NOTHING is more annoying to the installer than to wire good equipment and hear bad sound come out. Assuming that the equipment is of reasonable quality, in good condition and properly connected, the trouble is probably due to some sort of noise or interference getting into the signal channel, where it is amplified along with the desired program.
This is not a troubleshooting handbook, therefore we will (with a few exceptions) refrain from dealing with the correction of troubles within a single unit of equipment, such as an amplifier.
The aspects of noise and interference treated here are those that arise from external causes, such as diathermy equipment, nearby radio transmitters and power lines. Incorrect connections between units, such as might occur in the course of wiring an installation, are also discussed.
Hum
Hum is seldom a problem in any single unit of good manufactured equipment, but it can show up when several units are interconnected into a complete system.
Incidentally, by hum we mean the fundamental power-line frequency or one or more of its higher harmonics. Thus, for a 60-cycle supply line, this would mean a 60-, 120-, 180-, or 240- cycle signal in the output with no program input.
If hum shows up, disconnect the preamplifier and all other units supplying the signal to the power amplifier, and check the output again with only the power amplifier running. If the preamp and power amplifier are built as an integral unit, remove the last tube in the signal path ahead of the stage where the overall feedback loop begins (if there is such a feedback loop in the unit). If the hum persists with only the power amplifier on, try adjusting the hum bucking control (if the unit has one).
Such controls are often found misadjusted after shipment or stockroom storage. Also try reversing the position of the power-cord plug at the wall socket, and the addition of a direct ground from the power amplifier chassis to a water pipe or to the metal box containing the electric outlet in the wall. If none of these measures removes the hum, try replacing the tubes in the power amplifier one at a time. A tube can become defective because of shock or vibration, even when the equipment is not in service, although it happens rarely.
If the hum disappears with all units disconnected, (except the power amplifier) reconnect the preamplifier only and turn it on.
If the hum returns, look at the settings of the controls. With the bass control at the flat setting and the volume control turned to its lowest setting, hum and hiss should be little more than barely audible when listening within a foot or so of the speaker in a fairly quiet room. If the noise level is acceptable with the volume control at minimum but comes up when the control is advanced, the hum or hiss is coming from an early stage. With some types of input connector, hum may be excessive with the gain of the amplifier at maximum and nothing connected to the input. Try shorting the input, or shielding it with an appropriate cable connector mated to the chassis connector.
By following these general procedures with each unit as it is connected into the system, the unwanted noise can usually be traced to the source, whether it be in the preamplifier, AM or FM tuner, pickup or other input signal source.
Ground Loop
One frequent cause of a fairly low-level but hard-to-locate hum is the condition called a ground loop. Hum from this cause shows up only when two or more units are connected together into a system. It arises when there is a fairly strong 60-cycle ac induction field present in the room (this condition is extremely common), and there are two or more ac ground connections to the hi-fi system. What often throws the installer off the track and into a head-scratching frenzy is the fact that frequently, at least one of the ac ground paths is a hidden one, perhaps through the POWER AMPLIFIER capacitance between the power-transformer primary winding and the core, which is usually grounded to the chassis. A look at Fig. 701 will show how the hum voltage arises. When the power amplifier and preamplifier are at some distance from each other and plugged into separate wall outlets, hum can arise because of the two chassis and the power line itself. Note that even one or two millivolts of signal developed across the impedance of the interconnecting cable shield will appear as a signal at the input to the power amplifier.
Fig. 701. Hum can arise in a system because of ground loops through
interconnecting cables.
The closed loop can be formed in any of the several ways shown in the drawing: through the primary-to-core capacitance of the transformers, the noise-bypass capacitors often wired from one side of the primary to ground, or even through the capacitance of the secondary (not shown). In all cases, the remedy is the same: get rid of the closed loop in which the induced voltage appears. One effective way to do this is to plug one unit into a convenience outlet on the other. In most cases, this is most convenient because it permits turning both units on and off together.
Another step is advisable for safety reasons, and often necessary, anyhow, to eliminate stubborn traces of hum. This is to make sure that the power-cord plugs are inserted in the wall sockets with such polarity that both the noise bypass capacitors across the transformer primaries are effectively connected from the cold (grounded) side of the ac line to the unit chassis. If this is not done, the chassis of the unit concerned will be at line voltage above ground through the noise bypass capacitor, and will there fore be a shock hazard. The risk of shock with the reversed connection can be removed by providing a direct, solid ground from the chassis to the conduit or metal shell of the ac outlet box.
However, this brings in the possibility of hum pickup again, and there is no assurance that the ground wire will not be disconnected accidentally when furniture is moved for cleaning, reviving the shock hazard. A recent (and laudable) trend among appliance makers has been toward the use of a 3-wire power cord, with a 3-pronged plug that automatically grounds the device when the plug is inserted in a wall socket of a matching type.
Unfortunately, this type of cord and plug is not yet widely used in, the audio industry, probably because of some inertia, and the fact that most household electrical outlets are still the ordinary 2-pronged type.
If the client will authorize the slight additional expense, re placement of the standard power cord on a piece of equipment with a new cord of the 3-wire type, complete with a 3-pronged plug, is a very good solution. Even if the wall socket is not of the 3-hole grounding kind, a simple adapter can be inserted that has an exterior pigtail attached to its grounding socket. This pigtail is permanently grounded under the screw holding the cover plate on the outlet box, and the adapter is left in place even when the equipment is unplugged. The only joker here, of course, is the assumption that all outlet boxes are grounded. They should be, but this doesn't mean that they are. You can check easily enough with the help of a test lamp or ac voltmeter. Attach one lead to the plate. Insert the other lead into one, then the other, prong of the outlet. If you get no indication, the plate isn't grounded.
If this procedure is not acceptable to the client, he may agree to a change of the power-cord plug to the type having one of the flat prongs wider than the other. In properly wired household electrical outlets, the grounded side of the 115-volt ac line is wired to the wider of the two slots in the receptacle. If the plug is wired to the cord in such a way that the hot side of the line goes directly to the on-off switch in the equipment, the unit will definitely be safe when the switch is in the "off" condition.
One solution to the shock hazard problem in cases where a direct ground is not feasible is to clip out the noise bypass capacitor entirely. This usually reduces the shock hazard to negligible proportions, even when no direct chassis ground is provided, since the capacitance from transformer primary to core is very much smaller than the .01 or .005-pf capacitor often included for noise suppression. If noise trouble is encountered, a choke-and-capacitor-type line filter installed at the wall socket, with the case and ground terminal solidly grounded to the ac conduit or box, is the best solution to both the shock and noise problems.
Hiss
Little can be said here about curing hiss problems, since they usually indicate a defective resistor in an early stage of a high-gain system and, as such, are a troubleshooting problem inside the unit. Heater-to-cathode leakage in a preamplifier tube may cause ...
Fig. 702. Overdriving the grid of a low-level stage generates distortion
that cannot be corrected by adjustments in later stages.
... hiss, but it is almost invariably accompanied by a generous helping of hum, except in those rare cases where the preamp heaters are supplied with dc. Nevertheless, it is worth trying a tube change if hiss is excessive.
In some otherwise good preamplifier designs, a 1/2-watt carbon resistor will be found in the plate or grid circuit of an early stage.
Sometimes these cause no trouble, but all that's needed to make many of them a prolific source of hiss is a little too lengthy application of the soldering iron during assembly.
Overdriving
Overloading is another of those words that doesn't fit very well the idea it's meant to convey. As commonly used, it means that the amplitude of the input signal applied to a given circuit point exceeds the amplitude that the stage can amplify linearly.
A better word to describe this condition might be overdriving.
Such overdriving results in clipping of one or both peaks of the signal wave and plenty of distortion, even though every unit in the system is in perfect shape (Fig. 702). Usually it arises when a high-output device such as a crystal pickup of an FM tuner is connected to a preamplifier input meant for a low-level input.
Backing down the gain of the preamplifier will do no good, for the gain control is seldom between the input and the first grid, which is where the clipping takes place. Obviously, reducing the gain at a point beyond the source of the distortion cannot cure the trouble.
The answer, of course, is to knock down the input signal to a level that the stage can handle. At one demonstration of a specially-built preamplifier for the Audio Engineering Society, half the program was presented with raucous distortion while some of the better brains of the industry frantically switched tubes and examined pickups through jeweler's loupes. You guessed it; the specially-built preamplifier output was overdriving the input to the power amplifier all the time. A hastily inserted pad knocked the 5-volt output of the preamp down to the 1 volt or less the power amplifier was designed to take, and all was serene again, although the blushes took some time to subside.
Man-made interference and external electrical noise It is customary in electronic work to distinguish between noise and interference in some manner, although the distinction is often not an easy one to make. For our purposes, we can treat as interference all electronic disturbances that are deliberately generated to perform some desired function in some service (such as transmissions in mobile radio) but which also get into the equipment used by other services in such a way as to interfere with the normal functioning of that service. This broad definition is made necessary by the present extremely varied use of electrical and electronic equipment, and the certain expansion of such use in the future.
Signals generated by radio transmitters, electronic heating equipment used by doctors and by industry, and even signals generated internally in television and other receivers, can some times interfere in hi-fi and public-address equipment. This may seem a bit weird, since the interfering radio transmissions are made on frequencies far above the audio range. The culprit turns out to be our old friend nonlinearity, which comes about because of overdriving somewhere, the same process previously described.
Almost without exception, this overdriving and the resulting nonlinear operation that results in interference, occur in the first stage of a high-gain amplifier, although the interfering signal may be picked up on the transducer supplying the normal input signal to such a stage, or on the interconnecting cable, even if it is of the shielded type. Once the interfering signal reaches the grid at sufficient amplitude, the process is simple. Whenever the grid is swung by the interfering signal into a voltage region where the plate waveshape is not a faithful linear reproduction, detection takes place, just as in the detector of any radio receiver.
Any modulation on the interfering signal will now be heard through the hi-fi or PA system with a degree of clarity that can he both startling and embarrassing.
The remedy in most cases is simple in principle and may require either or both of two expedients.
Fig. 703. A low-pass rf filter at the input
grid will solve many interference problems.
The first is to prevent the offending signal from entering the hi-fi equipment, and the second is to bypass it to ground before it can reach the grid of the first tube. Preventing the signal from entering the equipment can usually be accomplished by thorough shielding of all input signal sources (such as a phono pickup or tape-recorder playback head) , with a shield that has a very low impedance path to ground for rf voltages. Unfortunately, this is sometimes easier said than done, particularly in areas where the interfering signal is very strong.
The difficulty arises in trying to shield the transducer effectively without interfering with its normal function. Often, the most complete shielding that can be practically provided will still permit a certain amount of interference pickup to occur. Fortunately, the insertion of a simple low-pass rf filter as near the grid of the tube as possible will usually remove the last traces of the interference that the shielding does not keep out.
Filters
Suitable filter configurations are shown in Fig. 703. The cutoff frequency of such a filter should be well above the audio-frequency range for most kinds of interference. A cutoff even as high as 100 kilocycles is usually just as effective as a lower cutoff and obviates any fear that high-frequency response will suffer.
It is necessary to mention one type of interference that definitely cannot be removed by such a filter: interference from the horizontal oscillator frequency of a nearby television receiver, which occurs at 15,750 cycles and innumerable harmonics thereof.
A great many TV receivers unfortunately have little or no pro vision to prevent radiation of this interference. If you can get at the offending set, a grounded shield placed around the tube and tuned circuit of the horizontal oscillator and as much of the yoke wiring as possible, may help considerably. Shielding the pickup or other input device at the hi-fi set will also help, and a line filter between the wall ac outlet and the power cord for the hi-fi may also improve things. It may also be necessary to install a shield braid over the power cord itself, grounded at the wall end and to the power supply chassis.
Only as a last desperate measure should a low-pass filter that cuts off below 15,750 cycles be installed, because this definitely does impair the high-frequency response which is one of the features of hi-fi equipment the customer pays good money to get.
In connection with interference from radio transmitters, particularly those operated by radio amateurs, it is important to bear in mind that all radio services are licensed by the FCC and have a legal right to operate. Owners and operators of licensed radio equipment and services are required by law to cooperate in eliminating interference, but this does not mean they are required to suspend operation. Almost invariably, even the most stubborn and difficult interference problems can be solved by cooperation between the parties concerned.
It is necessary to mention one other way in which interference can arise (although it is too large a subject to be covered here).
When the signal feeding the hi-fi system is the audio output from an AM or FM tuner or other type of radio program receiving equipment, the tuner itself may experience direct radio-frequency interference on the channel to which it is tuned. This may arise from a number of causes. If the offending signal is entering in this manner, it may be necessary to consult a book devoted entirely to the subject of interference elimination to solve the problem.
Electronic and acoustic feedback
By feedback, we mean here unwanted feedback in a hi-fi system, often regenerative, which can cause drastic distortion and even oscillation of the whole system. Unwanted feedback of the electronic or electrical sort can occur when output and input cables are laid close to each other for a distance, or when one or both lines are not shielded. The remedy is usually just a matter of providing adequate separation of all inputs from all outputs, and perhaps the addition of shielding to the speaker line, with the shield grounded to the power amplifier chassis (but not to a building ground at the other end!). This latter expedient is more likely to be required in systems using a very powerful power amplifier, run at high level, and perhaps an unshielded 500-ohm output line feeding distant speakers.
Acoustic feedback refers to actual sound vibrations, sometimes through the air but more often through solid materials like the wood cabinets. The output of the speaker reaches some sensitive part of the equipment by one of the phase shifts involved and proceeds to ruin reproduction. The two most likely places for the vibration to re-enter the system are through a vacuum tube at the input (where the gain following the stage is very high), or through the phono turntable, the material of the record, and the pickup itself.
Heavy mechanical vibration of the preamplifier chassis can cause the grid-to-cathode spacing of a tube to vary with the vibration and thus inject an electrical signal equivalent to the vibration waveform into the signal path. In a tube with loose mica spacers or some similar defect, this effect will be much more severe, which is why some tubes that seem perfect when tried in a tester free of vibration give bad results in actual service. Assuming that no defective tube is found, the only remedy is to provide vibration isolation between the acoustic output and the place where the vibration is re-entering the system. This may mean putting padding under one or more electronic units (but don't block the paths for ventilation), removing them to another cabinet, or even mounting the speaker and its enclosure on shock mounts. The direct airborne sound from the speaker will seldom vibrate even the most sensitive tube hard enough to cause this kind of feedback, unless the amplifier is right in front of the speaker cone, where it has no business to be.
Without doubt, the most common cause of acoustic feedback is vibration from the speaker enclosure traveling through a solid path to the turntable, where the vibration can actuate the pickup to inject the feedback signal at the point in the whole system where the following gain is usually highest. Here again, the solution is to provide isolation in the solid path. This may be done by means of foam rubber or similar padding under and behind the speaker enclosure, the turntable base or both. When both are housed in the same cabinet, the problem may be a little more severe, but even here sufficient isolation can usually be attained to prevent trouble. Stereo systems seem to be more prone to this difficulty than straight monophonic systems, probably because of the sensitivity of the stereo pickup to a vertical component of motion as well as horizontal.
If you suspect that a small amount of acoustic feedback is causing distortion, but the feedback is not great enough to make the fact a certainty, it is advisable to make the following test: monitor the output of the amplifier with a good pair of head phones bridged across the speaker line, while the speaker load and an equivalent resistor dummy load are alternately switched across the output. If, when the speaker is connected, the suspicious sound increases on large low-frequency sounds like the thump of a bass drum and disappears when the resistor is acting as the load, you've got acoustic feedback.