Smith-Kettlewell TECHNICAL FILE
A QUARTERLY TECHNICAL JOURNAL FOR BLIND AND VISUALLY IMPAIRED READERS
Sponsored by REHABILITATION ENGINEERING CENTER SMITH-KETTLEWELL INSTITUTE OF VISUAL SCIENCES
2218 Fillmore Street
San Francisco, CA 94115
SPRING 1984, VOL. 5, NO.2
TABLE OF CONTENTS
Bob Gunderson, W2JIO
Don't let the title fool you; it has nothing to do with my good or bad disposition. For nearly three years, I have been solar-powered with a 25-watt, 2-meter transceiver, plus a 100-watt Yaesu FT-707 all-band transceiver, along with a sun-powered SCA receiver, and an AM broadcast band receiver at my workbench.
As a boy, I was fascinated by the idea of communicating, and that fascination still persists. Two tin cans with a thin string conductor was perhaps my earliest telephone, and then came the earphones connected between houses by a double wire, or even a single wire, with the ground serving as the return leg. Battery-powered transmitter and receiver came later, and I can well remember the bicycle frame connected to a Model-T Ford generator by means of a rope. I literally pumped my heart out trying to keep the battery charged.
The idea of battery-operated equipment still appeals to me, and after reading an article in QST a couple of years ago, I decided to look into the problem of solar powering my gear.
Good fortune came my way when I became acquainted with Ted Handel of Los Alamos, New Mexico. Through Ted's kind interest I was able to purchase three 36-cell solar panels for my experimental work, plus a "how-to-do-it" taped response. I felt that I should share my good fortune with others, inasmuch as the price of solar cells is supposed to get much lower within the foreseeable future.
My panels were made by Spectrolab, Inc., and they will deliver an open-circuit voltage of about 22.5 volts in bright sunlight and a short-circuit current of about 550-600 milliamperes. They are essentially constant-current devices, which means, of course, that the current is pretty much independent of the output voltage. This also means that the short-circuit current provides a good indication of the panel's capacity; shorting them won't do any damage either. Although the Spectrolab panels are no longer available, equivalent ones can be gotten from Arco Solar,* Tideland Signal,* and Solec International. *
The matter of conversion efficiency is always of prime interest when discussing solar panels. Ted advised in his tape that the average yearly output of a panel located on a mountain top in New Mexico is in the vicinity of 100 milliwatts per square centimeter of solar surface. In order to obtain this high output, the air must be comparatively free of pollutants--in other words, this figure is assumed under ideal atmospheric conditions. The average here in New Jersey is considerably worse--running to a yearly average of something in the vicinity of 17 milliwatts per square cm. This means that the efficiency is only about 17 percent, based on the New Mexico maximum.
I have the three panels connected in parallel, and the peak output is about 2 amperes in bright sunlight. However, with this 17 percent efficiency, the yearly average is only some 340 milliamperes; yet, this value is considerable, based on the yearly average.
The panels are mounted on a telephone pole which also supports several antennas. The solar cells are mounted in an aluminum channel about 48 inches in length and perhaps 4 inches wide. The panels are mounted on aluminum angle at each end, and the top end is hinged to a wrap-around bracket on the pole, with the bottom end mounted to a similar wrap-around bracket, but attached to two threaded rods, allowing the tilt angle to be set. These rods are fitted with lock nuts so that the position of the panels can be fixed.
The panels are about 10 feet above the ground, so that it is simple to clean them periodically from a step ladder. I tried them at 50 feet and found the output to be about the same.
During the winter the panels are tilted up at about 15 degrees, so that snow and ice can run off the channels easily; during the summer the tilt angle is about 45 degrees. The panels are set in a fixed position, pointed a bit west of south, affording considerable sunlight during the average day. I do plan to build an automatic steering device which will allow the panels to "home" on the sunlight once each hour; more of this later.
The leads from the individual panels are terminated in a waterproof metal box mounted on the pole, and a blocking diode is connected in series with each panel so that the potential of each panel will be independent of the others. This is important if one cell in a particular panel shorts out, causing that panel's output voltage to decrease; were it not for the blocking diodes, the total output voltage would be governed by the lower panel voltage.
The output leads coming from the panel on the pole are then brought into the basement, some 60 feet from the pole, through PVC tubing underground. The storage battery is located in the ham shack on the first floor. This battery is a J.C. Penney so-called "sealed unit," and has an ampere-hour capacity of 95.
The manufacturers of lead-acid batteries recommend that these batteries be charged using the constant potential method, which means that the voltage of the charging source must be maintained at a constant value, for long battery life, this value is of the order of 2.33 volts per cell, or a total of 13.98 volts for a 12-volt battery (or 14 volts in round figures). Since the output voltage of the solar panel assembly is 22.5 volts, a voltage regulator, connected between the panels and the battery, is indicated.
On the other hand, if a 12- to 13.8-volt nickel-cadmium battery stack is used, these cells are charged via the constant-current method, and an appropriate resistance value is simply inserted between the 22.5-volt output and the battery to limit the initial charging current to a safe value (which depends on the size of the cells). (For more information on nickel-cadmium batteries, see SKTF Fall 1981.)
Should it happen that you wish to use a single solar panel, it is essential that you insert a blocking diode in series with the panel, so that in darkness, when the output voltage is zero, the battery cannot force a current through the panels. The blocking diodes at the pole take care of this in my installation.
I have a take-off point at the 22.5-volt output post for nickel-cadmium batteries, which feeds a variable resistance and a metering circuit for checking the charging current. In addition, I have a high-frequency solid-state buzzer (Sonalert type SC628) controlled by a pushbutton, which allows me to determine whether or not I have output.
The output also feeds a 2-transistor regulator and a zener diode to set the voltage at 14 volts. The output of the regulator feeds through a 2-arnp current meter to the battery line. The meter was included for the benefit of my sighted friends so that they can check the charging rate when visiting W2JIO. The meter is also shunted by an auditory transistorized Gimmick circuit which consists of an audio oscillator which changes pitch with changes in charging current. This Gimmick circuit is arranged to feed its built-in loudspeaker, or an outdoor speaker on the pole can be switched in. This allows me to adjust the position of the panels for maximum sunlight, indicated by the highest pitch in the speaker. (The Gimmick circuit appears at the end of this article.)
Voltage Regulator Circuit
The regulator, a Darlington configuration, consists of a 2N3055 series pass transistor driven by a small-signal transistor (type 2N3241A NPN), with a zener diode assembly connected in the input to this latter unit. The positive 22.5-volt line from the panels goes to the collector of transistor Q2 (2N3055), with the emitter of Q2 serving as the 14-volt regulated output for charging. The base of Q2 goes through a 1,000-ohm 0.5-watt resistor to its emitter. The Q1 (2N3241A) collector goes to the collector of Q2, with its emitter going to the Q2 base.
The Q1 base goes to the cathode of the zener diode (zener to be discussed later), with the anode grounded and connected to the negative dc output from the panels. The Q1 base also goes through a current-limiting resistance to the Q1-Q2 collectors. The zener diode is bypassed by a 100uF capacitor, with its negative grounded. This capacitor is also shunted by a 0.01uF disc ceramic capacitor. This completes the circuit.
[Note -- Of course, I have a 3-amp fuse in the line corning from the basement to the battery, right up at the battery for obvious reasons.]
The zener diode determines the output voltage, and I had some difficulty in finding the correct one for the necessary 14 volts from the regulator. However, I finally finished the project, using a 13-volt zener, plus a silicon and a germanium diode connected in series; my actual voltage turned out to be 13.98 volts. This zener string is made with the cathode of the zener connected to the Q1 base; its anode is connected to the anode of the silicon diode, with this silicon diode cathode connected to the anode of the germanium diode, and the germanium diode cathode grounded.
The value of the dropping resistor in series with the zener diode chain will depend upon the output voltage of the solar panels and upon the base current taken by the input transistor. The transistor types are not at all critical, although the pass transistor must be sufficiently large to handle the charging current of the battery. The 2N3055 is a good choice. If memory serves, the value of the limiting resistor in series with the zener diode is about 3.3K ohms, 1 watt. Its value must be chosen so that the zener voltage will remain stable, with and without the load on the regulator, which in this instance is the charging current for the storage battery. This current will, of course, reach a maximum when the battery is run down.
The 0.01uF capacitor in shunt with the 100uF electrolytic is an important component. When I first put the regulator into operation without it, I experienced all kinds of radio and television interference; evidently the negative resistance characteristic of the zener diode produced a broadband oscillator; the disc cured the trouble.
The regulator is built into a small mini-box, with Q2 mounted directly to its surface so that the box serves as a heat sink. (If mica insulating materials are not used in mounting this transistor to the case, remember that the metal cabinet will be common to the hot output of the solar panel, not common to ground.) The remainder of the circuitry is built on a small piece of perforated Vector board inside the box.
The system was put into operation on May 1, 1981, and it is used almost every day. Admittedly, the panels are not able to supply enough energy for continuous 24-hour-a-day duty, but I do operate it just about any time I wish. I have had to charge the storage battery from the line twice during this period when we were without sunshine for long intervals.
In order to achieve better circuit efficiency, the panels should be moved into maximum sunlight during the day as the sun moves across them. I plan to motorize the system with a reversible dc motor, controlled by a pair of photo-electric cells connected in a Wheatstone Bridge arrangement for comparing the light on each edge of the solar panels. When the bridge is balanced, the motor will stop. The direction of rotation will be governed by the polarity of the voltage at the bridge centers. A dc motor is essential, since it would be morally wrong to operate the system from the 117-volt ac house current.
A solid-state timer will be arranged to turn the system on once each hour to readjust the position of the solar surfaces. Of course, the system will shut down at sundown.
During overcast days, there is always some light falling on the panels, and on a cloudy day the charging current is perhaps 100-250 milliamperes. I often operate the entertainment devices--the a-track tape player (12volt), an SCA receiver, or an automobile CB set--at my workbench, directly from the output of the regulator with the battery disconnected; in other words, powered directly from the sunlight.
Is this system economical? I hardly think so. I have invested $300 in the panels, plus the cost of the storage battery, plus a great deal of time for experimenting with-the components. I do consider the experiment a great success, and I have achieved my goal--that is, learning a little about solar cells. If I had it to do over again, you may be sure I would gladly do so. Only this time I would try to obtain perhaps 10 panels with this current capacity, since the battery would be better off at a higher charging rate, particularly during deep discharges.
Auditory Charging Current Monitor
As mentioned in the text, a current meter monitors the charging current from the regulator into the lead acid storage battery, across this current meter is connected a "Transistorized Auditory Gimmick" which gives a variable frequency audio tone in accordance with the current.
A 2-amp meter is inserted, in series, between the regulator and the 3-amp fuse on the positive terminal of the battery (the positive meter terminal toward the regulator). The negative meter terminal is the common side of the circuit for the Auditory Gimmick.
The positive meter terminal goes to the base of a germanium transistor, with the emitter of this transistor going to common and to the negative meter terminal. The collector of this transistor goes through 47K to the base of a PNP silicon transistor.
The collector of the PNP silicon unit goes to common. Its base goes through 0.01uF to one end of the primary winding of an output transformer. The centertap of this primary goes to the emitter of the PNP unit. The free end of the primary goes through an on- off switch to the positive terminal of a 9volt battery, with the negative side of this battery going to common. In shunt with the transformer primary is another 0.01uF capacitor. The transformer's secondary goes to a small loudspeaker.
[Editor's Note-It's getting harder and harder to find parts for the good old Gimmick. For example, the Radio Shack no longer lists a germanium NPN transistor (276-2002). However, it is always worth asking for, since some stores have older items in stock. Even the output transformer is slipping into obscurity. However, I have unearthed the following alternative sources.]
Radio Shack still lists the output transformer (273-1380) which has a primary impedance of 1000 ohms (center-tapped) and an 8-ohm secondary. Mouser, on the other hand, has three nice units whose primary impedance is 500 ohms (center-tapped) and whose secondary impedance is either 4 or 8 ohms. These units are 42TM001 (open-frame type, PC mount), 42KM001 (open-frame type with flexible leads), and an encapsulated PC mount cube-shaped unit, 42HM001.
The PNP silicon transistor is not critical, and the very common 2N2907 will work. For the NPN germanium, RCA still lists the SK3011 unit, while GE lists two such transistors, the GE6 and the GE7. At one time, these were extremely common, and a request of a local ham (whose junk box is well stocked) will no doubt get you something suitable.
As described, the zener diode is made up of three units: A 1N4742 12V 1-watt zener, a 1N4148 silicon diode, and a 1N54 germanium diode (available from Mouser as the 581-1N54). An alternative would be to use the 14-volt 0.5 watt zener (1N5244) available from Mouser as the 597-1N5244-B. Finally, the blocking diodes mentioned can be 1N4001's, which are very common.
- *Arco Solar, P.O. Box 4400, Woodland Hills, CA 91365; (213) 700-7000.
- *Mouser Electronics, 11433 Woodside Avenue, Lakeside, CA 92040; (714) 449-2222.
- *Solec International, 12533 Chadron, Hawthorne, CA 90250; (213) 970-0065.
- *Tideland Signal, P.O. Box 52430, Houston, TX 77052; (713) 681-6101.
Patterned after the capacitance bridge by Albert Yeo in the winter issue, this is an auditory resistance bridge using two integrated circuits and very few components.
The classical approach to an auditory resistance bridge was to drive a Wheatstone bridge with a dc voltage, and detect the condition of balance by way of a chopper and amplifier. This works fine, and I have such a unit designed by Bob Gunderson in the early 1950's.
In this instrument, however, an ac signal drives the Wheatstone bridge; an audio amplifier simply "listens" to the output of the bridge. The advantage of this scheme is that the circuitry is much simpler than using a chopper system. The disadvantage is that inductance cannot be present in the unknown if a measurement is to be had. In other words, transformer windings and other inductances cannot be measured with this device.
A Signetics NE555 is used as the driving oscillator. Its output is used to drive its own charging resistor, thereby producing a square wave.
With the bridge being driven at an audible rate, an audio amplifier (using a National LM386) is used to detect the condition of balance at the "bridge centers." Because the oscillator signal and the amplifier inputs have no common connection, separate 9-volt batteries are used for each.
Pin 1 of a 555 goes to the negative side of its 9-volt battery. Pins 4 and 8 are tied together and go through one pole of a double-pole, single-throw, on-off switch to the plus 9V oscillator supply. Pins 2 and 6 are tied together and go through 0.01uF to pin 1. Pins 2 and 6 also go through 100K to the output, pin 3.
The bottom of the braille-calibrated pot (10K linear precision) goes to the negative supply lead of the oscillator battery. The arm of this pot goes to the arm of a calibration rheostat (15K), with the top of this rheostat going to the output of the oscillator, pin 3 of the 555.
Pin 3 of the 555 also goes to the arm of a 6-position switch. The following resistors have one end going to the following switch positions; the far ends of these resistors are tied together. (These resistors should be 1 percent precision units.)
- Position 1-100 ohms
- Position 2--1K
- Position 3-10K
- Position 4--100K
- Position 5--1 meg
- Position 6--10 meg
The tied-together ends of the resistors go to one binding post (the positive test terminal), with the other binding post going to the negative terminal of the oscillator battery.
Pins 2 and 4 of an LM386 are tied together and go to the negative side of the 9-volt amplifier battery. Pin 6 goes through the other pole on the on-off switch to the plus terminal of this battery. Pin 6 is bypassed to pin 4 by 250uF. Pin-7 is bypassed to pin 4 by 22uF. Pin 5 is bypassed to pin 4 by 0.22uF. Pin 5 also goes to the positive end of a 100uF capacitor, the negative end of which goes through the speaker to pin 4.
Pin 8 goes through 47e ohms to the negative end of a 100F capacitor, the positive end of which goes to pin 1.
Pin 3 goes through 0.1uF to the arm of a 10K volume control; the bottom of this control goes to pin 4. The top of the control goes through 39K to the junction of the resistors and their binding post (the positive test terminal). Pin 4 and the negative side of the amplifier battery go to the arm of the braille-calibrated pot.
Connect a resistor of known value across the binding posts and set, the 6-position switch to the appropriate range. Adjust the braille-calibrated pot for the desired reading; then adjust the calibration rheostat until a null occurs. The instrument is then calibrated for all six ranges.
The precision of this instrument is good to the combined error percentages of the calibrated pot's linearity and the tolerance of the "standards." If a pot whose linearity is good to within 1 percent is used, and if 1 percent resistors are used on the range switch, an accuracy of about 2 percent can be expected with the instrument.
For those who have never used a bridge like this, it might be worth mentioning the operating procedures. With the range switch in position 1, the braille-calibrated dial will cover the range of from zero to 100 ohms. With the range switch in position 2, the calibrations cover from zero to 1000 ohms. The third range covers 10,000 ohms, and so forth.
When measuring a resistor which is completely unknown, start with the range switch at position 6 and with the calibrated pot set at zero. Then switch to lower and lower ranges until the tone heard from the bridge is substantially loud. Try to obtain a null with the calibrated pot. If the tone gets softer but never reaches a null, even at its maximum, advance the range switch and try again. If the null occurs with the calibrated pot below 10 percent of full scale, decrease the range switch and try again. The optimal range will be lowest switch setting possible for getting a null.
Construction details are identical to those used in Mr. Yeo's capacitance bridge. The back panel of a Bakelite box contains the two binding posts. The top surface of the box contains the braille-calibrated pot, the 6position range switch, the volume control, and the on-off switch.
- 1--0.01uF disc, ceramic
- 1--0.1uF disc or Mylar
- 1--0.22uF disc or Mylar
- 1--10uF electrolytic
- 1--22uF electrolytic
- 1--100uF electrolytic
- 1--250uF electrolytic
Resistors (1/4 watt, 5%)
- 1-470 ohms
Resistors (precision, 1%)
- 1-100 ohms
- 1-1 meg
- 1-10 meg
- 1-10K volume control
- 1-10K linear precision
- 1--15K PC-mount trim pot
- 1--DPST toggle
- 1-Single-pole, 6-position rotary
- 2--Binding posts (insulated if using metal box)
- 2--9V batteries
Tom Fowle, WA6IVG
A digital counter is described which, when preset, beeps after a desired number of pages (from 1 to 99) have been copied. While installing this counter on our Thermoform machine, we made other improvements which made the job of copying a bit faster and easier.
It is extremely difficult to keep track of the number of copies you have made. Once you have lost track, it is annoying to count the finished pages; static electricity causes them to stick together, etc. Since photocopiers have them, why not provide the Thermoform machine with a counter.
A complicated approach would have been to build a talking counter which would announce the number of copies done. I chose a simpler approach, one of providing two 10-position switches which could be preset to a-desired number of copies. The counter emits a beep when this number has been reached. (Obviously, the two switches are to set 10's and units of copies; i.e., one switch goes from zero to 90, while the other goes from zero to 9.)
Triggering the counter could be done in a number of ways. (We are currently investigating the use of optical devices or magnetically triggered read relays which can be quickly attached to the side of the oven.) I chose the electrical approach--to install a relay, one contact of which triggers the counter, and another contact of which automatically shuts off the vacuum pump (something which is done by hand on the unmodified machine).
The switch on the machine which starts the timer is pressed by the oven as it is brought forward over the locking frame. A second contact is available on this switch which is connected to activate a double-throw, double-pole relay.
As the arm of the switch is connected to one side of the ac power line, and the normally open contact is tied to the timer, the normally closed contact was taken to one end of the coil of the relay. The other end of the relay coil goes to the other side of the ac line. Thus, when the oven is pulled forward, the switch starts the timer and opens the relay; and when the oven is pushed back, the switch closes the relay again.
One of the two poles of the relay has its normally closed side connected in series with the pushbutton which is ordinarily used to turn off the pump. When the oven is pulled forward, the timer starts, the relay is deactivated, and the pump can operate normally. When the oven is moved back, the relay closes, turning off the pump automatically.
As the relay opens and closes once each time a copy is made, the other pole of the relay can be used to drive an external copy counter. The electronics of the counter then count the contact closures made by the relay, and when the preset number of copies has been finished, the counter produces a one-second beep. The circuit for the counter is given below.
Circuit Operation of the Counter
The external counter consists of two CD4017 decade counter chips connected in cascade. The first of these chips is fed from the relay contacts through an NE555 timer chip set up as a special "contact debounce" circuit. (All normal debounce circuits failed because the open and close of the contacts are separated by several seconds, and the usual debounce "one-shot" delivered output pulses on both open and close.)
The CD4017 has ten outputs: 0 through 9 and a "carry-out" pin for feeding cascaded counters. As the clock pin is pulsed high, one after another of the 0 through 9 outputs is brought high in sequence. In other words, the counter chip is positive-edge triggered.
The carry-out pin is high from 0 through 5, goes low from 6 through 9, and goes high again when the count returns to 0. This produces a positive-edge triggering pulse to trigger the next chip.
Thus, two CD4017's cascaded can count to 99, producing a unique set of conditions on each possible count. The ten outputs of each of the counters are taken to ten positions of two rotary switches, and the swingers of these switches are logically AND'ed together to produce an output only at the number to which the switches are set.
The output pulse of the AND gate drives the first half of an NE556 dual timer which is configured as a monostable "one-shot" with a period of about one second. The output of this one-shot drives the second half of the NE556, which is a free-running oscillator set up to generate an audible tone (about 1kHz) that drives a speaker.
The reset pins of the two CD4017's are driven from the output of a logic OR gate whose two inputs are: (1) A front panel pushbutton used to reset the counter in case of an error in setting the selector switches; and (2), the output of the previous logic AND which indicates that the desired count has been reached. Thus, when the count is reached, the counter beeps and is automatically reset to zero, ready for a new group of copies.
Control Relay Circuit
A double-pole, double-throw relay having a coil activated by 110VAC is installed on the inside of the Thermoform machine's front panel. One end of its coil goes to the unused contact on the microswitch, which is hit by the oven. The other end of the coil goes to the contact on the "pump off" pushbutton, which does not go to the timer box.
The lead from the "pump off" button to the timer box is broken and routed through a pair of normally closed contacts on the relay. A pair of wires is connected to the other set of normally open relay contacts and brought out of the machine through the hole in the access door on the left side of the case. These wires are connected to a quarter-inch phone plug, mated to a jack in the external counter.
The negative of a 9-volt battery is grounded, while its positive goes through a single-pole, single-throw switch (on/off) to the plus V line. The plus V line is bypassed to ground by 100uF (negative at ground).
The sleeve of the quarter-inch jack from the relay contacts is grounded. The tip of this jack goes through 10K to plus V. This tip also goes through 100K to pins 2 and 60f an NE555. Pins 2 and 6 also go through 3.3uF to ground (negative at ground). Pins 4 and 8 of the 555 go to plus V, while pin 1 is grounded. Pin 7 (discharge) is not used.
Pin 3 (output) goes to the clock pin 14 of the first CD4017. Pins 8 and 13 of this 4B17 are grounded, while pin 16 goes to plus V. Pin 12 of the first 4017 goes to pin 14 of a second 4017. Pins 8 and 13 are grounded; pin 16 goes to plus V. Both 4017's have pin 16 connected through 0.1uF to pin 8 (close to chip).
Each 4017 has its 0 through 9 outputs connected consecutively to positions 1 through 10 of its single-pole, 10-position rotary switch. The 0 through 9 outputs are, respectively: 3, 2, 4, 7, 10, 1, 5, 6, 9, and 11. Thus pin 3 (the "9" output) of each chip goes to position 10 of its own switch, and so on... Pin 11 of each chip (the "9" output) goes to position 10 of its switch.
The arm of each rotary switch goes to an input of a NAND gate (one of four gates in a CD4011). (The gate used had pins 1 and 2 as its inputs and pin 3 as its output.) Pin 3 of the 4011 goes to pin 13, one of the inputs of another NAND gate. This pin 3 also goes to the NE556, to be described later.
Pin 12, the other input of the second gate of the 4011, goes through 10K to plus V, and pin 12 also goes through 0.1uF to ground. This capacitor is shunted by a normally open pushbutton (reset). The output of this second gate, pin 11, goes to pin 15 (reset) of both CD4017's. The other gates of the CD4011 are not used, their inputs, pins 5, 6, 8, and 9, are grounded.
Pin 7 of the 4011, the negative supply pin, is grounded. Pin 14 goes to plus V, and is bypassed to pin 7 (close to chip) by 0.1uF.
As mentioned before, pin 3 of the 4011 goes to pin 6 of the NE556. Pin 7 of the 556 is grounded. Pins 4, 10, and 14 of the NE556 are tied together and go through 10 ohms to plus V; they are also bypassed to ground by 10uF (negative at ground). Pins 1 and 2 of the 556 are tied together and go through 100K to plus V; they also go through 5uF to ground (negative at ground).
Pin 5, the output of this one-shot, goes through 47K to pins 8 and 12, the other half of this chip. Pins 8 and 12 also go through 0.01uF to ground and through 20K to pin 13. Pin 9, the output of this oscillator, goes through 47 ohms, then through the speaker to plus V. This completes the circuit.
Due to the contact debouncer, the beep is delayed slightly after the last contact closure, but it comes just as the last copy is removed. Opening and closing of the relay must be separated by about 0.5 seconds; however, massive relay bounce will not cause false counts, and the timing fits the normal operation of the machine.
One final trick makes copying considerably easier. Because of static electricity, the master tends to follow the finished copy off the machine. This problem is easily solved with a strip of masking tape placed diagonally across the lower right-hand corner of the braille paper master; the left and upper edges of the master press up against rubber gaskets, which tend to hold these in place.
[Editor's Note: Out of interest or out of boredom (I'm not sure which), I have measured my copying rate without these improvements to be 3 pages a minute. With all Mr. Fowle's improvements in place (including not having to worry about counting the pages), my copying speed is slightly over 4 pages a minute--a substantial improvement.]
Resistors (1/4 watt, 5%)
- 1--10 ohms
- 1--47 ohms
- 1--DPDT relay, 110VAC coil
- 1--SPST toggle
- 1--Normally open SPST pushbutton
- 2--Single-pole, 10-position rotary
- 1--1/4 inch open-circuit phone jack
- 1--1/4 inch phone plug
- 1--9V battery
This task is more demanding of your skills than any other. You can rest assured that any such connector has been designed to be of minimum size. Also, there will always be nearby wire insulation that you can abrade with the iron. Nevertheless, I have learned tricks which help--both in finding the pin and in avoiding others--some combination of which may get you out of a jam.
There are literally hundreds of varieties of multipin connectors, any of which you are likely to encounter. However, since the problems are similar (even for doing small switches), I have chosen two examples for this discussion. One is the DB25 connector (which has become the standard for computer interface), and the other is the DIP (dual in-line plug) that fits into IC sockets. If you can solder wires onto these, you can do anything.
[At the end of this article, computer connectors are listed which permit soldering of the pins onto their wires before insertion, thus making this job much easier. However, these are specialty items, and your Radio Shack store won't have them.]
The DB25 computer connector has two closely spaced offset rows of pins; 13 on one row and 12 on the other. (Actually, these connectors come in sizes of 9,15,25,37, and 50 pins, the 25-pin animals being most common.) The back ends of the pins (or sockets, since the females are similar) are scoop-shaped; being tubular to start with, their ends are cut off at a steep angle to leave "scoops" into which wires are laid and soldered. The "inner diameter" of each scoop is just large enough to accommodate wire of 20 gauge, and they have no eyelets or holes through which the wires can be hooked. In short, each wire must be held in its scoop, the pin then heated with the iron, and solder deposited into the scoop to surround the wire.
The DIP plugs are made up of standard flat IC pins mounted in a flat insulator plate; the top ends of these pins protrude only 1/16 inch above the insulator board, and these back ends are slotted to form little "forks" into which the wires are soldered. (Up to 22-gauge wire can be accommodated.) Each wire is laid in its fork so that its end reaches out- toward the side of the plug. It is then soldered to the fork, after which the end of the wire is clipped off close to the pin.
It should be obvious from the above descriptions that good systems for holding these items should be instituted. Vises and holding clamps are not only required for the connectors, but also for the wires and cables. For- example, one relatively heavy vise would be used to hold the connector. A smaller vise can be used to hold the cable (perhaps a foot back from the connector), and a nearby alligator clip can be set up to hold the individual wire, not more than an inch back from the connector. (These are only suggestions, but something has to keep the little wires in contact with their "scoops" or "forks.")
[Holding the individual wires will probably not be necessary in cases where the traditional solder lugs, having holes through which the wires can be mechanically secured before soldering, are present. In these cases, bend the stripped and tinned end of each wire almost double to form a "hook"; hook the lug with the wire and squeeze the hook closed with needle-nosed pliers. As the wire now holds itself in place, no complex array of clamps will be necessary. However, you will thank yourself later if you hold the cable so that wires come straight back off the pins, rather than at odd angles; they will be easier to trace in the future, and bridged wires will be less likely.]
The vise for holding your connector should be able to swivel and tilt in several directions. As the thicket of wires becomes significant, you will want to position the connector so as to avoid marring previously done wires with the iron. Also, depending on the connector's design, soldering on female terminals may have to be done with the pins horizontal, so as to keep solder from running through the terminal and ruining the socket. You will often have to use your imagination in order even to grab a connector in the vise. For example, there is not enough meat on a DIP plug to hold securely. In this case, the plug can first be inserted into a good, deep socket to form an assembly of significant substance to be clamped securely.
A good vise for holding the connector is the "Panavise" listed in the "Tools" section of Soldering, Part II (SKTF, Winter 1981). (Of particular usefulness are a swivel-type base such as the No. 380 "Vacuum Base" and the standard vise head, No. 303.)
Unfortunately, however, the jaws of the Panavise are plastic; they will readily contaminate the tip of the iron if struck accidentally, and they are easily marred by the iron. For this reason, it is advisable to line the jaws with braille paper, leaving flaps of paper folded back over them to protect the jaws from the iron.
I have used an extraordinary variety of arrangements for holding individual wires behind their terminals. An alligator clip suspended by a piece of coathanger makes an adjustable structure for this purpose. The far end of the coathanger can be held in another clamp (such as your board clamp). Some arrangement with locking forceps in a board clamp is also effective. Actually, the wire being worked on can be fixed to one which has already been done by biting them both in a small alligator clip. Holding a wire parallel to its neighbor is a rather secure way of doing it.
[There are commercial holding devices which are built around alligator clips. One such instrument -is the Radio Shack 64-2093, known as the "Helping Hands." Mounted on a small cast iron base, ball joints are used to support two alligator clips, one at either end of a horizontal boom. The base of the "Helping Hands" is not very heavy, however, and building a new base for it would be highly advisable.]
Any arrangement for holding the cable a foot away from this paraphernalia will work. Very often, I will just sandwich it between a couple of heavy books or transformers; tying the cable to a C-clamp on the edge of the workbench is ideal. However you do it, the idea is to keep the cable from pulling back or twisting as you set up the clamping system for the connector.
Preparing the Cable
Short leads cannot be formed or held so as to stay in position for solder. Furthermore, arranging leads to your advantage cannot be done without flexing the cable; short leads that were previously attached will be stressed to the breaking point. Finally, if heat-shrinkable tubing is threaded over each wire in anticipation of insulating the finished connections, short leads will conduct enough heat to the tubing to shrink it while soldering--it will no longer accommodate the terminal. Therefore, the prepared ends should be at least 1-1/4 inches long.
After you have stripped and tinned the ends, inspect them for marred and peeled insulation. Cut the stripped portion very consistently to perhaps 3/16 inch in preparation for attaching the wires to their respective terminals.
[Of course, the appropriate length of the stripped and tinned ends depends entirely on the type of solder terminals on the connector. For example, large solder lugs with holes in them demand that 1/4 inch of lead be available to hook through them. Also, in the case of DIP plugs where you are soldering to very tiny forks, you may wish to make the bare ends quite long (perhaps 1/2 inch) so that you can use this wire as a landmark to guide the iron. In any case, you will find that consistency is your best friend; errors are quickly spotted among a grove of otherwise uniform terminals.]
Soldering--Doing More Good Than Harm
Probably the single most significant tool to enhance the chances of successful work on multipin connectors is the Japanese Solder Guide (the JA3TBW Solder Guide, SKTF, Spring 1983). Invented by Mike Bhagwandas of Kobe, Japan, it provides the user with a good means by which a desired terminal can be found with absolute certainty. A brief description follows (appropriate suppliers are listed at the end of this article).
The solder guide consists of a 3- or 4-inch piece of thin-walled stainless steel tubing whose inner diameter is about 0.05 inches. Flux-core solder of about 0.03 inches in diameter is fed through the tube so that it just emerges from the "bottom" end, this bottom end is then rested on the desired terminal. In operation, the tube is used to guide the hot iron to the terminal, and solder is then fed to the work from the "top" end of the tube.
A sort of "handle" can be provided near the center of the tube, this can be made from a drilled-out braille stylus handle, or a simple shaft collar can be secured to the tube with a setscrew.
Solder is fed down the tube with the thumb and first finger, while other fingers are used to support the tube above the "handle." The iron is brought to the lower section of the tube, whereupon it is slid down to contact the work pieces.
Soldering can still be done the old way (without the tubular solder guide) by using tactile feedback to verify that the right terminal has been found with the iron. without the tube, a system of nearby landmarks will greatly enhance your chances of hitting the right point with the iron; these landmarks might consist of, first, the vise, second, a nearby alligator clip, etc. With the solder in one hand and the iron in the other (the solder resting on the work), verification of hitting the terminal with the iron will come by noting the vibration of the solder, then by watching the solder melt.
Because the pins are close together, a common error is the accidental bridging of two adjacent ones. Shorts between pins can occur in the following four ways:
I often lose the pin temporarily with the solder--either I knock it off with the iron, or it melts before I have expected it to--and I must then fish around with the solder to find the hot pin. In re-finding it, it is easy to spill a little extra solder into the connector, usually on the pins below. This is, however, one of the easiest mishaps to correct. If metals onto which solder has been spilled never reach soldering temperature (they usually will not if they receive no direct heat from the iron), solder droplets will not firmly adhere to them. The droplets can be dislodged later with a braille stylus or a probe-type soldering aid. Remember to check for droplets if you suspect that spillage has occurred.
Icicles of solder often form if the iron is abruptly pulled away from a connection-sliding it off the connection will usually break the surface tension and prevent icicles. If you abruptly leave the pin in the direction of an adjacent one, you can form an icicle which decreases the distance between pins. On the other hand, there really is not room to talk about sliding the iron off these closely spaced terminals. Therefore, go ahead and jump off the pin--knowing that occasional icicles will result--but do so in a benign direction (straight up, for example).
If wires are touching each other directly behind the pins, their insulation will some-times melt from transmitted heat, whereupon the insulation will give way and allow the wires to touch each other. If this happens, they can often be pried apart and then reinsulated with tape (or tubing, if it has been previously installed onto them).
The most sinister kind of bridge is the bonafide "solder bridge"; this occurs when two adjacent pins are simultaneously heated, thus allowing a bridge of solder to alloy with both of them. To clear this type of bridge, the connections must be simultaneously reheated and a probe of some sort passed between them. [Sometimes a knife or a small "mill file" (a mill file has teeth on its edge) can be used to destroy the bridge without reheating the pins, but you risk breakage of the pins or the wires with the inevitable rough handling.]
There are four ways of detecting shorted terminals. These are the indications you should look for:
- Connector pins often exhibit some mobility in their socket block; they can often be wiggled slightly. They should all wiggle separately--one at a time. If two pins insist on moving together, they're bridged.
- Some kind of probe that fits between the terminals can be used to explore them--a stylus, a soldering aid, a dental tool filched from your dentist, a jeweler's screwdriver, etc. Make sure that you can feel the insulating material of the socket block at the base of the pins; then make sure that you can bring the tool unimpeded all the way back to the insulated portion of the wiring.
- If little pieces of tubing have been installed onto the wires in anticipation of insulating the pins (thus burying your sins, as my supervisor used to say), a good indication of success is if the little segments of tubing slide all the way down over the pins to the insulator block. Once again, consistency will be your best friend; if you have cut all the tubes to exactly the same length, unevenness at their back ends will show you which ones to investigate further.
- Finally, the best (and the most tedious) test of the connector is checking it with a continuity tester, going pin by pin. Each pin can be checked for shorts to all of its neighbors (four neighboring pins, in the case of the DB25); each pin can be tested for continuity back to its wire at the other end of the cable. (I don't always go through this rigor myself. However, the electrical items of lowest reliability are connectors; if you really want to know that these are not your problem when the stuff doesn't work, check the connectors thoroughly.)
[For making the above continuity checks, special test probes that fit the connector are very helpful. These can be made by ravaging a spare or damaged unit of the opposite gender. First, remove any metal framework with a hacksaw--then dissect the insulator block with a coping saw to get a few sample terminals. Then, make test probes out of them, using heat-shrinkable tubing or tape to cover all but the business end. The probes can be plugged into the connector you are testing, and the tubing on the outside will keep you from getting false alarms when fishing around adjacent pins.]
The Paper Dam
This trick has indeed served me well; I first tried it when I was called upon to wire connectors having 120 pins (a slice of my early work history that I would just as soon forget). My first "dams" were made from little strips of braille paper-this worked most of the time, but starting a conflagration is possible. Other materials worked better in the face of extreme temperatures.
My favorite material is "fish paper," an insulating material that is not easily damaged by the heat. (GC Electronics of Rockford, Illinois markets fish paper in a 10- by 24-inch rolled-up sheet. Their name for it is "Fyberoid," No. 560.) Another good source of fish paper is the scrap heap; old TV sets and other consumer products are loaded with it. Though bits of it look like scrap to someone else, I save every square inch.
Another material is Teflon, strips of which can be cut from liners of broken tape cassettes. (Make sure that the plastic liner of your cassette is Teflon, other soapy-feeling plastics which are not heat-resistant are used in cheaper tapes.) You should be careful of Teflon, however; once it does reach the temperature of its destruction, it gives off gases which are highly toxic--keep a breeze going around the workbench.
You can test a material for its suitability by bringing the barrel of your iron into contact with it. If it melts or smokes, don't use it.
The principle of the dam is to isolate the terminal of interest from its neighbors. It will not only prevent bridging, but it will do much to keep you from marring previously done wires in your attempt to find the terminal with-the iron. The dam, a strip which is perhaps 1/4 or 3/8 inch wide, is "woven," as you might say, through the terminals.
My usual configuration is an L-shaped affair with the bend of the "L" being a small u-shaped gutter that accommodates the terminal being soldered. (In terms of items you might recognize, its shape is rather like a sewer trap.) Suppose, for example, you are soldering wires along the top row of pins on a DB25 connector--also picture yourself facing the end of the connector. Suppose also that one or more wires have been attached near the end farthest away from you. One end of the dam would be sticking straight up between your next terminal and the ones previously done. The dam then courses under the terminal of interest, after which it lies flat on top of the pins to come.
The dam then becomes your main source of landmarks with the iron, you can follow the horizontal surface of the dam until it drops down under the terminal you want. Or, with the iron, you can first find the vertical fin that prevents you from hitting previously done wiring--then follow this vertical surface down to the pin. In any case, there is only one pin to which you have easy access, and it's the one you want.
A much simpler example is that of using a dam for doing a pin at the end of the row. Although I seldom use one here, a simple vertical strip between this and the other pins is sufficient.
If the pins are long (which they are not in the case of the DB25 connector), they alone will hold the dam in place. If they are short, someway of holding the dam in position will have to be devised. The fish paper dam can be held in place by a small bit of masking tape; lay the tape across the horizontal leg of the dam and secure it to the frame of the connector. Teflon, on the other hand, resists adhesion to sticky surfaces, and it is best held in position by a small loop of wire or an alligator clip.
The kind of tape you use is very important. Don't put plastic tape around where you can accidentally find it with the iron; it will soil the tip and render the soldering iron inefficient. On the other hand, no adhesive tape is impervious to the heat of the iron. However, I have had good luck with paper tapes (such as masking tape), and you should apply it with the understanding that it should not be used for resting the iron against while soldering.
[In reviewing my procedure in preparation for writing this article, I came across a DB25 connector of unknown brand whose insulator block is easily melted by the iron--a sinister practical joke indeed. My only recourse was to lay masking tape over the side of the connector with the edge of the tape "lipping" down over the insulator block above the pins. Heating the tape was unavoidable, but it was better than soiling the tip of the iron and ruining the connector block.]
Additional Comments on the DIP Plug
Because of its construction (having very short pins to which wires are soldered at right angles), I have found no way to isolate these pins from one another while soldering. Success depends mainly on how stable the clamping setup is--whether or not the wires tend to stay in position while you work.
The wire of interest can serve as a good landmark for finding the pin. If previously done wires are cut off close to the terminals, and if a substantial length of your prepared lead is allowed to protrude past the edge of the plug, this lead becomes "the only tree in the forest." The iron can be used to find the lead while the solder is held against the pin to detect vibration when doing so.
If you wish to take the trouble, there is a very secure system of holding wires in position which I have recently tried. Insert the DIP plug through a piece of perforated board, then plug it in to a good-quality socket to hold it there. When stripping and tinning each lead, provide one full inch of bare wire beyond the insulation. Before laying the wire in its fork, poke it into a hole in the perforated board adjacent to its pin and bend it over so as to hook it in position. Then stretch the wire across the fork and tape it down against the board on the other side of the socket. With this system, jostling the wire loose from the fork will be unlikely.
After one row of pins has been completed, you will find that avoiding this wiring while doing the second row is very difficult. This again is a good application for masking tape place masking tape across the completed forest of wires to protect them from the iron.
This article has been very difficult to write--connectors are widely different, and there is no foolproof procedure to assure successful soldering. All I can do is suggest hints and techniques which have been of use to me in various circumstances.
As you develop your own methods, you will no doubt sacrifice connectors. Always buy extra ones, knowing that spare parts from the old ones may be useful as test probes, and in some cases may be used to replace damaged terminals.
Where possible, do your soldering before inserting the terminals into the socket block, there will be no nearby terminals with which you can bridge connections, etc. A good example of this was the 120-pin connectors I was called upon to solder in my work as a technician. After wiring them (and having much trouble with the finished devices), I discovered that my supervisor had inserted the terminals into the blocks--he was trying to be helpful. Wherever those instruments are, the connectors are still a source of trouble.
It is customary to number connector pins, viewing them from the pins of males or from the backs of female sockets. In addition, if a "key way" exists (some mark around the perimeter of the connector), this key way will generally appear between the lowest and highest numbers. For example, DIP plugs will usually have a notch at one end, holding the plug with the notch up and with the male pins facing you, count them in a clockwise direction starting with the pin to the right of the "key way." A 14-pin DIP plug would have its key way placed between pins 1 and 14, the column of pins on the right-hand side would be numbered 1 through 7, while the pins on the left side would be 8 through 14 (7 and 8 near the bottom).
The DB-type computer connectors are numbered a bit differently. For example, viewing a DB25 connector with the male pins facing you, orient the rows of pins so that the row of 13 is on top. Then, pins 1 through 13 are on the top row (counting from left to right), pins 14 through 25 are on the bottom row (also counting from left to right).
The Amphenol "Poke-Home" Series
These units are shipped with the pins not being installed in their socket blocks. As you can imagine, soldering these pins individually would be much easier than soldering pre-assembled connectors. When soldering, the forward ends of the connectors should be held in something which does not act as a big heat sink (they should not be directly held in a metal vise). Line the jaws of your vise with braille paper or fish paper before grabbing the pins. Another idea is to hold the pins in something of less heat capacity, such as gripping them in your locking forceps.
The numbering system is simple. For example, the male set (insulator block and 25 pins) of a DB25 plug is an Amphenol No. 17-20250. The female set is an Amphenol No. 17-10250. Connectors with a different number of pins require the "25" figure to be changed. For example, a DB15 male would have the number 17-20150, and a 9-pin male would bear the number 17-20090. In like manner, a 37-pin female would bear the number 17-10370.
Several people make equivalents to the "Poke-Home" series, and it is always worth asking your supplier for this breed. However, a more common version of the "Poke-Home" series has crimp-on pins, and an expensive crimping tool is needed to attach them. Be sure you know what you're getting.
Solder Guide Parts--6-inch lengths of hypodermic tubing, catalog No. HTX-15, are available from Small Parts, Inc., 6901 North East 3rd Avenue, Miami, FL 331381 (305) 751-0856.
Shaft collars are available as "set screw collars," catalog No. 889, from the Player piano Company, 704 East Douglas, Wichita, KS 67202; (316) 263-3241.
Amphenol Industrial Division, 1830 South 54th Avenue, Chicago, IL 60650; (312) 242-1000.
GC Electronics, 400 South Wyman Street, Rockford, IL 611911 (815) 968-9661.
I have used heat-shrinkable tubing for a myriad of purposes, even for making assemblies for Light Probes. I will leave the purposes up to you; this paper describes the equipment and the pitfalls encountered in using it.
This spaghetti tubing is ideal for insulating closely spaced terminals, for putting stress-relief sheaths behind emerging cables, and for making your own tightly bound cables containing the conductors of your choice. Although polyvinyl chloride is the usual material, heat-shrinkable tubing can also be gotten in Neoprene, Teflon, Kynar, and Polyolefin (whoever she was).
The molecular structure of the plastic in this tubing has been stressed in the making. What's more, it remembers the dimensions it once had before being stretched and cooled by the manufacturer. It is simply aching to return to these dimensions (about half the specified diameter), and when softened by the heat, it does just that.
In operation, the piece of tubing--which is too big for the job--is installed. Then, hot air from a heat gun is played over the tubing --gently moving the gun o'er the project so as to heat all sides of it. The tubing will then shrink down and conform to the work pieces.
One disadvantage of heat-shrinkable tubing is that it is not particularly flexible. Therefore, unless a very small cable is made with it, stress at the termination points will be substantial, and good cable clamps should be provided at the connectors. (Those of you who are familiar with the Optacon will remember the lack of flexibility of the camera cable on early units. The designers had their work cut out for them, though-there were 32 separate enameled wires running through heat-shrinkable tubing.)
Another point is that once the tubing is cooled, it is rather inelastic; it will not tightly grip its cable after cooling off. Therefore, a short length of heat-shrinkable tubing will not, in itself, make a very good strain relief. Consider the example of attaching the most basic RCA plug (without a strain relief of its own) to a round cable, after which a piece of heat-shrinkable tubing is used to enclose this assembly. The tubing will not firmly hold the cable; twisting and turning of the cable will be possible. In fact, unless a blob of solder or some other lumpy feature holds it there, the tubing may slide back along the wire without much resistance. (A flat cable is better protected in this way; at least twisting will not occur.)
[Where flexing of wire or cable is expected, the ideal setup is to have an actual strain-relief clamp under the tubing; this prevents twisting. A sheath of heat-shrinkable tubing outside the clamp, reaching back along the cable for an inch or so, will then prevent flexing from chronically occurring immediately at the cable clamp.]
Choosing the Appropriate Size
Pick a size that easily (perhaps loosely) fits over its intended assembly. A common mistake is to start with a size which snugly fits--"then, it'll really hold when you shrink it." Not so. In the first place, its eventual inelasticity is such as to never "really hold" anything. Second, since it is already stressed, it tears and ruptures easily if forced over something it doesn't quite fit. On the other hand, when it is allowed to shrink, the tubing "recovers" some of its wall thickness.
Where widely contrasting sizes are present (for example, the aforementioned RCA plug on a cable), you should resist the temptation to choose a size that will shrink to the smaller diameter, but which must be stressed to fit the larger item. A good alternative is to use two or more concentric pieces--adaptors, if you will-to get from one size to the other.
Remember, the shrinkage you can reasonably expect is such as to reduce the inner diameter by half. In the literature, this attainable dimension is called the "recovery size."
You can buy long lengths of heat-shrinkable tubing in two forms, on spools or in straight lengths. While the spools are convenient for storage, they have a major drawback. Because of its inherent inelasticity, this tubing kinks when it is bent. When it is spooled, it just gives up and collapses; it's flat as can be when you get it on spools. Users hotly debate the virtues of each style of packaging.
"HOW can you use that stuff--squashed flat like that!? You squeeze and you pinch; you can never get wires through it."
"Yeah? I make cables all the time. You just don' know how to use it. You and your 4-foot lengths; it costs a mint to ship it! What do you do when you want a 5-foot cable? Then, too, every time you make a shorter cable, you end up with a bunch of odd lengths lying on the floor of the closet--no good place to store 'em."
They both have a point; while the spools are easier to store, the flattened tubing cannot be used without knowing a couple of tricks. Thought the following suggestions will be lifesavers when using kinked or flattened tubing, they will help you no matter which kind you buy.
Once you open the end of the tubing (using a pointed instrument of some sort), you can squeeze its edges so as to crease and flatten it in the other direction. Any creases will completely disappear as soon as it is heated.
One good way to get wires through is to solder one end of all the conductors to a "leader wire," a piece of solid wire which is long enough to go completely through the tube before the conductors. Once it has emerged, clamp the end of the leader wire in a bench vise and use it to help you "milk" the conductors on through after it.
There is a little-known way to get your conductors through without a leader wire (which is a good thing, since the leader breaks off as often as not). Once the conductors have been fed about a foot down the tube, find their forward ends and grasp the tube firmly so as to hold them in position. With your other hand, slowly pull and stretch the tubing back so that it covers more of the conductors; firmly grasp the end of the tubing to hold it in position against the conductors. Now, let go with the first hand. The tubing will shorten to nearly its original length, thereby forcing the conductors forward into new territory.
When pulling back on the tubing--in order to stretch it--work slowly so as not to heat it up. Otherwise, you'll have a finished cable before its (it's) time.
People will tell you that you can shrink this stuff by bringing it near the barrel of your soldering iron or with a cigarette lighter. These methods only "sort of" work, the jobs come out all wrinkled and lumpy, if not burned. Though a heat gun may cost you $60 or $70, it works wonderfully.
Though I have never tried them, there are infrared heat guns available which have no moving parts. I wonder, however, if you could point these at a cluster of tubing-covered terminals and expect all the tubes to shrink uniformly. With a forced-air heat gun, there is no doubt about it; you warm everything up, including your nearby coffee cup.
Choosing a Heat Gun
Unless you are planning to make cables to your specifications, a small heat gun will be sufficient. In fact, it will be preferable, since you can damage a project by flooding it with hot air of perhaps 700 deg.F. For insulating terminals and for reinforcing cables as they emerge from connectors, small guns such as the Ungar Model 6966C or the Mastermite 10008 will be appropriate.
On the other hand, if you plan to make cables of several conductors, or if you use the tubing to "package" large components (NiCad batteries and test probe assemblies are often encapsulated in heat-shrinkable tubing), your heat gun had better be a big one. Using a gun which is too small is inadvisable; as you dwell on various portions of the work, items inside will be subjected to excessive localized heat, and damage may result. For large jobs, get a heat gun that heats up a large surface area quickly; once this is accomplished, get the project out of there.
[Once I had the job of bundling together cable of 120 wires in firehose-sized tubing (1-3/4 inch). A Master Appliance I5-amp heat gun was used, Model HG501.]
CautionThe on-off switches of heat guns provide two modes of operation --heated air, and just plain circulating air (the fan running without the heating element being on). This is done so that the fan can be run to cool off the element before turning the unit completely off. If you follow this procedure, the heating element will live longer.
List of Heat Guns
I have not found a supplier of Alpha heat guns to recommend specifically (call them up and ask for an appropriate one in your area); this is a big company. Also, notice that there is one which runs off of 230 VAC.
Alpha Hot Air Guns:
- FG2--4.5 amp
- FG2A-2.3 amp, 230 VAC
Other Hot Air Guns:
The following heat guns are listed in the catalog of Fordham Radio.
(1) Master Appliance:
These are industrial guns; although they may have more power than you need, they are not much more expensive than their junior rivals. You can use them for small things; just get in and out of the project quickly.
- HG301--12 amp, 300 to 500 deg.F., about $75.
- HG501-15 amp, 500 to 700 deg.F., about $80.
- HG751--20 amp, 750 to 1000 deg.F., about $90.
- 10008--4.5 amp, about $60.
- 6966C--25B watt, 75B deg.F., about $7B.
- 6975--975 deg.F., about $8B.
Most of what you see as heat-shrinkable tubing is apparently polyvinyl chloride; this is the only material with which I have had experience. [There seems to be a remarkable similarity between designation numbers of the various manufacturers. For example, the number 105 shows up in every case. Were I a chemist, I'd probably have something to say about this.]
Tubing is sold by the inner diameter at the "expanded size" (the size before being heated). The "recovered size," at least for PVC, ends up being about half the given inner diameter.
I have chosen not to list every parameter. For example, the "recovered wall thickness," which may be from 0.0020 to 0.035 inches, is usually listed in the print literature. The length changes slightly as well, although this effect is hardly noticeable.
Only PVC tubing numbers are listed here, although they have heat-shrinkable tubing in Neoprene, Kynar, Teflon, and polyolefin as well.
The smallest spooled lengths they list are 100 feet (up to 3/4 inch, at which point you can get 53-foot rolls of larger sizes). They also sell straight lengths, 4 feet being standard.
Their small sizes are given in thousandths of an inch, i.e., from 0.022 (PVC105-24) to 0.330 (PVC1-5-0). Fractional sizes then begin at 5/16 inch (PVC105-5/16), with the largest size being 2-1/2 inches (PVC105-2-1/2). In small sizes, 6 colors are available, and they are listed by number as follows:
[Calling their distributor here in Oakland, the salesman suggested that the color be named rather than numbered. Nowhere in the book do they show how to attach this number to the designation number.]
From PVC105-2 (0.263 inch), up to 1 inch, only clear (1), black (2), and white (3), are available. Only two colors, clear and black, are available above 1 inch.
Adding an "S" to the number denotes that your order is to be on spools. The length of the spool must be specified separately. If straight lengths are desired, the length of the section is to be specified (4-foot lengths are normally stocked); i.e., a 4-foot length of 3/4 inch tubing would be listed as follows: "PVC105-3/4-4-black." The same tubing spooled would be: "50 feet of PVC105-3/4-S-black."
Small sizes (given in decimal fractions of an inch):
- PVC105-24, 0.022
- PVC105-22, 0.027
- PVC105-20, 0.034
- PVC105-19, 0.038
- PVC105-18, 0.042
- PVC105-17, 0.047
- PVC105-16, 0.053
- PVC105-15, 0.059
- PVC105-14, 0.066
- PVC105-13, 0.076
- PVC105-12, 0.085
- PVC105-10, 0.106
- PVC105-9, 0.118
- PVC105-8, 0.133
- PVC105-7, 0.148
- PVC105-6, 0.166
- PVC105-5, 0.186
- PVC105-4, 0.208
- PVC105-3, 0.234
- PVC105-2, 0.263
- PVC105-1, 0.294
- PVC105-0, 0.330
- PVC105-5/16, 5/16ths
- PVC105-3/8, 3/8ths
- PVC105-7/16, 7/16ths
- PVC105-1/2, 1/2
- PVC105-9/16, 9/16ths
- PVC105-5/8, 5/8ths
- PVC105-3/4, 3/4
- PVC105-7/8, 7/8ths
- PVC105-1-IN, 1 inch
- PVC105-1-1/8, 1-1/8 inch
- PVC105-1-1/4, 1-1/4 inch
- PVC105-1-3/8, 1-3/8 inch
- PVC105-1-1/2, 1-1/2 inch
- PVC105-1-3/4, 1-3/4 inch
- PVC105-2-IN, 2 inches
- PVC105-2-1/4, 2-1/4 inches
- PVC105-2-1/2, 2-1/2 inches
In contrast to other listings in this article, this company at least supplies them directly; you call them up and order what you want without having to go through a representative.
Although tubing can be bought on spools of 1000 feet, 10-foot lengths (folded up) are what they list which is of interest to us.
- 536-HS105-1/16X, 1/16 inch
- 536-HS105-3/32X, 3/32 inch
- 536-HS105-1/8X, 1/8 inch
- 536-HS105-3/16X, 3/16 inch
- 536-HS105-1/4X, 1/4 inch
- 536-HS105-5/16X, 5/16 inch
- 536-HS105-3/8X, 3/8 inch
- 536-HS105-1/2X, 1/2 inch
- 536-HS105-3/4X, 3/4 inch
- 536-HS105-1X, 1 inch
Their selection is disappointing; they only list a packet of assorted sizes, 278-1627.
Besides the number, specify the color and add the suffix "S" if spooled tubing is what you want. For example, 1/8 inch black spooled tubing would be: "No. PVCR-105-1/8S-black." Without the "S," 4-foot lengths are available.
- PVCR-105-3/64, 3/64 inch
- PVCR-105-1/16, 1/16 inch
- PVCR-105-3/32, 3/32 inch
- PVCR-105-1/8, 1/8 inch
- PVCR-105-3/16, 3/16 inch
- PVCR-105-1/ 4, 1/ 4 inch
- PVCR-105-3/8, 3/8 inch
- PVCR-105-1/2, 1/2 inch
- PVCR-105-3/4, 3/4 inch
- PVCR-105-1IN, 1 inch
- PVCR-105-1-1/2, 1-1/2 inches
- PVCR-105-2IN, 2 inches
- Alpha Wire Corporation, 711 Lidgerwood Ave., Elizabeth, NJ 07207; (201) 925-8000.
- Fordham Radio, 855 Conklin Street, Farmingdale, NY 11735; (800) 645-9518.
- Mouser Electronics, 11433 Woodside Avenue, Lakeside, CA 92040; (619) 449-2222.
- Remtek Corporation, 2021 North Capitol Ave., San Jose, CA 95132; (408) 946-8400.