SAFETY

For all equipment which operates from domestic mains supplies the first concern in servicing should be safety, particularly from electric shock hazard. While a risk of shock from internally generated voltage is present all the time the covers are removed, the most dangerous shock hazard is that of completing a circuit from mains live to earth. Since TV aerial systems can be expected to be earthed, and many other surfaces and equipment which surround the engineer may be earthed, it is essential that the main supply to the equipment under service is isolated. A double-wound and fully insulated isolating transformer is required, and to provide a suitably ‘stiff’ (low impedance) supply a 500 W type is recommended. Suitable fuses are a 5 A HRC type in the primary circuit, and a 2.5 A anti-surge type in the secondary circuit.

For field engineers the shock hazard on site is greater than in the workshop. Possible solutions are a large rubber mat on which to work, or the use of a portable isolating transformer: provided service is confined to reasonably modern TV sets a 250 W type suffices for diagnosis work. Test equipment for TV field work should not be earthed; it should be tested regularly for insulation integrity.

A second aspect of safety is the need for consumer equipment to maintain BS EN60065 and BEAB safety standards, as it did when it left the factory. These standards are mainly concerned with user safety, and are implemented by the OEM (Original Equipment Manufacturer); the requirement of the service engineer is that he does nothing to compromise the equipment safety. In practical terms, shock, fire and implosion risks are avoided by using OEM-sourced and approved components wherever a safety component (clearly marked by the symbol Δ or shading on the circuit diagram and/or parts list) is replaced, and by mounting it in the same way as the original. Particular care is necessary in the event of repairs to – or replacement of – back covers, cabinets, picture-tubes and user controls.

TEST EQUIPMENT

In effect the test equipment forms the eyes and ears of the engineer, who depends on its accuracy and reliability to interpret what is happening in the circuits being dealt with; but it is important to understand the limitations of the test gear and the effect of its use on circuit operation. Brief descriptions of the main classes of test instrument follows.

PRACTICAL SERVICING

In general the servicing process consists of three distinct steps: (a) diagnosis of fault; (b) repair or replacement of faulty component or assembly; and (c) setting-up, alignment and test. All three apply equally to mechanical, electronic and optical aspects of TV and video equipment. It is important to separate these three aspects of service work as far as possible. While diagnosis by substitution may be necessary where information, expertise, test equipment or (in some circumstances) time is lacking, it is far better to carry out careful analysis of the symptom and a logical process of reasoning – aided by progressive test readings – to identify the faulty component.

There are several exceptions to the rule, especially where (as is usually the case) the labour time in diagnosis must be held to the absolute minimum. Where the fault is an intermittent one, cross-substitution (e.g. of picture-tube cathode feeds, or of L and R channels in audio systems) is invaluable, especially where – as in the examples given – the fault can be made to give a positive indication of its origin when it appears. For complex faults the problem area can be considerably narrowed down by substituting complete sections or PC boards where possible. In cases where components (ICs, tuners, modules etc.) are pluggable and a substitute available, it is expeditious to test in this way. On the other hand, haphazard trial-and-error methods of faultfinding are likely to damage printed circuits and delicate components, multi-legged ICs and wound components, especially where tightly packed PC boards (many double-sided) are used in miniature and portable equipment. The time spent in this pursuit can often be more profitably used in analysing TV-screen images, voltages and waveforms.

It is good practice to spend the first few minutes (or more) of service attention in a careful and close visual examination of all sections of the equipment – a high proportion of failures are due to ‘physical’ causes like corrosion; damaged, split or cracked circuit boards; burnt, broken or misplaced components; intermittent joints or contacts; faulty plug/socket connections; and spark or corona discharge of one type or another. If no obvious problems can be seen, the next logical step is a check of supply voltages to appropriate sections of the equipment – not only in respect of correct voltage, but (particularly in the case of digital circuits) the degree of ripple and spurious noise riding on them. The presence of input signals, and correct operation and connection of peripheral components should next be checked for. Failure to do this could result, for instance, in a long diagnosis session in a videorecorder servo only to find that the control track head is dirty; the replacement of a TV tuner because the aerial has been blown down; or the replacement of a TV picture-tube whose heaters are not being energised! Only when it is proved that there are no physical problems, power lines are present and correct, input signals present and output or load devices connected need fault diagnosis begin in the signal-, power- or data-handling stages themselves, in the sequence: signal in, operating conditions, signal out. Reference to the OEM’s service literature is essential at this stage, not only to establish the layout and operating mode of the circuit and components, but to refer to the d.c. voltages and waveforms quoted for normal operation.

ICs are always direct-coupled internally, so that an incorrect pin voltage should lead to a check of the externally derived voltages on associated pins (refer to the IC block diagram) before condemnation of the chip itself. Where a clamping or gating action takes place inside an IC, incorrect ‘static’ voltage levels may well be due to a missing or badly shaped keying, trigger or gating pulse. In digital and especially processor chips, all relevant inputs should be checked in the event of an incorrect or missing output, bearing in mind that apparent loss of an input signal can be due to a stuck-low input gate within the chip itself.

Many circuits, analogue and digital, have feedback loops. While this may appear to complicate the task of fault-finding, it can often help, in that under fault conditions the two ‘loose ends’ of the loop move (electrically) in opposite directions in an attempt to restore normality. Thus the loss of an FG feedback signal will usually vastly increase a videorecorder’s capstan speed to aid diagnosis; a low gain UHF tuner in a TV will result in a ‘go-to-full-gain’ a.g.c.-line message to the tuner even with high applied r.f. input level; the chopper transistor in a switch-mode PSU will have a high duty-cycle waveform applied to its base in the event of no output due to its collector lead opening, and so on. Again, analysis of all symptoms, and rational thought, should lead to a speedy diagnosis.

Experience, both of diagnosis and fault patterns in general, and especially of the particular equipment under service, counts for a great deal. Many a problem has been very quickly solved on the basis of ‘hunches’ and a knowledge of the habits of various classes of components. Large electrolytic capacitors tend to dry up and reduce capacitance, especially when mounted in hot places; some sorts of transistors are prone to go b-e open-circuit in certain circuit configurations; heavily loaded drive belts in video decks will slip under heavy loading conditions – the list is very long. An experienced engineer is aware of all these things, and thus often able to short-cut much of the diagnostic procedure.

Some quick checks are easy to understand and follow. A 10 kΩ wire-wound resistor with 200 V across it is intact if it is very hot, and open-circuit if it is cold to the touch. A multirange meter’s prods form a handy and instant shorting link with the meter switched to its 10 A range. An analogue test meter, provided its sensitivity is known (e.g. 20 kΩ/V d.c., 1 kΩ/V a.c.), is a useful switchable substitute resistor provided the applied voltage will not overload the movement. The test voltage on an ohms range will switch on a transistor junction. These are but a tiny selection of time- and trouble-saving techniques.

In support of the approach to fault-diagnosis suggested above, the off-screen photos on the following pages have been chosen not as simple cause-and-cure cases (diagnosis by ‘stock-fault’ lists is seldom relevant to modern equipment) but as illustrations of logical testing and analysis, and of how much information can be gained by careful study of picture symptoms. The same principles apply to situations (i.e. deck malfunctions, digital logic faults etc.) where the convenience of a ‘screen’ fault display is not present, necessitating close observation or the use of test instruments at the outset.

SYMPTOM ANALYSES

Fig. 23.5 shows a snow-and-noise bar across the extreme bottom of a videorecorder playback picture from a known-good tape. The effect is due to mistracking, and because it is confined to one section of the picture it is unlikely to be due to a servo fault, or maladjustment of the tracking control. The bottom of the picture corresponds to the top half of the tape ribbon, and is read out by the head as the tape leaves the head wrap. The implication is that the tape exit angle from the drum is wrong, and it is likely that the trouble is confined to the exit guide or the adjacent section of head rabbet. The need for cleaning and position checking is more likely than a need for height adjustment unless the guide is loose or has been disturbed.

In Fig. 23.6 appears another videorecorder fault, this time in the form of a glass-like bar of picture disturbance on replay of a tape recorded in a faulty machine. In fact the bar is the visible effect of a wrongly timed head-switch point during record. Normally it is hidden at the bottom of the picture by the action of the head-drum servo which should ensure that the changeover point takes place at a point in time just before the field sync pulses. If the offending bar is stationary the first check, should be on the setting of the record switch-point preset and the monostable it controls. If the bar drifts up or down, an unlocked head servo is indicated; the correct action then is to check for correct playback servo lock on a known-good recording. If good lock is achieved the likelihood is that off-air field syncs are not reaching the head servo during record. If not, the head servo itself is suspect, and a first check would be for correct output pulses from the drum tachogenerator.

The display of Fig. 23.7 is symptomatic of white crushing, and could arise in a camera, TV set, monitor or video recorder. It can normally only take place in a video signal amplifier, and the trouble source is easily traced with an oscilloscope, progressively checking the video waveform until the non-linearity becomes evident. At this point the bias, supply voltage, input level and feedback arrangements should be checked.

It is possible for the symptom to arise from overdeviation of an f.m. modulator or signal distortion (‘bottom crushing’) in an a.m. i.f. amplifier – check bias and a.g.c. voltages.

Another video-stage fault is pictured in Fig. 23.8. The picture is heavily shaded from left to right and badly smeared. An immediate conclusion is that the luminance signal(s) is deeply modulated at line rate, and that it stems from a large ripple on some l.o.p.t.-derived power supply line. The most heavily loaded and worked is that for the 180 V or 200 V h.t. supply for the RGB output stages.

Most sets use flyback rectification to derive this supply, and since the picture is dark (high supply voltage) at left, which immediately follows flyback, and brightens (low supply voltage) as line scan progresses, the obvious conclusion is that the 180/200V -line reservoir is low capacitance due to leakage or drying up.

The display of Fig. 23.9 is a familiar one, and is due to a poor S/N ratio in the vision signal. First check (preferably with another TV) that aerial signal strength is adequate, and that the r.f. connections to the tuner are correct. If so, the next check should be of the applied a.g.c. voltage to the tuner, sometimes easiest done by applying a known correct voltage to the a.g.c. input pin. The persistence of the symptom with correct gain-control voltage incriminates the tuner, whereas restoration of a good picture shows that the r.f. a.g.c. circuit (or its control pot) is in need of attention. Any ‘low gain’ fault which gives this degree of noise must invariably be in an early stage of the signal-handling circuits; low gain in the final i.f. amplifier stage or in post-demodulator circuits gives a flat ‘milky’ picture rather than a snowy one.

The multiple images of Fig. 23.10 illustrate a case where the TV screen is better diagnostic instrument than any piece of test equipment. The problem here is due to signal reflections, and the strength and position of the ghost image with respect to the main one indicates the amplitude and delay of the reflected signal, the latter easily calculated on the basis of 1 μs equals 1/52 picture width, which can then be expressed if required in terms of signal path length.

While ghost images can arise from reflection of the UHF signal from hills, tall buildings etc., they can be generated within the equipment: this photo was taken with the ground connection of the luminance delay line open to give multiple reflections along the line.

A likely effect of much shorter-term signal reflections appears in Fig. 23.11. Here a page of text shows corruption, in which the incorrect and missing characters result from misinterpretation of data. It can be caused by a faulty text decoder, but this is unlikely where its power line, grounding and video input signal level are correct. Apart from an eyeheight check with a ‘scope on the data lines, see Chapter 8, little can be done with ordinary test instruments. If the r.f. tuning and vision demodulator ‘tank’ coil tuning are correct, here is a justifiable case for testing by substitution – starting, perhaps, with a complete text-equipped TV set, because the aerial, distribution system, and downlead or feeder are prime suspects in these cases.

Moving onto the timebase sections of a TV or monitor, a picture display like that of Fig. 23.12 again rewards careful analysis with an almost certain diagnosis. Here the screen is grossly overscanned in the vertical direction, with evidence of bad non-linearity. This indicates a large increase in the power applied to the field scanning coils. Of several possible causes, those of an increase in the intrinsic gain of the field amplifier, or a large rise in the supply rail voltage to the field timebase are very unlikely; in the latter case, other symptoms would probably be present.

An increase in generated ramp height is again unlikely, and could be quickly checked with an oscilloscope. The first check, though, should be made at the main output negative feedback loop which here is open-circuit.

The opposite effect, field collapse, is the subject of Fig. 23.13. Careful study of the resulting single scanning line, however, reveals a slight undulation of the line at the extreme left, and this is the effect of a ringing pulse induced into the field scan coils from the line-scan pair at flyback. This arises from a lack of damping in the field coil circuit, and is indicative of an open yoke circuit. The field scan coils themselves are rarely responsible, and investigation is generally confined to the coupling components between coils and output stage.

A single vertical white line, where scan and e.h.t. are derived from the same flyback transformer, is likewise indicative of a break in the immediate circuit of the scan coil, typically due to a defective or dry-jointed S-correction capacitor or scan-balance coil.

A line timebase fault is depicted in Fig. 23.14. The basic problem here is a gross timing error between line scan and video signal, causing the line blanking period to appear on the screen. Its likely causes are perhaps less obvious without some thought. It is seldom possible for this error to arise in the line sync phase detector itself in a flywheel system; much more probable is that either its reference (incoming line sync) or feedback (line flyback) pulses are suffering an abnormal time delay. A check with a dual-trace scope will probably eliminate the former, and identify a problem in the pulse-feedback circuit between l.o.p.t. and line sync chip. A change in resistor value or leakage in a capacitor is often responsible.

Faults in PSU systems are common, and while many of them result in complete loss of picture and sound, others give a display symptom which immediately identifies the fault area, and sometimes the actual component. Fig. 23.15 shows one such. The picture has two ‘wasp-waists’ which shows that the main power line bears a strong 100) Hz ripple. This can only come from a full-wave mains rectifier system, and is almost certainly due to failure of the primary reservoir capacitor in the PSU. The bulging effect will generally move slowly up or down the screen because mains a.c. is not synchronous with TV transmissions.

A single wasp-waist is an almost certain indication that one of the four diodes in the rectifier bridge is open, resulting in half-wave rectification.

The screen image shown in Fig. 23.16 shows horizontal overscan, which would most likely be due to a fault in the flyback tuning of the line output transformer (check the capacitor(s) across the line output transistor c-e junction); a fault or maladjustment in the E–W raster correction circuit (start by checking whether the width control works); or an increase in supply voltage to the line output stage. While the latter possibility should certainly be checked, it is an unlikely one because the PSU over-voltage protection would normally prevent the fault being displayed.

TEST CARD FEATURES

The final screen picture, Fig. 23.17, is an (almost!) fault-free reproduction of the test card used in the previous shots, which is representative of the general-purpose types used by broadcasters and in factories and workshops. The main features of the card shown here are as follows:

TRACING INTERMITTENT FAULTS

Of all equipment malfunctions, intermittent faults are the most difficult and frustrating to trace and diagnose. Provided that operating conditions are not ‘borderline’, i.e. weak tacho pulses, low supply-line voltage, PSU starting resistor gone high resistance etc., they usually have their origins in mechanical or ‘thermal’ effects. Bad jointing on PCBs or within components, as well as hairline board cracks etc., can usually be traced by flexing, probing and tapping of PC boards – once the fault has been made to appear, use test equipment to prove its origin for a definite diagnosis.

Intermittent faults which are temperature-dependent can often be traced to a single component (very often a semiconductor) by the alternate use of gentle heat from a hairdrier and an aerosol freezing spray. The most vulnerable and suspect components for thermal faults are diodes, transistors, ICs and to a lesser extent ceramic and electrolytic capacitors. ‘Mechanical’ intermittent problems are most often due to dry joints (sometimes concealed under an apparently good solder blob), potentiometers, switches, plug/socket interfaces and board-print faults. The most vulnerable joints are those which carry heavy (e.g. scanning and primary power) currents, and those which support heavy or iron-tagged components like transformers and large electrolytic capacitors.

The main aim in diagnosis of intermittent faults should be to get positive indications of the identity of the faulty component or section. Thus a solder-up of suspect board joints or replacement of suspect components followed by a fault-free test period is not as satisfactory as a cross-substitution test of stages, panels or individual components, in which the fault can be reproduced in isolation. Whether this takes the form of transferring an intermittent blue picture to an intermittent green one by interchanging B- and G-output transistors, or an indisputable meter reading at some crucial point in the circuit during the presence of the fault, it gives complete confidence in the diagnosis and the repair.

It is important to glean every clue from the equipment user in cases of alleged intermittent faults. Analysis of these can considerably narrow down the field of search. In difficult cases, it is necessary to leave the equipment with diagnostic instruments (ideally redundant and aged test gear saved for the purpose) hooked semi-permanently on key test points.

PCB PRACTICALITIES

The second aspect of service after fault diagnosis is repair and component replacement. Ease of repair and access varies enormously between different make, type and vintage of equipment. Unsoldering of PCB-mounted components, especially multi-legged types, is most easily done with a small hot soldering iron and commercially prepared desoldering wick, which consists of resin-impregnated copper braid to absorb old solder as it is melted. A useful alternative, especially for use with double-sided and plated-through-hole boards, is the solder pump, a spring-loaded vacuum device; the most effective is the type in which it is combined with a mains desoldering iron whose bit is made in the form of a hollow nozzle.

The very tightly packed double-sided PCBs used in cameras, camcorders and portable videorecorders are the most difficult to deal with – see Fig. 1.3. Many of their components are SMD (Surface-Mounting-Device) types, which are soldered direct to the print panel on both sides of the board. These pinhead-size transistors, resistors, links and capacitors are glued to the board surface in manufacture, then soldered to the print in a bath-wave operation. To remove them, one or more small hot soldering irons are required to simultaneously release all connections. For multi-legged SMDs like miniature ICs it is usually necessary to use a specially designed desoldering station with hot-air blower and heat shields for both de-soldering and subsequent fitting of the replacement chip. It is important to be sure of the diagnosis before changing components on these high-density assemblies, and to closely consult the manufacturer’s service information in all cases.

STATIC PRECAUTIONS

Increasingly, C-MOS and similar IC technology is being incorporated in consumer equipment, and careful handling of these devices is required to prevent damage from static electricity charges. The necessary precautions are to ground the soldering iron and (by means of an earthed wristband) the operator; to keep the replacement device in its conductive packing until the moment it is to be fitted; to temporarily ground the circuit into which it is being fitted: and if possible to keep its leadouts strapped together until the fitting/soldering operation is complete. A wide range of accessories is available to combat the risk of ESD (Electro-Static Discharge) damage. They include conductive bench- and floor-mats, heel/toe dischargers and devices to check for static charges and for the efficacy of wristband dischargers.

SERVICE MANUALS AND DOCUMENTATION

Unless one is very familiar with the equipment under service it is foolhardy to attempt diagnosis on any but the simplest equipment without a service manual, preferably that produced by the OEM. Most manufacturers and some specialist and trade magazines produce fault-finding guides and algorithms, lists of common faults, modifications and spares ordering information. All this information should be carefully filed, along with personal notes and findings, by make and model.