holder pin configurations, so that it would be advisable to place various types of crystal sockets on the panel. I have four, plus an ordinary 5-pin socket on my unit and this pretty well accommodates most types. In addition, I made up a 5-pin plug with two flexible leads connected to pins 2 and 4, with clips at their far ends so that odd crystal units may be clipped into the circuit. These parallel-connected sockets will provide some additional capacitance, but it doesn't appear to have much effect.

The Pierce circuit is essentially a Colpitts oscillator in which the feedback is a function of the input and output capacitance (from gate to ground, and from drain to ground, respective ly) providing the necessary feedback. At the low frequencies, additional drain-to-ground capacitance is needed, and I have provided three toggle switches so that the necessary additional capacitance can be placed in the system.

Caution -- When you use an oscillator of this type, you will probably use some of the older crystals which are comparatively large, meaning that their output amplitude is quite high. If the positive-going swing exceeds the breakdown voltage rating of the transistor (from gate to source), the FET will be destroyed. I quickly learned this fact, and found it necessary to add a diode clamp from the gate to the source and ground. Of course, there is another option -- you may add a source biasing resistance, properly bypassed, to limit the positive swing. I decided in favor of the diode clamp because of my limited supply of 9 volts. The source bias arrangement would, no doubt, work just as well.

My original unit -- the one built some years ago, and used until the MOS transistor went West, used the original transistorized gimmick which I devised back around 1954. However, Earl Quay, W4MKC, recently sent me a simple gimmick circuit which is essentially an assymetrical multivibrator, and I have included it in the rebuilt unit. This newer circuit is simple and works very well -- more of this when we describe the circuitry. One day when I attempted to show the unit to a group of radio amateurs, I found that it didn't work, and upon checking, found that the FET had burned out. The replacement lasted about 10 minutes, and it was then that I decided to add the diode clamp. Measuring the negative-and positive-going output voltages at the drain showed that the positive signal reached something like 17.5 volts peak, while the negative signal was only 5 volts. This occurred with the larger quartz crystals. These measurements were taken, using a small-signal germanium diode (a type 1N270) working into the vacuum tube voltmeter. The two halves of the signal can be measured by reversing the diode leads, and reversing the input polarity of the VTVM. The addition of the diode clamp yields an output signal which isn't symmetrical, but the positive excursion is only something like 7 volts rather than 17.5; the negative excursion is still the same.

As mentioned previously, the trick in making a satisfactory wide-range oscillator is the use of sufficient feedback from the amplifier's output to its input; this has been taken care of through the use of 3 separate feedback capacitors in the output circuit. These may be switched in separately, or they may be placed in parallel to give the sum of the 3 capacitors. The second essential part of the circuit is sufficient impedance in the drain of the FET. For the higher frequencies, a 2.5 millihenry RF choke is used, and for the low-frequency end, a large 80 millihenry unit provided the necessary increase in drain impedance. These two chokes are connected in series, with the smaller value directly at the drain. At low frequencies, this 2.5 mH choke hardly exists, and the larger unit does the work.

The output of the RF oscillator is coupled in two directions -- to a coaxial connector for application to the receiver, counter, etc., and also to the rectifier for forward-biasing the auditory gimmick when the aural indicator is to be used. I have thought that it might even be a good idea to couple the output to a source follower, or an emitter follower, with a potentiometer as its load, so that an attenuated signal could be available (we'll leave that to the new and improved version built by another reader).

The layout of parts isn't at all critical. This unit is built into a Keystone bakelite box measuring 2-1/2 inches high by 6-3/4 inches long and 5-1/2 inches front-to-back. These are outside dimensions. It might be better to place the unit in a metal container to provide better shielding. However, I use the oscillator for direct pickup on the station receiver, so that the unshielded construction is ideal. The panel is aluminum stock, with the crystal sockets at the left end; the output coaxial connector at the upper center, with the on-off-function switch at the lower center. The 3 toggle switches for switching the capacitors into the feedback circuit are just to the right of the crystal sockets. The connector for the power (charger) transformer and the speaker are at the right side of the panel. The greater part of the electronics are mounted on a small piece of perforated Vectorbord measuring about 6-1/4 inches long and 2 inches wide, mounted vertically along one edge of the panel, in from the lower edge, far enough to allow the unit to fit inside the plastic box. The circuit board is fastened to the metal panel with small brass angle brackets. When the 3-position switch is "off," the charger is connected to the 9-volt battery for recharging the unit.

Pierce Oscillator Portion of the Circuit

All crystal sockets are connected in parallel, and one side of this arrangement goes to the gate of the oscillator transistor (3N139); the gate also goes through the parallel combination of 33 pF (mica capacitor) and 100K ohms, to the source, to the circuit ground and to the minus 9-volt line. This gate also goes to the anode of a 1N914 small-signal diode, with the diode cathode also grounded. (This is the clamping diode used to limit the positive output swing.) The source and substrate of the FET are connected together and grounded.

The other side of the parallel-connected crystal sockets goes through 0.01uF to the transistor drain, with the drain going through the 2.5 mH RF choke, thence through the 80 mH RF choke to the plus 9-volt line. This is a switched line which will ultimately connect to the 9-volt battery positive. The drain also connects to a 100 pF mica capacitor, with the other side of this capacitor going to the hot side of the output coaxial connector (SO-239). Of course, the other side of the output is connected to the panel common.

The drain connects to one end of three additional mica capacitors (350, 500 and 1,000 pF); the other ends of these capacitors each go to one end of their separate toggle switch. The other ends of all three toggle switches are connected together and go to the circuit ground. With this arrangement, any one of these capacitors, or all three, may be connected from the transistor drain to ground.

This completes the oscillator, except for the coupling arrangement used to rectify the oscillator's output, which biases the auditory gimmick -- more of this later.

If you wish, the diode clamp may be omitted, and you can apply source resistor bias to the circuit to provide the necessary protection for the insulated-gate FET. The only circuit change consists of omitting the diode from gate to ground, and connecting a resistance of perhaps 470 ohms, suitably bypassed by a 0.01uF disc ceramic capacitor between the source and ground. The remainder of the circuit is the same. I chose the diode clamp because of the 9-volt limit.

You will find that some crystals are a bit sluggish, and may not oscillate unless additional capacitance is switched into the drain-to-ground circuit; the 3 switches will pretty well take care of this problem. You may even add a pair of terminals for adding even more capacitance if you wish. One of these would be grounded, while the other would connect to the drain. You can even use a capacitance decade at these terminals.

Output Indicator

A small capacitance (100 pF mica) couples the oscillator to a shunt diode rectifier, and the output of this diode provides the DC to forward-bias the auditory gimmick, so that it produces an audio tone when the crystal is oscillating; no output will indicate that the oscillator isn't functioning. My original design used the earlier gimmick circuit, as mentioned previously; however, the circuit provided by W4MKC worked so well and required fewer parts, that I put it into this newer version.

Most of the readership is familiar with the simple transistor gimmick. Just to make sure, the circuit for it is described below:

Standard Gimmick Circuit

A PNP transistor is used in a blocking oscillator, using a small audio output transformer. The bias for this oscillator is provided by the output from an NPN transistor. The input of the NPN transistor is fed from a DC source -- from a meter movement, or in this case, from the output of the shunt diode. If a silicon transistor is used as the DC amplifier (the NPN transistor), forward bias should be applied to its base via a voltage divider, inasmuch as this type of transistor requires something like 0.7 volts before conduction can occur.

Circuit Description

Any general-purpose PNP and NPN transistors may be used. The emitter of the PNP oscillator transistor goes to the center tap on the transformer primary, with the bottom end going to the plus 9-volts. The other end of the primary goes through a 0.01uF capacitor to the base. This winding is tuned by a capacitor of about 0.01 to 0.05 uF. The secondary of the transformer goes to the speaker. Minus 9 volts is grounded. The base goes through a resistor (10K to 47K ohms -- value not at all critical) to the collector of the NPN transistor. The emitter of the NPN unit, and the collector of the PNP oscillator transistor, are grounded.

The input signal is applied to the base-emitter terminals of the NPN transistor. If this circuit is used around transmitting gear, the input leads should be suitably filtered by means of RF chokes and appropriate bypassing. If forward bias is to be applied to the DC amplifier, connect a high-resistance potentiometer across the 9-volt lines, and return the low side of the meter to its arm, rather than to the circuit ground. Then, adjust the potentiometer until the circuit just begins to oscillate.

The sensitivity of this circuit is quite dependent upon the value of the base blocking capacitor -- the smaller the capacitance, the greater the sensitivity. Values from 0.1 to 0.01 uF can be used, with 0.01uF providing a much greater change in pitch for a given input signal. The audio output level does decrease, however, with the smaller capacitance value.

W4MKC's Gimmick

This is really simple, and it takes less room on the circuit board. I haven't really played with this circuit, except as it applies to the test oscillator; however, it does look promising. It uses a PNP and an NPN transistor, both general-purpose types in an assymetrical multivibrator.

Circuit

The emitter of the PNP transistor goes to plus 9 volts. Its collector goes directly to the base of the NPN transistor. The NPN's emitter goes through the speaker to minus 9 volts. The NPN unit's collector goes through 10 ohms to plus 9 volts, and this collector also goes through a 0.005uF capacitor to the PNP transistor base. The PNP base also goes through a 470K resistor, thence to the anode of a 1N270 diode (1N34 will do), with its cathode taken to a point X, which I shall explain in a moment. The diode is shunted by a 1 megohm resistor. The signal from the oscillator is injected through a coupling capacitor to the anode of this diode.

If the diode cathode and 1 megohm resistor junction is returned to the circuit ground, this will forward-bias the PNP transistor, and the multivibrator will "free-run." Returning point X to plus 9 volts will hold the circuit in the idle mode until the diode provides the necessary bias. The return point for the resistor-diode junction at plus 9 volts may be by-passed by means of a 0.01uF capacitor. The plus 9-volt lines for both circuits are connected together and bypassed by a 1,000uF electrolytic capacitor.

Switching

The on-off switch is a 2-circuit, 3-position unit. The arm of unit 1 goes to the plus of the 9-volt battery, and the "off" position for this section goes to the charger output. Positions 2 and 3 of this unit go to the plus 9-volt line of the circuit.

The "off" position of the B side of this switch is left blank. Position 2 of unit B is connected to the plus 9-volt line; position 3 is connected to the junction of the diode (1N270) anode, the 1 megohm shunting resistor, and the series 470K ohm resistor which goes to the PNP transistor base. The arm of this second switching unit goes through a 100pF mica capacitor to the drain of the crystal oscillator.

It was necessary to use this switching arrangement because there was sufficient capacitance between the switch contacts to excite the auditory indicator with the switch in its second position; I wanted this position for just the crystal oscillator. The multivibrator output is a square wave, and it finds its way into other circuits, so that it is often a good idea to have just the crystal oscillator without the indicator.

The Battery Charger

If you plan to use the ordinary 9-volt battery, you may skip this portion of the project. However, the charger is very simple. I used a low-current 100 PIV full-wave bridge rectifier and the transformer used to operate the pilot light in the Princess type telephone. If you have room in the cabinet, you might wish to include the transformer. I ran out of real estate, so I took this approach. The transformer plugs directly into the wall outlet, and I fitted it with a length of zip cord and an RCA phono connector, plugged into the appropriate connector mounted on the panel. The connector is hooked to the AC input leads on the bridge, with the minus output lead connected to the circuit ground, and the plus connected through a current-limiting resistance (620 ohms in my unit) to the "charge" position (the "off" position) on the on-off switch. The secondary voltage of this transformer is 15-18 volts, and the charging current for the 9-volt Nicad is 10 mA, for a period of 14 to 16 hours.

My first unit used a standard 9-volt battery, but I often left it turned on, and had a dead battery whenever I went to use it.

One final note about the sensitivity of this second auditory indicator -- A small capacitor in the feedback circuit, from the second collector to the first transistor base, provides higher sensitivity than does a large one. I had originally tried a much larger value and had just about given up the circuit. Of course, the higher capacitance reduces the free-running frequency of the circuit; the 0.005uF capacitor seems to be a good compromise in this circuit.

Note -- The series capacitor connected between the drain and one side of all the crystals is not absolutely necessary; however, if the two leads are accidentally short-circuited, or if you have a short-circuited crystal, this places the 9 volts on the gate and on the diode anode. The 0.01uF blocking capacitor will have inductance above the resonant frequency, and this places a limit on the operating frequency. I have found that putting a 2-inch loop of wire in place of the crystal provides an oscillator frequency of about 100 megacycles.

Note -- Another interesting observation I have made is that some very low-frequency crystals appear to oscillate; however, the oscillation frequency is not the one marked on the crystal. Rather, it is usually a much higher frequency. For example, I have some crystals marked "76 Kc," and find that they oscillate at about 210 Kc. These low-frequency units are mounted to oscillate in their "sheer mode," or end-wise. Thus, the crystal plate may be very thin. An X-cut 50 Kc unit would indeed be a very large bar, if it were cut to oscillate at its "thickness" frequency. I also have a 16 Kc crystal, and when plugging it into this oscillator, I find that it does oscillate -- not at its marked frequency, but up around 100 Kc. No doubt, it is mounted in such a way (between its plated electrodes) that it has a thickness mode, however unintentional, allowing it to oscillate at this odd frequency value.

For the most part, however, the oscillator provides output at the specified crystal frequencies. If you use it along with a counter, you will have no difficulty. I have not been able to make this circuit operate at the very low frequencies, from 1 kilocycle up to 30. These sheer mode crystals are available for use in various clock circuits.

Parts List

Capacitors [Those used in the oscillator include: 33 pF, 350 pF, 500 pF, and 1,000 pF (all mica or polystyrene). An additional bypass capacitor (0.01uF disc or mica, etc. should be included if source-resistor bias is used).]

Coupling capacitors:

Diodes:

Resistors: (All resistors are 1/2 watt)

ransistors:

Miscellaneous:

Basing Diagram for the 3N139 Metal-Oxide Field-Effect Transistor

Hold the transistor with the case facing down and leads up, with the tab at the edge of the case toward you. Counting clockwise from this tab, the connections are: 1. drain; 2. source; 3. insulated gate; and 4. substrate. These leads are arranged in a square configuration.

I mounted them on the circuit board with small hollow terminals made especially for the purpose. These terminals fit into the holes in the board; they are then peened over with a special tool -- an automatic center punch, fitted with a suitable point which fits an anvil, allowing the terminal to make a snug fit with the top side of the board. The hollow terminal makes for convenient wiring on the under side of the board, plus allowing the transistor to be plugged into the holes and then soldered. These terminals, along with the punch and anvil, are supplied by Keystone Electronics, 49 Bleeker Street, New York, N.Y. 10012.

A GREAT GIFT IDEA FOR SOMEONE ELSE'S KID

Abstract

To go right along with the train whistle (SKTF, Summer 1984), this is a toy siren. It emulates the new-style electronic sirens; it has a button for ascending and descending wail, as well as a "fast-warble" feature. Circuit afficionados will appreciate the power FET on-off switch, as well as the fine example of "zener-diode DC coupling." True to my tradition, it's in the Summer issue, so you'll have it by Christmas.

Circuit Operation

This circuit uses a Signetics NE566V VCO chip (see SKTF, Winter 1985) to generate a nice-sounding triangle wave for the siren. An option is included to modulate this oscillator at a slow rate, so as to simulate the "warble" that many of the modern sirens can produce. The "warbler" is another NE566, producing a 3 Hz triangle wave. The main oscillator is fed into an LM386 amplifier.

The main pushbutton, which gives you the standard "wail," runs the VCO control pin (pin 5) by way of an RC circuit with different time constants for charge and discharge. The charging of a 5uF capacitor is taken care of by a 47K resistor which goes from its negative end to the top of the pushbutton. Across this capacitor is a voltage divider made up of 22K and 47K; this sets the VCO voltage to the appropriate level, and its total resistance, 267K, governs the discharge rate (determines the descent of the tone).

This main pushbutton also "refreshes" the "on" switch. Pushing it turns on a PNP transistor which, in turn, turns on the gate of an N-channel MOS/FET, the channel of which is the power switch of the device. A parallel RC circuit on the gate of this FET (consisting of 1uF in parallel with 220K) keeps the circuit alive in order to let the siren "run down" to sub-audible frequencies before shutting off the power.

The warbler is set up to free-run; it has its requisite voltage divider on the VCO pin made out of fixed resistors and everything (see sample circuits in SKTF, Winter 1985). As can beseen from the literature, there is an oddball bias on the output of the 566; this had to be accounted for when driving the VCO pin of the siren -- which requires funny control levels of its own. The warbler output and the siren's control pin are DC-coupled via a 6.8V zener diode, which just acts like a connecting rod of the desired length. (This use of zener diodes for coupling was pointed out to me by an instructor in college; it's a technique which can be forgotten if left to disuse. The catch is that zener diodes are awfully noisy, and they should never be used in low-level circuits.)

Whether or not the main switch is held down, the warbler switch seems to keep the bias on the transistor high enough, on the average, to keep the power switch alive in my unit. This may vary from unit to unit, depending on the transistor gain and the average value of the triangle wave being coupled by the zener. If it does not work quite this way, the skilled operator will soon learn to keep the main switch on when using the warbler button.

These oscillator chips really are meant to operate on volt ages over 10V; control of them is unpredictable under this value. For this reason, I chose the battery arrangement shown, which consists of a 9V battery in series with an AA-type penlight cell. Replacement of the AA cell need only be done for every second or third 9V battery.

It should be noted here that this project could have been designed around one of the commercial noise-maker chips that were made for use in arcade games (pinball machines). The thing was that these chips are as wide as the board onto which I fit my circuit, and the arrangement for hooking them up is not trivial. On the other hand, those chips are worth knowing about, and if I ever get any spare time, I'll describe one. (Why don't one of you ambitious builders write that article for us, yes?)

Circuit

A 9-volt battery is connected in series with a 1.5-volt penlight cell to get 10.5V. The positive terminal of the resultant battery goes to the VCC line. The negative terminal of the battery goes to the source of a power MOS/FET (Siliconix VN0300M or VN10KM, or any N-channel unit). The drain of this FET goes to project ground. Between the gate and the source is the parallel combination of 220K and 1uF (negative of the capacitor at the source). This gate also goes to the collector of a PNP silicon transistor (2N2907). The emitter of this transistor goes to the VCC line. Its base goes through 470K to one side of the main pushbutton (SPST), with the other side of this pushbutton going to the negative side of the battery (not to ground).

On the NE566's, the following power (and other) connections are made: (I put them both into a single 16-pin socket; this is why I left out the power supply bypass on the warbler, which has its pins 1 and 8 immediately adjacent to pins 4 and 5 of the siren chip.) Pin 1 of both is grounded; pin 8 of both goes to VCC. Between pins 1 and 8 of the siren chip is 0.1uF, located close to the chip. As per instructions from the data sheet, there is a 0.001uF disc capacitor between pins 5 and 6 of each chip.

Siren Oscillator

Pin 7 of this 566 goes through 0.1uF to ground. Pin 6 goes through 4.7K to VCC. Pin 5, the control pin, goes to the junction of a voltage divider; i.e., pin 5 goes through 47K to VCC, as well as going through 220K to the negative end of a 5uF capacitor. The positive end of this capacitor goes to VCC. The negative end of this capacitor also goes through 47K to the top of the pushbutton (the side of the switch that also drives the transistor base).

Warbler Circuit

Pin 6 of this 566 goes through 18K to VCC. Pin 7 goes through 5uF to ground (negative at ground). Pin 5 goes through 470K to ground, as well as going through 47K to VCC. The output, pin 4, goes to the anode of a 6.8V zener diode; the cathode of this diode goes through 4.7K to VCC. This cathode goes through another SPST pushbutton switch (warble switch) to pin 5 of the siren.

Audio Amplifier

Pins 2 and 4 of an LM386 are grounded. Pin 7 is bypassed to ground by 5uF (negative at ground). Pin 6 goes through 10 ohms (1/2-watt) to VCC; pin 6 is bypassed to ground by 100uF (negative at ground). Between pins 4 and 5 is a 0.1uF capacitor. Pin 5 goes to the positive end of a 47uF capacitor, the negative end of which goes through the speaker to ground.

The output of the siren, pin 4 of the first 566, goes through 220K, then through 470K to ground. The junction of these two resistors goes through 0.1uF to pin 3 of the 386.

It was a tight squeeze, but I built this animal into a box measuring 4.75 by 2.56 by 1.4 inches (Radio Shack 270-222). The circuit board is just over 1.2 inches wide (thereby being twelve holes in width), and it stands on edge along one side of the box. If I had it to do over again, I believe I would try a larger board lying flat on the bottom, and covering about 2/3 the length. The way I did it just barely leaves room for a 2-inch speaker on the bottom of the box. The pushbuttons are on the opposite side from the board, and the batteries are mounted on the lid.

As far as board layout is concerned, the two NE566V's are near one end, the large RC circuit and the on-off switch components extend slightly beyond the middle, and the amplifier is at the other end. A single 16-pin socket accepts both 566's, and this socket's long sides are parallel to the long edge of the board. The 8-pin amplifier socket has its rows of pins perpendicular to the long edges of the board.

Parts List

Capacitors:

Resistors (1/4-watt 5%, unless otherwise specified):

Integrated Circuits:

Miscellaneous:

[Actually, with purposefulness, the thing isn't very loud. As designed, this is no instrument of terrorism.]

THE IN'S AND OUT'S OF THE JAMECO CATALOG

by Susan Fowle, NY6D

[Editor's note, August 2005: While much of this article is dated, it does show the format of a typical electronics catalog and how to give instructions to get around a standard electronics catalog.]

Abstract

The goal of this article is to familiarize you with the general contents of the Jameco parts catalog, their ordering forms and procedures, so that you -- together with an electronically naive reader -- can order parts. Or, if you order by phone, you will know approximately what is available and how to present your order.

Basic Rules and Procedures

The address and phone number are:
Jameco Electronics
1355 Shoreway Road
Belmont, CA 94002
415-592-8097 (M-F, 7:00-5:00 PST)

Jameco's index is on the inside front page. There are two order forms in the middle of the stapled catalog, and the information needed is: quantity, part number, data sheet, price each, and amount extended (price each times quantity). Attached to one of the order forms is an addressed envelope without postage. More order forms and envelopes are sent with your order. IMPORTANT: Jameco now requires an order of $20 minimum. Prior to July 1985, the minimum was $10 [GRRRR!].

Jameco offers data sheets for 30 cents each on "most items." This is mainly for IC's, and formerly included all the information found in data books, such as pinouts, typical applications, graphs of various operating parameters, and truth tables. Now, however, the data sheets contain only basic recommended operating conditions and electrical characteristics. Jameco appears to be photocopying only the first two pages of the data book information for each IC. They do say more information may be requested at an additional cost. The 30-cent data sheet I received on one IC did not even include the pinout of the chip, so I recommend requesting all available information if you must obtain chip data this way.

Jameco requests $1 for postage and handling to obtain their free catalog, but they will send future catalogs and sales flyers free to active customers. If you order from Jameco, they assign you a customer number, which is found on your invoice, as well as on the upper right side of the address label on the catalog's back cover; they request this number on the order form.

In 1985, for domestic (USA) orders, postage and handling is 5% of the total order ($1 minimum). Insurance, if you want it, is $1.50, and Jameco is not responsible for uninsured parcels. Tax is 6-1/2%, and is applicable for California residents only. Jameco will accept checks, Mastercard or Visa. There is a $10 fee for returned checks. COD is available, but you cannot pay COD with a personal check.

All claims for defective material must be made within 60 days; include the defective material for testing purposes, Jameco's invoice number, and the date of purchase.

If parts are returned to Jameco due to customer error, you must pay all extra fees plus a 15% restock charge.

Claims for shortages or damages must be made within 10 days after receipt of parcel. If a part is on backorder they will tell you; this information is probably on the invoice, so check it first if you receive only part of your order. In my experience, Jameco usually has parts in stock.

Jameco sends sales flyers at least twice a year. These flyers are numbered; if you order from a flyer, be sure to include the flyer number on your order form. The flyers contain new prices for old items, items not listed in the regular catalog, "special sale" prices, and some items which are listed at their usual prices.

What's In It?

At last we can get to the fun stuff, and contemplate the contents of the catalog.

As much as possible, Jameco uses the "generic" part number for IC's. IC information includes part number, number of pins (helpful in determining socket size and, perhaps, whether you are ordering the proper chip), function and price. IC's are listed by the following types:

7400 Series (TTL):

Part numbers are preceded by SN (a Texas Instruments tradition), and followed by N (signifying plastic packaging). Ex., SN7400N.

The following types of 7400 IC's have no prefix or suffix:

Intersil IC's are listed by Intersil numbers.

CMOS IC's:

4000 series IC's are prefixed with CD (an RCA tradition). Ex., CD4000.

74C00 (CMOS) and 74HC00 (High-Speed CMOS) have no prefix or suffix.

The linear IC's have a more complicated order and part number. These IC's are listed in numerical order, regardless of the alphabetical prefix (ex., LM317T precedes LF353N). The pins column lists the "type of package" (if it is not DIP). It would be difficult to be sure of the suffixes used for the linear IC's (and some of the IC's listed later), but the following may give you a clue.

In general, the suffix N is used for most, but not all, of the DIP (Dual-Inline Package) IC's.

The suffix K denotes TO-3 package (diamond-shaped metal package with round capsule on top, and two pins underneath -- the flanges of the case being a third connection).

Suffix T denotes TO-220 (rectangular metal back with square IC package mounted on one side, and having three pins on one end).

Exar linear (ex., XR2242) and RCA linear (ex., CA3130E) are listed separately.

Also separately listed are Signetics IC's, DTL (Diode-Transistor Logic), HTL (High-Threshold Logic), the National 8000 series, and Fairchild 9000 series. The Digitalker chips are also listed separately.

There is, for different manufacturers' numbers, a cross reference guide for RAM's, PROM's and EPROM's, as well as listings for Memory subdivisions: Static RAM's, Dynamic RAM's, EPROM's, PROM's, ROM's.

There are subdivisions of Microprocessors and associated components (subdivisions: 5000, 50000 series, Miscellaneous circuits, A/D and D/A converters, CRT controller chips, Disk drive controller chips, UART's, Generators, Telephone/keyboard chips).

Microprocessor components are subdivided, but basically pertain to Z80, Z8000 series, 6500 series, 6800, 68000 series, 8000, 80000 series).

(Hope you wanted to know about IC's. That was a lot to describe, as well as to skip over.)

IC sockets are available in several types:

Information on precut crystals is given for computers, TV colorburst, etc. -- from 1 MHz to 49.89 MHz. This includes what they are used with, case size, and load capacitance (for parallel resonance).

For diodes and transistors, generic numbers are used as their part numbers. Important specifications for these semiconductors are also listed.

1/4 watt, 5% carbon film resistors are the only fixed resistors available in specifiable values. Values range from 2.2 ohms to 10 megohms. Resistors must be ordered in quantities of 5 each per value. All values are mixed in one package (unless individual packaging is requested, which, for less than 100 pieces, costs an additional 5 cents per value). Above 100 pieces per value, cost includes individual packaging for each value. There are discounts for quantities beyond 500 pieces.

Potentiometers include the popular square top-adjustment, single-turn type, and the long rectangular 15-turn type. For either type, specify standard resistance values of 50, 100, 500, 1K, 2K, 5K, 10K, 20K, 50K, 100K, 200K, 500K or 1M. Part numbers are 63P-(value) for single-turn pots, and 43P-(value) for 15-turn pots.

Several types of capacitors are offered: Miniature aluminum electrolytics are listed in "axial" (leads at both ends), and "radial" (both leads on one end) versions. Other specifications for each value include WVDC, surge VDC, approximate dimensions in inches, and then the part numbers. For a 100 uF 25 WVDC axiallead electrolytic, the part number is A100/25. For the same value in the radial-lead type, the part number is R100/25. Values available are: A.47/50 to A2200/16, and R.47/50 to R1000/16. Some values may be obtained in more than one WVDC.

Other types of capacitors are listed as follows: dipped tantalum (values in uF, are specified with part numbers of TM 0.1/35 to TM 100/6). Disc capacitors have part numbers such as: (pF values) from DC 10/50 to DC 470/50, and (uF values) from DC 0.001/50 to DC 0.1/50. Mylar capacitors have numbers of: (uF values) from MY 0.001/100 to MY 0.47/100. Silver-dipped mica capacitors have part numbers which are more involved: (values from 10 pF to 1000 pF). Part numbers are DM15-100J for 10 pF, DM15-101J for 100 pF, and DM15-102J for 1000 pF.

For trimmer capacitors: TC 2-8 is part number for value of 2-8 pF; other part numbers and values are TC 3-12, TC 3-23, TC 4-34, and TC 6-70.

Assortments of each type of capacitor are offered at substantial savings. Values included are marked with * to the left of the part numbers.

Various types of switches and relays are offered.

Audible devices include a few small speakers, a buzzer, and the SC628 Sonalert.

Tools include soldering irons, solder, and crimping and nibbling tools.

Grab-bag specials are assortments of good parts, usually costing less per piece than you will find otherwise; generally, values are not specifiable, and you will have to sort items such as resistors yourself. The items may be different in each catalog, so a reader will be necessary here.

The page of transformers, fuses and battery accessories includes DC wall transformers (adapters) from 6 VDC to 10.5 VDC. If you buy these, be sure to check plug polarity and operational viablity of each unit. (I have gotten units that were shorted, open, vibrated and got hot, as well as with center conductors negative in one shipment and positive in the next.) Be careful!!

Computer cable assemblies are available in various forms, and you can even order custom-made lengths with specified connectors.

The pages of "OK Products" include tools for wire wrapping, DIP/IC insertion and extraction, desoldering and drilling.

Vectorbord comes in 3-hole patterns: the A pattern has hole centers spaced 0.265 inches apart, with 0.093 inch diameter holes; H has 0.10 inch spacing with 0.062 inch holes; P has 0.10 inch spacing with 0.042 inch holes.

Three materials are available: phenolic with hole patterns A or P; epoxy-glass with patterns H or P; glass-epoxy with one side copper-clad, having hole pattern P. Phenolic hole pattern A is 3/32 inches thick; all others are 1/16 inches thick.

Vectorbord comes in various lengths and widths. In the 1985 catalog, the part numbers are labeled "Stock No." Vector terminals for soldering and for wire wrapping are available, as are insertion tools. Be sure the terminals match the hole size of your Vectorbord, and that the insertion tools match the terminals. Various types of Vector etched circuit boards for general use (and for specific computers such as Apple plugboards) are listed.

Global Specialties products include about four styles of solderless breadboards and protoboards, accessories for same, IC proto [test] clips, test equipment, instrument cases and hardware.

Also listed are regulated power supplies, keyboards, kits, disk drives, diskettes, computers, peripherals and accessories.

Many items are available which have not been mentioned in this article (oh my gosh, you mean there's more?). The indexes of the Jameco catalog, and probably other often-used electronics parts catalogs, may be made available in some accessible format at a future date (clip that hedge!).

EDITOR'S CORNER

Well now, a new kind of thing has appeared -- as of this issue -- embodied in Ms. Susan Fowle's article, "The In's and Out's of the Jameco Catalog." This kind of thing can hand us (on a zinc platter) a new measure of independence. Our thanks to you, Ms. Fowle, for giving us this fine distillation of material; it makes me want to pick up the phone and buy stuff. (Could it be that the motivation for doing this mammoth job comes from your husband yammering all day, "Could you please read that?" "No, read that." "Heah! Read that!"?)

The "In's and Out's of Jameco" is a landmark beginning; the choice of company was perfect, since it is the most useful organization I know of to the individual experimenter. Yet this idea would die right here and now without sighted assistance in doing the next one. Until -- indeed, if ever --" data bases" get good enough so that a blind person, with his computer, can browse through the world of parts, this kind of "distillation" cannot be affected by your lights-out editor. (I have tried minor attempts at this kind of thing -- notably the article on solderless breadboards in 1981; the error rate is unacceptable, and communication with beset readers was stressful for all concerned.)

Although good-ol' Susan has made tenuous overtures in the direction of doing another such piece on another supplier, why leave this job all up to one person? Half of you guys are around sighted classmates or hams who could try their hand at authorship in a similar vein. Get the smartest guy you know by the lapels and bark, "Say, how would you like to become a published author?" So what if it takes a couple of tries; send the paper to me, I'll critique it and send it back for revision, and we may end up with material that gives us key phrases for the "phone call to the supplier." (Please don't put someone to hard work, only to find out that the outfit doesn't do mail-order business.)

In the meantime, I inherited an idea from my dad which may be of use. Every few years, he would hire a neighborhood kid to read the index of the Allied Radio Catalog onto a tape, just so he could get an idea about what was out there to be had. He would write out a list of things he wanted to know more about, and pursue these items in more detail. This sort of thing, I could do for us. As examples, tell me that more of this wouldn't be fun:

The following is taken from the catalog index of TechniTool, and covers only the letter "E":

E.C.2000; ETM Pliers; Economy Tool Boxes, Edge Protectors, board; Edsyn Products, Eklind Hex-Key Sets; Electric Edger; Electrician's Kit; Electrician's Knife, Electrician's Scissors; Electro-Mechanical Kit (I don' know what it is, but I want that); Electroplating Tape; Electronic Assembler's Kit; Electronic Designer's Kit; Electronic Handy Pack; Electronic Timer; Electronic Operator's Kit; Electronic Specialty Kit; Electronic System-Maintenence Kit; Electronic Pliers (ouch); Elements, Torch; Emergency Kit; Endeco Desoldering Irons; Engineer's Tool Kit; Engraver, Electric; Engraving Cutters; Epoxy Dispensers; Epoxy; Esico Solder Pots; EtchoMatic Etcher; Eveready Flashlights; Extension Cords; Extra-Deep Cases; Extractor, DIP; Extractor, Lamp; Extractor, Lead and Shield (I assume that this is a stripper for coaxial cable); Extractor, Screw; Eyelets.

Although the Edmund Scientific catalog has no index, it does have major headings that jump out at you like:

"Complete kit turns any 12-to-19 inch TV into a giant TV projection system." "It's a watch; it's a radio; it's an alarm." "A thinking light switch." "It's the best news in home heating yet." "SuperBug picks up voices on the telephone -- and right through concrete." "Now your incandescent lightbulbs will burn for ten years or longer." "Drive pests away safely with our Sonic Sentry." "The Fischer 'Gemini II' Metal Detector is made for professionals." "Construct Geodesic Dome buildings easily with 'Star Plates'."

Now, the next question is, should I reprint this stuff in the valuable pages of this magazine, or should I just issue tapes with the "smart kid" reading this material at break-neck speed? You decide, please. Send a card with your preference (and any other thoughts) to me, before I go ahead and do the wrong thing.

So far, I have permission to reprint from the following companies:

Cole-Parmer Instrument Company; Digi-Key Corporation; Edmund Scientific Company; Jensen Tools, Inc.; Kulka Smith, Inc.; Small Parts, Inc.; Techni-Tool, Inc.; and Vector Electronic Company.

In my usual tradition, have a good summer. (I couldn't say have a good fall -- that would be rude.)