The chief consideration in any hi-fi installation is an evaluation of how the components work together as a team. We are concerned with the overall frequency response, overall harmonic (or intermodulation) distortion, and the overall transient response.
Distortion- and transient-response tests are usually made at maximum rated power output. In case the system performance is unsatisfactory in some respect, the technician must make more detailed tests to locate the component that is causing the difficulty.
If a system has been in use for an extended period of time, it is possible that more than one component may have deteriorated past acceptable limits. A hi-fi system commonly includes an fm/am tuner, pre amplifier, multiplex adapter, power amplifier, speakers, record player, and a tape recorder or tape deck.
In this guide, the coverage does not include record and tape equipment, since these are specialized units.
Interested readers are referred to record-changer and tape-recorder servicing guides.
Common trouble symptoms associated with system operation are:
1. Stereo-reproduction failure or malfunctioning.
2. Poor bass or treble response.
3. Distorted reproduction.
4. High noise level.
5. Interference on fm or AM function.
6. Weak output.
7. Microphonic ringing or how ling.
GENERAL DISCUSSION
A typical hi-fi stereo system is illustrated in Fig. 7-1. Any system test requires the application of a suitable test signal at the input, and connection of a suitable indicating instrument at the output. For ex ample, the AM tuner function would be energized by an AM generator, and the amplifier output would be indicated by a VOM, VTVM, TVM, or scope. Since an AM generator usually provides only one tone signal, a check of the system frequency response re quires that the generator be externally modulated by an audio oscillator, as shown in Fig. 7-2. To mea sure the system harmonic distortion, a harmonic distortion meter is used in place of the VTVM, as shown in Fig. 7-3. If it is desired to measure the percentage of harmonic distortion at a frequency other than the modulating frequency provided by the generator, an audio oscillator may be used to externally modulate the generator.
To measure the system intermodulation distortion, an IM distortion meter is employed, as shown in Fig. 7-4. In most cases, the IM signal source is built into the IM distortion meter. If a tone-burst system test is to be made, the test setup shown in Fig. 7-5 is employed. Again, a system square-wave test is made with the test setup shown in Fig. 7-6. The foregoing tests require an accurate AM generator with low distortion. In addition, an audio oscillator used in the tests must have low distortion, and the external modulation function must be linear. Otherwise, distortions introduced by the test instruments would be falsely charged to the hi-fi system. It is good practice to use lab-type instruments in hi-fi test work. If service-type generating equipment carries hi-fi ratings, it will be satisfactory.
Next, if system tests are to be made through the fm tuner, the same instrument arrangements are used, except that an fm signal generator is used in- stead of an AM signal generator. Since most hi-fi systems include a multiplex adapter, the generator needs to be of the stereo-fm signal simulator type. To make an overall frequency-response test, the signal simulator may be externally modulated by an audio oscillator, or by an audio sweep generator. A test setup with an audio sweep generator is shown in Fig. 7-7. Since audio sweep generators (Fig. 7-8) are lab-type instruments and are comparatively expensive, most hi-fi shops prefer to use conventional audio oscillators. A stereo-fm signal simulator used with either type of audio generator must be quite linear and rated for hi-fi application.
Fig. 7-1. Typical stereo system arrangement.
Fig. 7-2. Test setup for checking system frequency response.
Fig. 7-3. Test setup for checking system harmonic distortion.
Fig. 7-4. Test setup for checking system IM distortion.
Fig. 7-5. Test setup for making system tone-burst test.
Fig. 7-6. Test setup for making system square-wave test.
Fig. 7-7. Sweep-frequency test of multiplex system. (A) Test setup. (B) Scope
display of typical response pattern.
Fig. 7-8. A typical audio sweep generator.
Fig. 7-9. Effect of de-emphasis on multiplex frequencies.
In the example of Fig. 7-7, the system frequency response will seem to be poor, even if we vary the setting of the treble control from its normal mid range position. The reason for this is due to the effect of de-emphasis on the audio test signal, as shown in Fig. 7-9. In other words, an fm broadcast signal is pre-emphasized, whereas the output from an audio generator is flat. De-emphasis is provided in the out put section of the multiplex adapter. Therefore, we normally find a flat frequency response at the input of the adapter, whereas we will find a frequency response that is 20 dB down at 20 kHz when testing at the output of the adapter. In other words, the normal frequency response at the outputs of the L and R amplifiers will start to drop off at 1 kHz, and will be 20 dB down at 20 kHz. The response at 20 kHz is 10 percent of the peak-to-peak response at 1 kHz.
Let us suppose that the frequency response is abnormal at the output of the R amplifier in Fig. 7-7.
The next step is to check the frequency response at the output of the L amplifier. In case a normal frequency response is found at the output of the L amplifier, we conclude that there is a defect in the R amplifier; amplifier troubleshooting is explained in Section 5. If we find the same abnormal response at the output of the L amplifier, we know that the trouble is occurring in the fm tuner or in the multiplex adapter. To localize the trouble, we transfer the vertical-input leads of the scope to the output of the fm tuner. If we find a flat frequency response in this test, we know that the trouble source is in the adapter; on the other hand, an abnormal frequency response indicates that the trouble is in the fm tuner.
Now, with the faulty unit known, further trouble shooting can be done; tuner troubleshooting is explained in Section 2, and multiplex-adapter trouble shooting is explained in Section 3.
When making a system harmonic-distortion test through an fm tuner, or through an fm tuner and multiplex adapter, it is advisable to employ a 400-Hz test frequency. Since the second and third harmonics of this frequency are 800 and 1200 Hz, respectively, the de-emphasis network has little effect on the reading of the harmonic-distortion meter. On the other hand, if a higher test frequency were used, the distortion reading would be abnormally low. It is difficult to apply correction factors, because the proportion of second to third harmonics is unknown in a routine measurement. Therefore, the practical procedure is to confine a system test to a frequency of 400 Hz, or lower if desired. In case the system percentage of distortion is unduly high at the output of the L channel, check next at the output of the R channel. An acceptably low distortion figure at the output of the R channel indicates that the trouble will be found in the L amplifier. On the other hand, if the distortion readings are too high at the outputs of both amplifiers, we conclude that the trouble is occurring in the fm tuner, or in the multiplex adapter.
Distortion measurements should be made at maximum rated power output. Apprentice technicians should not confuse a music-power rating with a steady-power output rating. A music-power rating is always higher, and specifies the ability of an output amplifier to process a pulse-type waveform. If the output amplifier is driven by a sine-wave signal and is operated at the music-power rating, a transistor type amplifier may be damaged. That is, a transistor amplifier should not be tested at an output greater than the steady-power rating when driving the sys tem from an audio oscillator. A tube-type amplifier will not be damaged in a sine-wave test if operated at its music-power output rating, but the distortion will be excessive. That is, the sine-wave distortion increases rapidly as the steady-power output rating is exceeded.
Intermodulation-distortion tests of a hi-fi system through the fm tuner are not practical unless the deemphasis network is bypassed. That is, typical IM analyzers employ test frequencies of 60 Hz and 6 kHz. Although the 60-Hz signal is passed without attenuation, the 6-kHz signal is normally attenuated about 10 dB. This action of the de-emphasis network makes the reading of an IM analyzer virtually meaningless. Depending upon the design of the multiplex section, it may or may not be practical to "jump" the de-emphasis circuit (s) for the purpose of making a system IM test. In any case, a percentage IM measurement can be made at the output of the fm tuner (input of the multiplex section). Next, the IM test signal can be applied at the input of the pre amplifier to check the sections past the multiplex adapter. In a situation of this type, our check of distortion in the multiplex section is limited to a harmonic-distortion measurement. As noted above, the test frequency should be 400 Hz, or a lower value, if desired.
System distortion tests are informative because we cannot necessarily predict the amount of system distortion from a knowledge of the component distortion values alone. In other words, there are some situations in which the distortion produced by one component is partially compensated by the distortion produced by the following component. For example, if a preamplifier has 1 percent distortion, and an output amplifier has 1 percent distortion, we cannot assume that the combination will have a total distortion of 2 percent. Each component has a particular transfer (input/output) characteristic, and it is the system transfer characteristic that determines the amount of system distortion. If one amplifier has a concave transfer characteristic, and the following amplifier has a convex transfer characteristic, the system characteristic may be more nearly linear than either of the individual characteristics. Fig. 7-10 illustrates the principle that is involved.
Fig. 7-10. How oppositely curved transfer characteristics tend to linearize
a system.
Apprentice technicians sometimes assume that a test record or a test tape is equivalent to an audio oscillator for checking amplifier distortion. This signal source has valid uses, but gives higher distortion readings than if an amplifier system is driven by an audio oscillator. The reason for this disparity is seen in Chart 7-1; that is, conventional recording systems are associated with 2 to 4 percent distortion, and the same order of distortion is associated with playback.
Since a good audio oscillator such as that used in a hi-fi shop has less than 0.5 percent distortion, it is the most accurate signal source for use in distortion measurements.
-------------
Chart 7-1. Typical Harmonic-Distortion Values
Frequency-Modulated Transmitters 100% modulation-5 to 7% 80% modulation-2 to 3% Disc-Recording Systems record/playback-2.5 to 4% Audio Amplifiers High quality-0.10 to l % Medium quality-2 to 5% Magnetic Recorders and Reproducers record/playback-2 to 4%
-------------
A system square-wave test such as that shown in Fig. 7-6 is straightforward so far as the AM function is concerned. On the other hand, if the square wave signal is used to modulate an fm generator for driving an fm tuner, we are confronted with the same basic de-emphasis problem that was noted previously in regard to IM tests. That is, a square-wave signal differs basically from a sine wave in that the square waveform is built up from a number of different frequencies. Therefore, system square-wave tests through the fm tuner are not practical unless the de-emphasis circuit(s) can be "jumped" during the test. In any case, a square-wave check can be made at the output of the fm tuner (input of the multiplex section). Next, the square-wave test signal can be applied at the input of the preamplifier to check the sections past the multiplex adapter. The same general procedure should be observed in making tone-burst tests through an fm tuner.
Square-wave tests differ from frequency-response tests in that a square-wave test takes the transient response of the system into account. The transient response is fundamentally dependent upon the phase characteristics of the various components. If each component has a linear phase characteristic, the out put square-wave will be free from transient distortion. However, most components have phase characteristics that are more or less nonlinear. Fig. 7-11 shows the phase characteristic for a simple tuned circuit, as in an AM tuner. Phase-shift charts for RC-coupled amplifiers are shown in Fig. 7-12. Experienced technicians know that phase characteristics combine in an entirely different manner from that of frequency characteristics. Let us consider the basic principles that are involved.
With respect to a system frequency response, we recall that the output of an amplifier is multiplied by the gain of the following amplifier at any given frequency. For example, if we apply 0.1 volt at 1 kHz to an amplifier that has a gain of 10 times at this frequency, the output will be 1 volt. Next, if this out put signal is fed to a following amplifier that has a gain of 6 times at 1 kHz, the output will be 6 volts.
In other words, the system gain is 6 X 10 or 60 times at 1 kHz. If we apply 0.1 volt at the input of the system, we will obtain 60 X 0.1 volt, or 6 volts out put at 1 kHz. However, phase characteristics do not combine in a simple or readily predictable manner.
Fig.
7-11. Frequency and phase characteristics of a tuned circuit, as in an
AM tuner. (B) For high frequencies.
Fig. 7-12. Universal phase-shift charts for RC-coupled amplifier at low
and high frequencies.
Fig. 7-13. Distortion of square wave by circuit under test. (A) Undistorted
test wave. (B) Low-frequency attenuation. (C) High-frequency attenuation.
(E) Lagging low-frequency phase shift. (G) Combination low-frequency attenuation
and leading low-frequency phase shift (differentiation) . (D) Leading low-frequency
phase shift. (F) Transient oscillation (ringing). (H) Combination high-frequency
attenuation and lagging low-frequency phase shift.
In general, it is much more difficult to compensate a nonlinear phase characteristic than to compensate a nonlinear amplitude characteristic. Rapid curvature in the phase characteristic of any component causes more or less overshoot, often accompanied by ringing. The basic square-wave distortions are shown in Fig. 7-13.
A square-wave test of a hi-fi system must take the tuner bandwidth into consideration, as this is normally the limiting factor. In the case of an AM tuner, the normal bandwidth is in the range from 5 kHz to 10 kHz. Since it is a rule of thumb that about 15 harmonics (odd harmonics) should be reproduced in a square-wave test, the repetition rate of the square wave will be chosen in the range from 20 to 200 Hz. In the case of an fm tuner, a square-wave repetition rate up to 400 Hz is suitable. The tone controls should be set for flat frequency response when making a square-wave test. If the system is unsatisfactory, the scope can be used as a signal tracer to localize the defective component. That is, the vertical-input leads can be transferred to the output of the preamplifier, and then to the output of the tuner.
ANALYSIS OF COMMON SYMPTOMS
An analysis of the common trouble symptoms listed earlier in this Section is presented in this section.
1. Stereo Reproduction Failure or Malfunctioning
Poor stereo reproduction or stereo failure is not necessarily due to a component defect. For example, a broken antenna lead can reduce the strength of a stereo-broadcast signal to a sufficiently low level that the subcarrier oscillator in the multiplex section is not tightly locked. In case the system stereo reproduction is unsatisfactory when checked with a generator, the antenna is eliminated from suspicion. It is often helpful to carefully check the system operation on each function; this evaluation can provide preliminary localization of the trouble.
Possible causes of stereo reproduction failure or malfunctioning are:
a. Weak output from fm tuner; troubleshooting procedures are discussed in Section 2.
b. Defect in multiplex section; troubleshooting procedures are explained in Section 3.
c. Subnormal output or no output from one of the audio channels; audio troubleshooting is covered in Section 5.
d. One speaker defective; speaker operation is discussed in Section 6.
e. Defective lead or plug to one speaker (see Fig. 7-1); less likely than other faults, but possible.
2. Poor Bass or Treble Response
When a hi-fi system develops poor bass or treble response, the first step is to check the response on each function as an aid to preliminary localization.
For example, if we find poor bass response on the phono function, but have normal response on the AM, fm, and tape functions, we will suspect that there is a defect in the equalization network of the phono-input section. On the other hand, if we find poor treble response on all functions, it is logical to conclude that the trouble will be found in the audio section. The speakers are included in the general audio section; thus, if the brilliance-control network in the speaker circuit has a defective component, poor bass or treble response can result. Although subnormal bass or treble response is the most common type of "tone" trouble symptom, the reverse situation occurs occasionally. That is, a defect in a negative-feedback loop that produces a midband "suck-out" is associated with abnormal bass and treble response-in other words, the mid-range response is poor when this malfunction occurs.
Possible causes of poor bass or treble response are as follows:
a. Tuner misalignment (see Sections 1 and 2).
b. Component defect in an equalization network.
c. Preamplifier malfunction.
d. Output amplifier malfunction.
e. Defective capacitor in crossover network (see Section 6).
f. Defective component in brilliance-control network.
3. Distorted Reproduction
If a hi-fi system develops distorted reproduction, we start the trouble analysis by checking the system response on each function. In a stereo installation, distortion may develop in only one channel, or in both channels; this observation is also helpful in preliminary localization. Occasionally, we can find a localization clue in the type of distortion that is occurring. For example, if there is objectionable hum modulation or interference, we turn our attention to the power supply at the outset. Acoustic-feedback (microphonic) distortion can often be localized to a system component by rapping your knuckles against each component in turn. If a distortion symptom disappears when the volume control is turned down considerably, we know that the trouble will be found in a section following the volume control.
Possible causes of distorted reproduction in a hi-fi system are:
a. Defective component in the output amplifier (see Section 5).
b. Speaker malfunction (see Section 6).
c. Component breakdown in preamplifier.
d. Leaky capacitor, defective transistor, or mis-alignment in tuner section.
e. Mismatch of output amplifier to a replacement speaker.
4. High Noise Level
A high noise level directs attention to the input sections of a hi-fi system because the noise voltage then becomes amplified by the following sections.
If noise occurs only on the phono function, for example, most of the system circuitry is immediately cleared from suspicion. On the other hand, we often encounter situations in which the noise level is the same on all functions. In this case, the preamplifier becomes the most logical suspect. Although various components can develop noise voltages, case histories show that transistors are most likely to become defective in this manner, and that poor connections are a runner-up. Resistors of the composition type may become noisy, particularly in low-level stages.
Possible causes of a high noise level in a hi-fi system are as follows:
a. Extraneous noise voltages feeding in from the power line; check the line filter in the power supply.
b. Plug not fully inserted into connector.
c. Cold-solder or high-resistance connection.
d. Noisy component, such as a transistor, resistor, or capacitor.
e. Broken antenna conductor (noise occurs only on AM or fm functions.
5. Interference on FM or AM Function
When investigating a complaint of interference on the fm or AM function of a hi-fi system, it is essential to determine the reception conditions that prevail in the particular area. For example, if the installation is in the vicinity of a high-powered broadcast station, other receivers in the same neighborhood will also be subject to interference. In a typical case history, all receivers in the vicinity of a high-powered radar installation were subject to brute-force interference. Environmental interference can be reduced by means of antenna and line traps, but complete elimination may require operation in a screened room. In difficult situations of this kind, the customer should be encouraged to discuss interference problems with his neighbors. This procedure serves to eliminate customer suspicion that the technician is "trying to sell a bill of goods."
Fig. 7-14. Location of radio noise.
Fig. 7-15. Block diagram for a professional interference-locating unit.
In the event that a complaint of interference on the fm or AM function of a hi-fi system is legitimate, possible causes of the malfunction are as follows:
a. Customer may be using the power line as an antenna via a capacitive-coupling device. This practice is conducive to noisy reception and various forms of interference.
b. An excessively large or high antenna may be in use; most tuners are designed to operate normally with comparatively short and low antennas.
c. Mobile installations are subject to ignition interference, etc., and should be installed by technicians with experience in this area.
d. Marine installations have the same basic interference problems as mobile installations, plus that of interference produced by specialized electrical and electronic equipment; reference should be made to handbooks or texts on this subject.
e. Interference due to tuner defects may be tracked down to component failure, or to mis alignment (see Sections 1 and 2).
f. Telephone-line crosstalk has been found in installations employing unshielded lines to the speakers with the lines installed parallel to telephone service conductors.
Whether a difficult interference problem is en countered in a domestic, marine, or an airborne installation, the basic attack is the same. In the event that sophisticated professional equipment is avail able, we can proceed by tuning in radio noise on a field-intensity meter. Sometimes, the signal can be identified by means of earphones, as shown in Fig. 7-14. An essential part of professional equipment is an electrostatically shielded loop probe; it is used as an antenna to locate sources of noise in machinery.
Moving the probe in the direction of the source ( or some conductor radiating the noise energy) causes the signal strength to increase. A block diagram for a professional interference-locating unit is shown in Fig. 7-15.
6. Weak Output
Weak output in a hi-fi system may originate at any component in the chain. Some defects impair only one function, whereas other defects can impair more than one, or all functions. Therefore, the trouble analysis starts with a check of operation on each function. If the fm tuner is defective, only fm reception may be affected; however, some defects cause impairment of both fm and AM reception. In normal operation, a typical fm tuner provides an output of 1 volt rms when driven by a 10-µV input at 75 kHz deviation. A typical AM tuner provides approximately the same output when driven by a 10-µ, V input with 30 percent amplitude modulation. This check can be made only if a lab-type generator is available. However, in case of doubt, it may be possible to make a comparison check with another tuner, using an ordinary generator.
Possible causes of weak output in a hi-fi system are:
a. Broken antenna conductor, or short-circuited lead-in.
b. Defective function switch.
c. Plug not fully inserted into connector.
d. Subnormal power-supply voltage.
e. Tuner defect or misalignment (see Sections 1 and 2).
f. Amplifier malfunction (see Section 5).
g. Defective speaker (see Section 6).
7. Microphonic Ringing or Howling
Microphonic ringing or howling ( acoustic feed back) is much more common in tube-type hi-fi systems than in transistorized designs. In most cases, a defective tube is the offender. However, other kinds of microphonic components are occasionally found both in tube-type and transistor equipment. For example, a phono cartridge sometimes causes this trouble symptom; in such a case, the ringing or howling occurs on the phono function only. The two main sources of microphonics in an AM tuner are the local-oscillator components and poor installation practices. Any loosely secured component in the local-oscillator section is a potential source of micro phonics. For example, a local-oscillator coil replacement suspended by its leads or a long thin wire hung loosely between the tuning capacitor and its associated components is a typical troublemaker.
Acoustic feedback is aggravated by poor installation practices, such as mounting the tuner in front or on top of a speaker. It is comparatively easy to localize a microphonic component by means of tapping tests.
Possible causes of microphonic ringing or howling in a hi-fi system are:
a. Cold-soldered or unsoldered contact that develops varying resistance when subjected to slight mechanical vibration.
b. Operation of system at excessive output level in a confined space.
c. Deteriorated resistor that responds like a car bon microphone.
d. Defective tube.
e. Malfunctioning phono cartridge.
f. Mechanically insecure component in local-oscillator section.
g. Poor system-installation practices.