Although the home entertainment electronics industry is still in the process of changing from vacuum tube circuitry to the use of transistors, another revolution also is in the making. The new gadget causing all the stir is the integrated circuit (IC), sometimes simply called a "chip." Incorporating this new device into a television receiver might appear as just another way for the manufacturers to make life miserable for servicing technicians, but in reality there are some pretty sound reasons why we are going to see more and more IC's in today's and tomorrow's TV. One of the advantages of using IC's is that they require so little space. This equates into smaller instruments, for consumer appeal, and also smaller chassis frames, less circuit board area, less weight, etc. Along with this goes the saving in assembly costs. Obviously, it is cheaper to mount a single unit on a board than it is to mount the dozen or more components which an IC replaces. Whether or not the IC itself is cheaper than the separate components which it replaces would be hard for us "outsiders" to know at the moment; but it is probable that improved manufacturing techniques will bring down the cost of IC's as their use becomes more general. The IC retains most of the good characteristics of the transistor- low power consumption and the at tending coolness of operation, no consumption of the material of which it is made (as contrasted with a tube filament and cathode), resistance to physical shock, etc. The IC also has about the same poor qualities as the transistor, including wide variations in characteristics as the ambient temperature is changed, susceptibility to damage from electrical overloads and arc transients, and susceptibility to dam age from excessive heat when it is installed or removed from soldered circuits. There are some limitations which are peculiar to IC's, but more on that later. Construction of an Integrated Circuit The construction of IC's has be come literally a science, and certainly no complete description of the processes involved could fit into this magazine. But a few simple drawings can, at least, provide an understanding of the basic construction technique. Suppose that we take a slab of P-type silicon material and diffuse into it some N-type impurities, as shown in Fig. 1. The depth of the diffusion will depend on the length of time the slab is exposed to the N impurities and the temperature at which it is diffused. The area of diffusion will depend on the amount of surface which is exposed and the amount which is covered with a protective mask, which can be re moved later. If leads are welded to the slab and to the N region, a useful diode has been produced. The next step is to mask all but a portion of the N surface and subject the device to another diffusion, but this time with P material. As illustrated in Fig. 2, a transistor is produced as a result of this second process. Obviously, if one transistor can be made in this manner, it is simply a matter of repetition to form a number of transistors on the same slab and then cut them apart. (As a point of interest, transistors produced simultaneously on the same slab will be quite uniform in characteristics, for reasons which should be obvious.) Just as two or more vacuum tubes often share the same glass envelope, there are many times when it is desirable to form several transistors in a single package. If the transistors produced in the manner just described were not cut apart, we would have such a "pack aged" device. But there is one problem: All the collectors would be connected together. Consequently, the device would not be very useful. To provide a means of isolating the transistors from one another, and also to provide an inactive mounting for them, one more pro cess of masking and diffusion is Fig. 1 The first step in construction of an integrated circuit is the formation of "N pockets." employed, the end product of which is shown in Fig. 3. Note that in Fig. 3 the original slab is no longer used as one of the elements of the transistors but, instead, forms a mounting base called the substrate. Because the substrate is made of P material, the junctions of the substrate and the collectors of the NPN transistors are reverse biased and, consequently, no conduction can take place between the substrate and collectors of the transistors. There is capacitance between these elements; however, it can be compensated for in much the same way that the stray capacitance of a vacuum-tube is either used to advantage or compensated for by the external circuitry. Although an array of closely similar transistors is sometimes useful, it is more often desirable to inter connect a pair of transistors within the unit, as in the Darlington pair configuration, for example. This can be accomplished by the following additional manufacturing processes. Fig. 2 By diffusing P material into the N region, the elements, or junctions, of a transistor are formed. P REGION N REGION After the transistors are formed, the points which are to be used for connections are masked. The un masked surface then is exposed to an oxygen atmosphere and baked. This forms a layer of silicon dioxide, better known as glass, which is an excellent insulator. Next, another masking operation covers all the surface except those points at which connections are de sired, and some type of metal, per haps gold, then is deposited on the unmasked areas. Finally, leads are welded to the metal-covered points, for external connections. The completed circuit is shown in Fig. 4. (The interconnection between the two collectors cannot be shown in the same cross section with the emitter-to-base connection, because it must take some other path across the surface of the chip.) Forming resistors and capacitors in the integrated circuit is somewhat simpler in concept, although not necessarily in actual practice. Fig. 5 shows diagrammatically how it can be done. On the left of the chip in Fig. 5 a resistor is formed by utilizing the resistance of a block of P material, which is isolated from the substrate by an intervening layer of N material. The size and shape of the P material determines the resistance; a long, narrow strip has higher resistance than a short, wide one. The capacitance between the de posited metal and the N material beneath the insulation forms the capacitor at the right. As with any capacitor, the value can be in creased by increasing the area of Fig 3 Formation of NPN transistors on a P substrate. The substrate can be cut to yield individual transistors, or additional processing can be performed to make an IC. the plates. In manufacturing integrated circuits, a great number of them, per haps a hundred or more, each having the equivalent of many transistors, diodes, resistors, and capacitors, are formed simultaneously on a single wafer, or substrate, a couple of inches in diameter. Obviously, the more IC's produced on the wafer, the lower the price per IC. As a rule of thumb, a 1000-ohm resistor requires about twice the area of a transistor and a 10-pf capacitor re quires about three times as much area as a transistor. For this reason, integrated circuits are designed to use as many transistors and as few passive components (resistors, capacitors and coils) as possible. This is just the opposite of conventional circuitry, in which the transistors usually are the most expensive of the four devices. Also because of cost, the values of resistance and capacitance are kept as low as possible, and often external components are used, if space permits. A common trick used to reduce the values of resistors is to stack diodes and use their junction potentials to drop voltage. For example, if 3000 ohms is required to drop 2 volts in a certain bias circuit, an area of the chip which could accommodate six transistors is required. But, by using three diode junctions in series, approximately the same voltage drop can be obtained using only the area of three transistors. Possibly a zener will be used, reducing the required area to that of a single transistor (see Fig. 6). At the present state of the art, the tolerance of resistors and capacitors in IC's are very broad, although the ratio of two resistances or capacitances can be held to relatively close specifications. Consequently, circuits are normally de signed so that absolute values of R and C are not critical. In practice, this equates, more or less, to "lots of gain and lots of negative feed back." Integrated circuits are packaged in about the same fashion as transistors, except, of course, there are many more leads coming out of the device. A common package is similar to the TO-5 transistor, but with eight to a dozen leads. They also are packaged in flat, rectangular enclosures made of plastic or ceramic with conductors extending from both long edges. The conductors are either flat or round, and sockets are available for either type. Applications in Television The IC first appeared in tele vision about five years ago. That first IC functioned as a combined 4.5-MHz amplifier, FM detector, and audio preamplifier. At present, we know of the following applications, which illustrate the design sophistication already achieved: 1) 4.5-MHz amplifier/FM detector/audio preamplifier with voltage-operated variable gain and an audio output of more than 1 volt P-P. 2) Video IF amplifier/discriminator/differential DC amplifier, with integral voltage regulator. This type is used for automatic fine tuning. 3) Chroma demodulator and color-difference amplifiers, with voltage-operated phase shift of the 3.85-MHz reference signal, to allow tint control. 4) Chroma-bandpass amplifier/ reference oscillator/burst amplifier/color killer/ACC/burst blanking, with voltage-operated gain control. 5) Complete IF amplifier with integral AGC, sound and video detectors, and video amplifier with a 6-volt output. Naturally, some of the IC's in current usage might not perform all of the functions listed above; or different combinations of functions may be performed by a certain IC. For example, a circuit designer might choose to include the reference oscillator in the same IC with the chroma demodulators. (The IC's used in the electronic tuner control of RCA's CTC 47 are simple logic devices which do not reflect the present state of the art.) Fig. 7 shows an early model IC used as a 4.5-MHz amplifier, FM detector, and audio preamplifier. It, too, is relatively simple by today's standards. As technology advances, it is likely that more integrated circuits will be developed for use in tele vision. On the other hand, there are three limitations of the IC which make the all-IC receiver seem un- likely for quite some time. First, it is difficult to build an IC which will operate with much more than 10 volts of supply, and this limits the usable output amplitude of an IC to about this peak-to peak amplitude. Secondly, heat dissipation be comes a problem in IC's, and so they are essentially low-power de vices. Third, large values of capacitance (in excess of perhaps 20 pf) are undesirable in an IC because of the size requirement and additional cost. These limitations pretty well exclude IC's from the deflection systems and the audio and video out put circuits of TV receivers. Troubleshooting As most technicians who are servicing solid-state television have discovered, there are fewer failures in them than in vacuum-tube receivers. Of course, a transistor failure might require more time to replace than a tube failure, because so many transistors are soldered in, and also because there are so few standard type transistors. Still, transistors are the most frequent cause of failure; but not because transistors are bad. For most applications, 1/2 -watt resistors are used in circuits, even though the actual dissipation might be only a fraction of this rating. Smaller resistors are no cheaper, and much of the equipment (or people, for that matter) which inserts components into circuit boards can handle the 1 / 2 -watt resistors more easily than smaller ones. Be cause of this oversizing, burned-out and off-tolerance resistors (usually caused by overheating) occur less often in solid-state circuits, particularly circuits using IC's. Fig. 4 The interconnections are made by covering the devices with an insulating layer having apertures in the desired locations and then depositing metallic conductors on top. Fig. 5 Capacitors and resistors are formed in this manner. Capacitor failures also are rare because capacitors in low-power, solid-state circuits often are operated far below their voltage ratings. Also, there has been a lot of improvement of capacitors in recent years. Because of these factors, the incidence of failure in solid-state equipment has been reduced to a level below the failure rate of com parable tube-type circuits. However, most of the failures in solid-state equipment result from faulty transistors. There haven't been enough integrated circuits used in television receivers to permit definite conclusions, but it appears that we are faced with the same situation that has prevailed in tube-type equipment-many failures will be caused directly or indirectly by the active component, the IC. Transistors are relatively cheap, simple to test, relatively easy to change, and often can be substituted. IC's are more expensive, practically impossible to test, and can not normally be replaced by some substitute type. Also, unless they are plugged in, which is rare, it is possible that the original will be destroyed when it is removed from the board. (Incidentally, plug-in IC's function normally in a computer operating in an air-conditioned environment; but contact contamination definitely has been experienced in television receivers employing plug-in IC's). Before you decide to replace an IC, several checks should be made. Just how extensive these checks are will depend on your equipment, your knowledge of the circuit in which the IC is located, and the trade-off between your labor cost and the cost of the IC. First of all, be sure that the IC actually is receiving the supply voltage it is supposed to receive. A shorted zener diode used as a regulator, or an off-tolerance power supply bleeder, might be supplying insufficient voltage; IC's are rather critical in this respect. Also, an open zener or some other fault external to the IC might increase the supply voltage to a level that will destroy the new IC as soon as it is installed. The next item on the pre-replacement check list seems obvious, but it is too important to be ignored. Be sure that the input signal actually is getting to the input terminal of the IC. It is possible that a coupling capacitor has opened, or that there is a defective solder connection or a cracked board. Defective solder connections have been known to turn up months, or even years, after manufacture. Plated through holes in circuit boards are still a bit new, and it is not uncommon for the through plating to be open. Usually there will be several terminals of the IC which are connected to ground via a resistor and bypass capacitor. It is fairly certain that a transistor emitter inside the IC is connected to this terminal. If either of these components is open or shorted, the transistor to which they are connected will develop little or no gain. Another item to be checked is the circuit which the IC is driving. If it uses a transistor, perhaps it is tie one which has failed. Inject a signal at a convenient point near the output of the IC; if the test signal doesn't pass through the rest of the system, it is a safe bet that the output of the IC cannot either. IC Replacement If the IC is mounted in a socket, replacement is simple, but there are a few tips which might save you time and money: • First, don't plug in another IC until you are certain that the sup ply voltage is correct. Excessive supply voltage seldom ruins a tube immediately, but it is likely to "zap" a transistor or IC before you know what happened. • Don't forget that the trouble might be socket and pin corrosion. It is possible that a new IC will fix the trouble, but the old one might too, if it is reinserted. If you do not have a replacement, try removing and inserting the suspected IC a couple of times before ordering a new one. Since IC's have come along, we have added another subject to our "do-not-argue about these-things" list. How to get an IC off the board is almost as inflammatory as discussions of religion and politics. Everybody has a pet method which he is prepared to defend against all others. Here are some of the leading contenders: One approach is simply to cut off all the leads and then remove the ends which are soldered through the board, one at a time. This works rather well with IC's which have long leads, but not so well with the plastic, in-line types which are mounted tightly to the board. As a rule, the IC is destroyed; but this is unimportant unless the IC was actually good, or if it is a warranty item and the manufacturer wants it back. (As a point of interest, I have always maintained that the customer rightfully should pay for any parts which are destroyed through no fault of the technician, or as a calculated risk-but it does run up the repair bill.) Another popular method of IC removal involves using braided ground strapping as a wick to draw off molten solder from around the terminals of the IC. Braid designed specifically for this purpose also is available. However, extreme caution is needed, because it requires a lot of heat to melt the solder and also jet the braid hot enough to carry the solder away. It is easy to get the printed wiring on the board so hot that the bonding between the copper and the board melts and the two separate. Nevertheless, some technicians become very adept and can perform a very neat job with this method. The flat-surface soldering iron, which is designed to heat all the contacts simultaneously, works well for removing AFC-diode packs having three leads. But this method sometimes gets difficult with IC's having up to 14 heads. Either the iron is too large or not large enough. Of course, if all integrated circuits had the same lead configuration there wouldn't be any problem with this method. Another approach uses a soldering iron with a tube and squeeze bulb attached. The idea is to get the terminal hot and then pump the bulb to blow away the solder. The one we tried didn't seem to work too well, but maybe with more practice we could have mastered the trick. One tool that seems to be relatively effective consists of a spring loaded piston inside a small cylinder which has a Teflon nozzle on one end. The piston is pushed down and the nozzle is held as near as possible to the connection to be un soldered. When the solder melts, the piston is released and, hope fully it draws up the solder. If there is room to operate, it works fairly well, but usually one has to repeat the process on some of the IC terminals. An uncle of mine has had a small air compressor in his shop for the past twenty years. He uses it to blow the dust out of cabinets, clean off the bench, etc. Now he uses it to blow the solder out of holes in boards. To prevent spraying solder all over the chassis, he wads a paper towel around mil beneath the reverse side of the area where he is working. Fig. 6 A voltage drop can be obtained by forming a resistor, a series of diodes, or a zener. Cost is a major consideration in determining which method is used. Fig. 7 The RCA type CA 3013 amplifier and discriminator, with typical external circuitry. Summary In the next few years, it can be assumed that the IC will be applied to many more circuits in television receivers. The reasons for this assumption are: simplified assembly, smaller and lighter chassis, reduced cost and performance equal to, or better than, tube and transistor circuits. At the moment, the production of the types of IC's suitable for most TV applications is, relatively speaking, in its infancy. The techniques for manufacturing IC's are being constantly revised; however, basic construction techniques probably will remain similar to those described in this article. From the servicing angle, it isn't really necessary to know precisely how an IC is made, or for that matter, precisely how it functions. It is important, however, to know thoroughly what each IC in a receiver does. Equally important, the technician must know the purpose of each component connected to an IC. If he does not, he will make a lot of wrong guesses, replacing good IC's needlessly and damaging new ones by installing them in defective circuits. Good soldering techniques must be employed when you are removing IC's. Whether or not an IC found to be good should be rein stalled is questionable-putting it back might be more costly than leaving the new one in place. But ruining the circuit board while re moving it is another matter! A patched up board can lead to future failures, and the alternative-re placing the board--is too costly even to consider in most cases. A sign we noticed in a shop not too long ago sums it up. "If you're too busy to do it right the first time, how will you find time to do it over?"
|
| Prev | Next | Index | HOME |