DG3J 35 Electronic Fault Finding

March 2008 SQA

Electronic Fault Finding

DG3J 35

Acknowledgements
SQA gratefully acknowledges the contributions made by Scotlands colleges in the authoring, editing and publishing of this material. No extract from any source held under copyright by any individual or organisation has been included in this publication.

Scottish Qualifications Authority Material developed by Robert Keddie. This publication is licensed by SQA to COLEG for use by Scotlands colleges as commissioned materials under the terms and conditions of COLEGs Intellectual Property Rights document, September 2004. No part of this publication may be reproduced without the prior written consent of COLEG and SQA.

SQA Version 1

Developed by COLEG

Electronic Fault Finding

DG3J 35

Contents
Acknowledgements Contents Introduction to the unit What this unit is about Outcomes Unit structure How to use these learning materials Symbols used in this unit Other resources required Assessment information How you will be assessed When and where you will be assessed What you have to achieve Section 1: Techniques of fault diagnosis Introduction to this section Assessment information for this section Introduction Logical fault-finding methods Sequential and non-sequential methods Block diagrams Exceptional faults Summary of this section Answers to SAQs Section 2: Implementing a fault location strategy Introduction to this section Assessment information for this section Introduction Safety Safe working practices Fault location strategy Worked example Summary of this section Answers to SAQs 2 3 5 5 5 5 5 6 8 9 9 9 9 11 13 14 15 16 17 21 33 37 39 41 43 44 45 46 48 49 50 55 57

SQA Version 1

Developed by COLEG

Electronic Fault Finding

DG3J 35

Section 3: Locating faults to component level Introduction to this section Assessment information for this section Introduction Analogue example Digital circuits Digital example Summary of this section Answers to SAQs Glossary

59 61 62 63 64 71 72 75 77 79

SQA Version 1

Developed by COLEG

Electronic Fault Finding

DG3J 35

Introduction to the unit

What this unit is about
This unit is designed to enable candidates to understand the concept of electronic fault finding and enable them to be proficient in designing and implementing a fault location strategy. This unit is particularly suited for candidates who expect to work as electronic technicians (especially in a maintenance role) but is also relevant to all those on an electronic study programme who require a practical understanding of electronic fault finding.

Outcomes
On completion of this unit, the candidate should be able to: 1. 2. 3. Explain the techniques of fault diagnosis in electronic circuits and systems. Implement a fault location strategy in an electronic system. Locate faults to component level in digital and analogue circuits.

Unit structure
This unit contains the following study sections: Section number and title 1 2 3 Techniques of fault diagnosis Implementing a fault location strategy Locating faults to component level Approx. study time 5 hours 10 hours 25 hours

How to use these learning materials

These learning materials have been written with the emphasis on fault-finding theory. However, practical experience is vital. You will spend more time performing practical fault-finding exercises than studying the theoretical principles. A good approach is to read through the notes first, then attempt to put the principles into practice, before rereading the material. That way, you should gain a clearer understanding, and be able to put the theory into context. Then you can transfer that knowledge to other systems and become a more effective engineer. Your tutor will also need to supply you with information about the systems and circuits on which you are to perform fault finding. This information cannot be included here as it depends on the equipment that is available at the centre. Understanding the manufacturers documentation is a key part of the fault-finding process.

SQA Version 1

Developed by COLEG

Electronic Fault Finding

DG3J 35

Symbols used in this unit

These learning materials allow you to work on your own with tutor support. As you work through the course, you will encounter a series of symbols which indicate that something follows that you are expected to do. You will notice that as you work through the study sections you will be asked to undertake a series of self assessed questions, activities and tutor assignments. An explanation of the symbols used to identify these is given below. Self assessed question 1 This symbol is used to indicate a self assessed question (SAQ). Most commonly, SAQs are used to check your understanding of the material that has already been covered in the sections. This type of assessment is self contained; everything is provided within the section to enable you to check your understanding of the materials. The process is simple: you are set SAQs throughout the study section you respond to these by writing either in the space provided in the assessment itself or in your notebook on completion of the SAQ you turn to the back of the section to compare the model SAQ answers to your own if you are not satisfied after checking your responses, turn to the appropriate part of the study section and go over the topic again.

Remember the answers to SAQs are contained within the study materials. You are not expected to guess at these answers. Activity 1 This symbol indicates an activity, which is normally a task you will be asked to do that should improve or consolidate your understanding of the subject in general or a particular feature of it. The suggested responses to activities are given at the end of each section. Remember that the SAQs and activities contained within your package are intended to allow you to check your understanding and monitor your own progress throughout the course. It goes without saying that the answers to these should only be checked after the SAQ or activity has been completed. If you refer to these answers before completing the SAQs or activities, you cannot expect to get maximum benefit from your course.

SQA Version 1

Developed by COLEG

Electronic Fault Finding

DG3J 35

Tutor assignment formative assessment 1 This symbol means that a tutor assignment follows. These are found at the end of each study section. The aim of the tutor assignment is to cover and/or incorporate the main topics of the section and prepare you for unit (summative) outcome assessment.

SQA Version 1

Developed by COLEG

Electronic Fault Finding

DG3J 35

Other resources required

A selection of analogue and digital systems is necessary to give practical experience. The unit specification states that various electronic systems may be used for fault finding. A commercial/industrial electronic system would be particularly appropriate but it is recognised that this will not always be possible. Other suitable systems could include: a programmable logic controller operating external equipment a desk-top computer driving a multiapplications board a microprocessor driving a multiapplications board an alarm system a multistage AF or RF amplifier a static inverter.

For practical work in Section 2 it would be best to have a complex system which can be subdivided into clearly separate stages. For Section 3, it is necessary to have access to an individual PCB or unit so that fault finding to component level can be performed. In order for the candidate to practice fault finding, several systems need to be available to test. A wide variety of systems would be helpful. Clearly, each system should have a fault on it somewhere, preferably of a type which does not make the source of the fault obvious. This might mean shorting a component out with a solder bridge, deliberately damaging a capacitor, cutting a track where it will not be noticed, or replacing a resistor with one of a much higher value. Alternatively, a set of switches could be used to implement shorts or open circuits. They could be wired to the track side of the PCB, then the wiring covered up to prevent the candidate seeing what faults can be switched in. This would not be appropriate if it makes it unreasonably difficult to access the board with test equipment or if surfacemounted components are used so that they are on the same side as the PCB tracks. Section 3 contains a couple of examples of simple circuits. It may be practical to build multiple versions of each design, with different faults on them. A simulation package could also be used to simulate a working circuit, then various shorts or open circuits implemented to confirm the effect of these faults. Simulations are not recommended for the practical assessments.

SQA Version 1

Developed by COLEG

Electronic Fault Finding

DG3J 35

Assessment information
How you will be assessed
Assessment for this unit should take the following form: Section 1 short written test lasting 30 minutes. Section 2 practical exercise on an electronic system Section 3 one practical exercise on an analogue system, one practical exercise on a digital system.

The written test for Section 1 may be composed of a balance of short answer, restricted-response and structured questions. Sections 2 and 3 are practical and it is recommended that candidates be assessed by the use of checklists and a report written by the candidate which includes a fault-finding log. Assessment for both the written test and the practical exercises will be conducted under controlled, supervised conditions.

When and where you will be assessed

Assessments will normally be at the end of each section. For Sections 2 and 3 the centre will inform you of the deadlines for the submission of reports.

What you have to achieve

Each section lists the knowledge and skills covered. You should check these lists before attempting the assessments. Evidence for the knowledge in Section 1 may be provided on a sample basis. This means that only a selection of the material will be covered in the assessment, but your responses should satisfactorily demonstrate your competence. Centres are recommended to use appropriate checklists to monitor the candidates fault-finding activities during the assessment of Sections 2 and 3. Written reports will need to contain relevant information. A list of what is required will be issued with each assessment and is included at the start of Sections 2 and 3 in these notes. Opportunities for reassessment Normally, you will be given one attempt to pass an assessment with one reassessment opportunity. Your centre will also have a policy covering exceptional circumstances, for example if you have been ill for an extended period of time. Each case will be considered on an individual basis and is at your centres discretion (usually via written application), and they will decide whether or not to allow a third attempt. Please contact your tutor for details regarding how to apply.

SQA Version 1

Developed by COLEG

Electronic Fault Finding

DG3J 35

SQA Version 1

10

Developed by COLEG

Electronic Fault Finding

DG3J 35

Section 1: Techniques of fault diagnosis

SQA Version 1

11

Developed by COLEG

Electronic Fault Finding

DG3J 35

SQA Version 1

12

Developed by COLEG

Electronic Fault Finding

DG3J 35

Introduction to this section

What this section is about This section covers a variety of techniques for fault diagnosis in electronic circuits and systems. These are presented in general terms so that they can be applied to all electronic devices.

Outcomes, aims and objectives At the end of this section you will be able to understand: sequential and non-sequential fault location methods systematic fault location methods, e.g. input to output, output to input, half-split fault location methods in complex systems, e.g. divergence, convergence, alternative path exceptional faults, e.g. manufacturing faults, multiple faults, catastrophic failure.

Approximate study time 5 hours.

Other resources required None.

SQA Version 1

13

Developed by COLEG

Electronic Fault Finding

DG3J 35

Assessment information for this section

How you will be assessed Evidence for the knowledge in this section may be provided on a sample basis. The evidence may be presented in responses to specific questions. Each candidate will need to demonstrate that she/he can answer correctly questions based on a sample of the items shown above under Outcomes, aims and objectives. In any assessment of this section, one item from each of the four parts should be sampled. Candidates must provide a satisfactory response to all four items.

When and where you will be assessed Evidence should be generated through an assessment paper lasting 30 minutes undertaken in supervised conditions. Candidates may not bring to the assessment any notes, textbooks, handouts or other material.

What you have to achieve At the end of this section you will be able to: explain the difference between sequential and non-sequential fault location methods explain the use of input to output and output to input methods explain the use of half-split methods explain the significance of diverging paths in fault finding explain the significance of converging paths in fault finding explain the significance of alternative paths in fault finding describe typical manufacturing faults explain the added difficulties involved when trying to locate multiple faults explain the significance of catastrophic failure.

Opportunities for reassessment In order to ensure that candidates will not be able to foresee what items they will be questioned on, a different sample from each of the four knowledge items is required each time the section is assessed.

SQA Version 1

14

Developed by COLEG

Electronic Fault Finding

DG3J 35

Introduction
Fault finding is an important activity for most engineers. Whether it is a matter of testing and repairing items as they come off the production line, or maintaining equipment on the production line itself, or debugging a new design as it is developed, almost every engineer will have some experience of fault finding. It is vital that fault finding is done in an effective and efficient manner. It has been estimated that repairing faulty goods costs around six times as much if done after the product has left the factory. Making an inadequate repair will not only cost more than fixing the item properly, it will also harm the manufacturers reputation. If equipment on the production line fails in use, then it can incur huge costs for the company. Consider the wages of workers who are unable to do productive work while machinery sits idle the company management will want the fault fixed as quickly as possible. However, they will not want a poor repair which will fail later, possibly causing additional damage. Neither will they want to risk injury to their employees. In todays marketplace companies need to innovate to survive. They need to get new products on sale faster than those of their competitors. When new designs are under development, it is often delays fixing problems which cause deadlines to be missed. This is especially likely when new technology is involved unexpected problems can arise, causing the launch of the product to be postponed. It is generally accepted that fault finding needs to be done in a logical manner. Familiarity with a particular piece of equipment can mean that problems can be fixed without resorting to a methodical approach, but this level of expertise only comes after years of experience. Also, it lacks transferability a good engineer should be able to fault find on unfamiliar equipment, and become competent quickly. This unit has been written to suit readers with a wide range of circumstances. As a result, it cannot go into great detail on particular systems. Instead, a variety of examples are quoted. Some of these involve equipment such as cars and domestic appliances, with which most people are familiar. Others are electronic systems, of which you should have technical understanding. Later sections involve practical exercises, but to begin with we shall cover the general principles of fault finding.

SQA Version 1

15

Developed by COLEG

Electronic Fault Finding

DG3J 35

Logical fault-finding methods

In everyday life you will occasionally encounter problems with equipment. Familiarity with its use means that you will often be able to fix the problem immediately without giving it much thought. For example, from time to time you may need to use a head-cleaning disc on a CD player. When listening to music you may find the CD jumping or repeating sections. Using the head cleaner makes the problem go away, and you conclude that the issue has been resolved. Dampness can penetrate the electrics of car engines. The symptoms are that the car can be hard to start, or may suffer intermittent losses of power until the engine bay heats up and the dampness has been expelled. A temporary solution is to spray the ignition leads with water-repellent spray. In the long term it may be necessary to replace the high-tension leads. In both of these cases we are using knowledge of the system to attempt a repair. However, these approaches rely more on statistics than logic. If you are sure that a CD is not jumping because the disc is scratched (easily checked), then dirt on the laser lens is the next most likely cause of the problem. If you own a car for a few years you become familiar with its faults, and it is easy to assume that a problem is due to the same cause as before. A moments thought will show that there are a number of faults which could cause similar symptoms. Erratic starting and running could be the result of fuel contamination, a blocked fuel filter, blocked jets, faulty sensors, or a problem with the engine management system. Some modern cars deliberately limit the engine power if they detect a fault this is intended to let the owner drive to a garage without causing further damage. It is not logical to assume that fault symptoms are the result of a particular failure, just because this was the case on previous occasions. It may be statistically likely, but by no means guaranteed. There is the risk of premature diagnosis, where one jumps to conclusions about what the source of the problem is. You need to check that your hypothesis is correct, and take care when interpreting symptoms it is easy to delude oneself that they support your fault-finding diagnosis. If you are lucky, there may exist a fault-finding procedure which can guide you. Instructions for domestic appliances usually contain these, but are of limited help to anyone with engineering competence. For example, the manual for a TV satellite receiver suggests that if you cannot find a channel you have previously watched, it may be because the channel only broadcasts for part of the day. The manual for the computer on which these notes were typed suggests that if the monitor is blank, the user should check that the cable is fully connected.
SQA Version 1 16 Developed by COLEG

Electronic Fault Finding

DG3J 35

Manuals for industrial equipment are usually much more helpful. They typically contain block diagrams, schematics, a brief explanation of how the system works, a list of suitable test points, and the voltages and signals which should be found at them. Complex equipment can have such a wide variety of possible failures that it may be impossible to list all possible faults and the procedures for correcting them. This is where you need to apply a logical approach. Finally, remember than the system may not actually be faulty. It is possible that the user is operating the equipment incorrectly, or perhaps has encountered unusual behaviour and assumed it is the result of a fault. For example, if an oscilloscope input is switched into ac-coupling mode, then a steady dc voltage appears as 0 V. If one were not paying attention, it might seem that either the CRO was broken, or the signal being measured was faulty. Also, if unreliable equipment is prone to giving incorrect signals, one is liable to ignore them even when they are correct. There is the case of an airliner which crashed after the pilots ignored a warning message from safety equipment which frequently triggered false warnings. Perhaps the moral of this case is that unreliable equipment is worse than equipment which is broken.

Sequential and non-sequential methods

Fault finding can be described as a six-step process, namely: collect evidence analyse evidence locate fault determine and remove cause rectify fault check system.

This may seem little more than stating the obvious in formal language, but there are important points to be made here.

SQA Version 1

17

Developed by COLEG

Electronic Fault Finding

DG3J 35

In many cases evidence of the fault will already be available the equipment will be handed in for repair with a description of the problem. However, the user may not have reported the symptoms accurately, or might have included extra information which confuses the issue. It is to be expected that users will see the problem from their perspective rather than that of the person attempting a repair. For example, a complaint might be the photocopies of my report are unreadable which is of obvious concern to the author. A better description would be "copies of text on coloured paper have the background printed too dark". This is more helpful in terms of identifying the fault. If possible, it is best to see the fault for yourself. This may be impractical in the case of a unit removed from a larger system. You may need to have specialist equipment to simulate inputs, or a variable power supply and signal generator may be enough. For example, you could test an audio amplifier for distortion by connecting a pure 1 kHz signal. Alternatively, other audio equipment such as a CD player or a video recorder might be suitable. In debugging it is said that a good test is one with a high probability of finding an error, and a satisfactory test is one which uncovered a previously unknown problem. This applies to fault finding as well. We can increase our efficiency by carefully choosing what to test. Milking the front panel is a term given to a technique for collecting evidence. Use switches and selectors on the equipment to determine which functions are working and which are not. This may help to determine which part of the equipment is at fault. This is the second step in the process analysing the evidence. A block diagram is useful at this stage. If the system can be split into separate parts with their own functions, the fault symptoms may indicate that one unit has failed. This block may be suspected as the culprit because all system functions using it are faulty, while functions not requiring that unit are operational. This is not a guaranteed method a faulty selector switch might give the same symptoms. To get confirmation, we need to locate the fault more accurately. This may mean using test gear to check signals and power going into the unit, and output signals. If the inputs are correct but the outputs are wrong, the unit is clearly inoperative. One issue we need to be aware of is that the original cause of the fault may be entirely separate from the final result. If the output from a voltage regulator is shorted to the input, then components running off that voltage supply may fail. The result might be that all the logic devices on a board overheat. The operator may then switch off the equipment after only one or two components have burnt out. If these components are then replaced without fixing the original short, the fault will then recur.

SQA Version 1

18

Developed by COLEG

Electronic Fault Finding

DG3J 35

We could say that the key issue here is that we need to be sure we are fixing the fault, not just fixing the symptoms. It is important to check the system after the supposed repair to ensure that it really is fixed. If there is an underlying problem which is still to be addressed, then the equipment will fail a second time. It is better that this happens while the equipment is in the workshop, than later when it has been returned to the customer with an invoice. One point often forgotten when considering the six-step method is that it is often an iterative process parts of it have to be repeated as necessary. Fixing a fault may only deal with part of the problem part of the equipment may still not be fully functional. This can happen not only with power supply faults, but also in many other circumstances. The user may continue to use equipment when it is operating below standard, then ask for a repair once more serious faults have arisen. You may discover a fault which has gone unnoticed. A common mode failure is one where apparently separate systems fail simultaneously in the same way because of an underlying problem. Four-engined aircraft are believed to be particularly safe because the chances of more than one breaking down on a flight are minute. In fact, there are a variety of reasons this could happen, such as fuel contamination, bird strikes, software problems and so on. Critical computer systems, such as those on an aircraft, usually have multiple parallel processors, sometimes sourced from separate manufacturers, in an attempt to ensure they cannot all fail simultaneously. In practice, this is not guaranteed either. For a start, such a system requires arbitration if the computers are giving different outputs, a choice has to be made over which signals to ignore. This arbitration system could itself fail. Also, the engineers writing the software could all make the same faulty assumption, leading to all the computers generating an incorrect signal. Finally, there is the possibility that the specification is wrong. Formal methods of fault finding can be classified as either sequential or non-sequential. Non-sequential methods are those used by automated test equipment (ATE). ATE performs a large number of tests on the system under investigation and uses a database of possible fault symptoms to determine the problem. As far as the operator is concerned, the tests are all done simultaneously. Also, the sequence of tests may always be the same, independent of any errors detected. In this sense, ATE is said to be using a non-sequential method. Sequential testing is where tests are done in a sequence, one after the other. The result of each determines what test is done next.

SQA Version 1

19

Developed by COLEG

Electronic Fault Finding

DG3J 35

Manual testing done to a proper strategy is systematic testing. Tests are performed according to a plan. The strategy can be based on reliability data, where the system under test has been in use long enough to give helpful information on where faults are likely to occur. Whether this is possible depends on individual circumstances. A factory building thousands of mobile telephones may quickly acquire such statistics. A small workshop repairing a variety of devices may find that every fault is different, and has to use a different approach. It is only worthwhile to perform extensive research into fault modes if either large numbers of the unit are produced, without design changes, or if testing is hazardous or expensive. This section is largely concerned with fault finding based on block diagrams. Three kinds of systematic test will be covered, as shown in Figure 1.1. For completeness, the diagram also includes random testing, which is non-systematic. This method is rarely used it is difficult to conceive of a circumstance in which it would be appropriate. Perhaps if the symptoms gave no indication of where the fault was located, and no block diagram for the system was available, this might be worthwhile. The larger the system, the more efficient it is to have a logical approach.

Fault Finding Methods Sequential Systematic Based on reliability statistics Based on block diagram (input to output, output to input, half-split method) Non-systematic (random testing) Non-sequential (used by ATE)

Figure 1.1 Classification of fault-finding methods.

SQA Version 1

20

Developed by COLEG

Electronic Fault Finding

DG3J 35

Block diagrams
Block diagrams are a useful way of describing how a system is constructed. They present an intermediate level between the instruction manual or specification, and the schematic diagram. The instructions state what functions the equipment performs and how to operate it, while the schematics show the components and the electrical connections between them. These are both necessary for fault finding, but for a system of any size a block diagram is an aid to comprehension. Figures 1.2 and 1.3 show two typical block diagrams. Large systems may need two or more layers of block diagram, breaking it down into units then subdividing them. It might also be helpful to use a block diagram to show how the equipment is connected to other devices, as in Figure 1.2.
DVD player

Aerial

Television

HDD recorder

Satellite dish

Satellite receiver

Figure 1.2 Block diagram for connections to domestic television.

ASCII Data

Interface

Character Store RAM

Character Generator ROM

Shift Register

Serial Video out

Crystal

512

8 Horizontal Sync

32

Vertical Sync

Figure 1.3 Block diagram for monitor circuit.

SQA Version 1

21

Developed by COLEG

Electronic Fault Finding

DG3J 35

Block diagrams can often appear complex. One would like to think that they have been designed to make them easy to follow sadly this is not always the case. If you are drawing a block diagram yourself, try to position the blocks so that the connections between them are as simple as possible. Try to ensure that signal paths cross no more than necessary apart from that, signals flow from left to right by convention. (Power rails are usually shown separately on schematic diagrams to reduce the complexity.) In some cases it is possible to have a clear relationship between the units in the block diagram, the schematic, and the circuitry. Analogue circuits are often easier to lay out than digital ones in this respect. An integrated circuit with four gates on it may have three of these used in widely different parts of the circuit, making it impossible to lay these out separately on the printed circuit board (PCB). Also, space constraints may make it impossible to design so that units shown in the block diagram are separate on the PCB.
A B C D E F G

Figure 1.4 Diagram for system with linear connections between blocks. The easiest block diagrams to follow are linear signals flow from one unit to the next, without splitting or joining. Figure 1.4 is an example. We shall consider three systematic methods of fault finding on such a system: input to output output to input the half-split method.

Input to output This is perhaps the most obvious approach. With a suitable input connected, the first block in the chain is tested. If the system is connected as in Figure 1.4, the output from unit A would be compared with the correct signal. If As output shows no error, it is assumed to be working correctly. Unit B would be tested next, followed by C, D and so on. The process is repeated until a faulty signal is found. The broken unit is then the most recently checked one. For example, if testing proceeds with all signals correct until the output of E is faulty, then E is the source of the problem. Input-to-output testing has the advantages of being simple and direct. Once the faulty unit has been repaired or replaced, the system can then be checked. If it is still not functioning correctly, there may be a second fault. In the example, one would check that the output of unit E is now correct, then proceed through F onwards.

SQA Version 1

22

Developed by COLEG

Electronic Fault Finding

DG3J 35

Remember that it is possible that the fault lies in the wiring all the blocks could be working normally. To check for this problem, test the signal at either end of the connecting cables.

Output to input This method operates in the reverse direction. If a system is producing a faulty output, one could start by testing the input to the final stage, G in Figure 1.4. If G has a correct input but faulty output then that is the source of the problem. Otherwise, we would check the input to F. The process would be repeated until a unit with a correct input signal is found; it should be the source of the problem. Of course, there is always the chance that more than one block is damaged. With output-to-input testing, additional faults may turn up in blocks which have already been checked. It could be argued that this makes it an inferior method to the previous approach. Fault finding by looking for the first signal which is not faulty might also cause confusion and lead to more mistakes.

The half-split method This is also known as the binary split method, and is a more sophisticated strategy. The first signal to be tested is in the middle of the system. If it is correct, the second half must be faulty. If the signal is wrong, the first half of the system has an error. The next step is to narrow down the search by looking at a signal in the middle of the faulty half, and repeating until the damaged unit has been identified. In Figure 1.4, if unit E is faulty we would expect to track it down as follows: This example has an odd number of blocks so there is no suitable point half-way through. Suppose we check the output from C first. It will be correct. Next, check half-way between C and the final output; this is the signal connecting E to F. It will be corrupted. Finally, check the input to E; this is in the middle between the last two test points. It should be correct, indicating that E is faulty.

If the last test had given a faulty signal then D would be identified as the source of the problem. Figure 1.5 illustrates this process. Note how each test halves the size of the area containing the fault.

SQA Version 1

23

Developed by COLEG

Electronic Fault Finding

DG3J 35

suspect area A B C test here A B C D E test here A B C D test here E F G F G D E F G faulty output

Figure 1.5 Identifying a fault with the half-split method. Now, this may seem an overly complex method compared to the other two. Its advantage is that it is theoretically the most efficient strategy, assuming the fault could be in any block with equal probability, and all the signals are equally easy to test. In the example, three tests are enough to identify the faulty unit, whichever one it is. To see how this is efficient, consider a much more complex system, with 50 units. Three tests would divide the system into halves, quarters and eighths, which would be enough if there are no more than eight blocks. To check a system of 50 blocks needs six tests five tests could only check 32 blocks (25 = 32) while six tests are needed for a system of 3364 blocks (26 = 64). In contrast, using input-to-output or output-to-input testing would require an average of 25 tests, assuming there is a single fault at a random location. A sensible block diagram would not have 50 blocks in it; this would be unreasonably complex. However, it is possible to have a system with 50 similar units, such as a network of computers or a delay line with multiple filters connected in series. In a practical situation, we would want to modify the half-split method to take account of a number of realities. Firstly, we usually cannot assume that the fault is equally likely to be in any block. Some blocks will be more complex than others, or less reliable, and in any case the fault symptoms will allow us to narrow down the search, even if only by eliminating some units as the source of the problem. Secondly, not all tests are equally expensive, in the sense of the time and effort taken to perform them. It may be easy to access test points at the output of some circuits, while others may require much dismantling. Testing may involve complicated
SQA Version 1 24 Developed by COLEG

Electronic Fault Finding

DG3J 35

equipment such as a logic analyser, which will take a while to set up and give results that are time-consuming to analyse. One should also consider possible hazards involved in testing. You could concentrate on low-voltage systems first and avoid live circuits where possible. There are other risks to take into account, such as possible damage to the equipment under test, either through dismantling and reassembling it, or as a result of testing it, or running it while a fault is present. In short, a more sophisticated approach to the half-split method would be to weight the testing to take account of the likely fault location and the cost or risk of testing. Aim to minimise the expected effort required to find the fault. Instead of picking a point halfway through the faulty section, try to find a node with a 50% chance of showing a faulty signal, with a bias towards tests that are easy to do.

1.1 Referring to the system in Figure 1.4, list the test points you would use with the halfsplit method, if the faulty unit is: 1. 2. 3. F B G

SQA Version 1

25

Developed by COLEG

Electronic Fault Finding

DG3J 35

Non-linear structures Block diagrams are rarely as simple as Figure 1.4. It is usually impossible to lay out the structure in a straight line, as some units have more than one input or output. If one part of the equipment has multiple outputs, feeding several other blocks, then the signal flow is said to diverge. A unit with multiple inputs is said to be convergent. It takes signals from more than one source. Complex structures with converging and diverging paths can make fault finding more complex, but they can actually help track down the problem. For example, consider the system in Figure 1.6. This is a divergent system, centred around block B, which has three outputs. There are three final outputs from the system, arising from D, E and H. If the output from D is faulty, while E and H are correct, then we can narrow down the search immediately before even testing internal signals. Blocks C and D are obvious candidates.
E A B C F D G H

Figure 1.6 System with diverging signal paths. A lot depends on exactly how Bs outputs are connected Figure 1.6 shows them as three different signals, but it could be one signal split three ways. For example, consider Figure 1.7. In (a), a signal generator is shown as a single block, with three separate outputs. It is conceivable that one of these is damaged, while the other two are functioning normally.
open circuit
E

Signal generator

square wave sine wave TTL

+12V regulator +5V regulator

Motor driver

C F

RF data link

(a) Block with three outputs

(b) Block with one output to three destinations

(c) Alternative block diagram

Figure 1.7 Alternative way of representing diverging signal paths.

SQA Version 1

26

Developed by COLEG

Electronic Fault Finding

DG3J 35

However, the system in Figure1.7(b) has a single output to three blocks. An open circuit could cause one circuit to fail while the others function normally. It might be better to draw the block diagram as in Figure 1.7(c) to make this explicit. B, the +12 V regulator, must be working if at least one of the devices running off it is functioning normally. Returning to the example of Figure 1.6, if D and E give correct outputs but H does not, then the signal paths from B to F and G to H need to be checked. Use one of the earlier strategies, such as input-to-output testing. There is a further possibility. All three outputs may be erroneous. In this case, we would suspect that the fault lies before the point at which the signal paths diverge. In the example, we would suspect units A and B. The basic principle is to check any unit where all the components which its output passes to give a faulty output in turn. Of course, it may be that there are multiple faults, but this will become clear as testing and repair proceed. Converging paths also need to be taken into account. If a unit needs multiple inputs to work, then if its output is correct all preceding units can be assumed to be functioning. Figure 1.8 shows an example. A good place to start when testing this system would be the output of block R. If this is correct, then only S and T need to be investigated. If R shows an incorrect output, its three inputs would be tested next. If these are all correct, then R is the source of the problem. Otherwise, it is likely that the search can be narrowed down to only two blocks.
J L N K M P R S T

Figure 1.8 System with converging paths.

SQA Version 1

27

Developed by COLEG

Electronic Fault Finding

DG3J 35

Alternative structures Figure 1.9 shows a more complex system. It has both diverging and converging paths. Note how blocks D and E both have two outputs, so they count as divergent. Also, there are three parallel paths between C and J. It is possible to suggest a logical strategy for fault finding such a system dependant on details of the fault symptoms. For example, if the external outputs from D, E and K are all erroneous the fault is likely to be in A, B and C. A good place to start would be to check the output from C first, then work backwards. This is applying the output-to-input method to sections AC. If D is producing a correct output but E is not, then E would be the likely candidate. K would probably generate an error also. If D and E are correct but K is not, then this is now similar to the converging system problem covered earlier. J may be faulty or the problem may lie with the two paths F/G and H. E may also be damaged, depending on whether its two outputs are separate.

D A B C F H

E G J K

Figure 1.9 System with diverging and converging paths. Figure 1.10 shows a system with alternative paths. It is a computerised control system, with four computers running in parallel so that the system will still work if one fails. (This is known as a quadruplex redundant system.) One of the computers is sourced from a different manufacturer to reduce the chance of common mode failure. A side-effect of such a system is that if one unit fails it needs to give a prominent warning so it can be fixed. If this did not happen, the operator would continue using the system until it failed completely. This would probably be disastrous, as redundant systems are only worth the extra expense and complexity for safety critical situations.

SQA Version 1

28

Developed by COLEG

Electronic Fault Finding

DG3J 35

A domestic example is the ring mains, where power to the sockets comes from a loop of cable wired to the distribution at both ends. This will still work if there is a single break in the wires. In electronics, it is sometimes necessary to wire several components in parallel to achieve the necessary properties. For example, several power transistors would be connected in parallel if the current being handled exceeds the limit for a single transistor. Power supplies tend to need a large smoothing capacitance, often achieved by connecting two or more capacitors in parallel. In both of these cases, if one unit fails the system will keep working, but may overheat and be damaged. When fault finding a system with alternative paths, it may be necessary to test each of the subsystems connected in parallel. There are also maintenance issues, with equipment being tested on a regular basis to ensure no failure has occurred undetected.
Computer A1 Computer A2 Computer A3 Computer B

Pilot Controls

Arbitration

Control Surfaces

Figure 1.10 System with alternative paths.

SQA Version 1

29

Developed by COLEG

Electronic Fault Finding

DG3J 35

1.2 In the three systems below, the shaded block is faulty. List which units will give a correct output and which will not. 1.
A B C D E F G H

2.
S

T V X

U W Y Z

3.
A B C D E F

G H J

Feedback A further type of system is one with feedback paths. These can be tricky to test as faults effectively circulate round the system, meaning that most of the signals are incorrect, even for units working properly. The feedback may be used to improve the performance of the system, perhaps to compensate for changes in the environment. In this case, it may be possible to disconnect the feedback, possibly replacing it with a substitute signal. However, if the system relies on feedback to work properly it may not be possible to break the loop. Closed loop systems require more thorough knowledge of how the device is supposed to work. It is often necessary to understand how each part of the system will work if given an incorrect input. Each unit is liable to have a faulty input and output you need to identify a device where the output is inconsistent with its input, or at least eliminate those which are working consistently.

SQA Version 1

30

Developed by COLEG

Electronic Fault Finding

DG3J 35

For example, consider the temperature controller in a piece of industrial plant. Figure 1.11 shows a block diagram. The fault symptoms are that heat is applied continuously, leading to an over-temperature condition, triggering an alarm and an emergency cutoff.
0-12V + 0-12V Signal processing 0-0.5V Thermocouple 12V Controller 4-20mA 0-100% 0-40kW Actuator Heater Plant

Set point

0-400C

Figure 1.11 Temperature control system with feedback. The way this equipment is supposed to work is that the operator sets the desired temperature on a dial, while the actual temperature is sensed by a thermocouple. A controller works a heating element, applying energy proportional to the difference between the required and actual temperature. The block diagram in Figure 1.11 shows the signal types and their ranges. For example, the unit shown as set point has an output in the form of a dc voltage. It has a dial calibrated as 0400C, and will be linear so that if the operator turns it to 200C, the output is at half of the range, or 6 V. Now, if the temperature is seen to rise continuously, past 200300C and beyond, we would expect most of the signals to be different from normal. For example, suppose we find an output of 12 mA from the controller. This causes the heater to run at half power, even though the temperature is now above the set point. However, checking the controller input shows it is receiving a voltage of 6 V. It is likely that the controller is working normally; it is operating the heater at half power because its input is wrong. Figure 1.12 shows some measurements of signals in the system. The thermostat is the likely cause of the problem as its output is 0 V even though the temperature is 300 C and rising. As a rough guess the output should be three-quarters of the maximum, 0.375 V in this case. The manual should allow a more accurate calculation.

SQA Version 1

31

Developed by COLEG

Electronic Fault Finding

DG3J 35

Set point

6V

+ -

6V 0V Controller

12mA Actuator

50% Heater

20kW Plant

300C

Signal processing

0.0V

Thermocouple

Figure 1.12 Temperature control system with faulty signals. Other blocks seem to be working normally, with outputs consistent with their inputs. One way to operate this equipment without feedback would be to connect a variable power supply in its place. Set this to 0 V to simulate the system being cold the heater should come on full. Turning the voltage up to 12 V should represent over-temperature and cause the heater to switch off. Slowly turning up the voltage should simulate the temperature rising to the set point, and should cause the heater to throttle back. The temperature control example is assumed to use proportional control the amount of heat applied is proportional to the difference between desired and actual temperature. Control systems which incorporate integral action are harder to fault find. Connecting a DC input instead of the feedback will not help as any error signal will cause the controller output to ramp up or down. Other systems may be much harder to test with feedback disconnected. A control system for a missile, for example, has to handle signals which change in a small fraction of a second. In such circumstances, one would hope that the manufacturer has devised a suitable test procedure and documented it in the manual.

SQA Version 1

32

Developed by COLEG

Electronic Fault Finding

DG3J 35

Exceptional faults
Other types of faults we need to consider are multiple faults, catastrophic failure and manufacturing faults.

Multiple faults So far we have tended to assume that there is only one fault in the system under test. This may not be the case. The obvious approach when dealing with multiple faults is to repeat the process, fixing one problem at a time until they are all repaired. The methods we have covered are appropriate for dealing with multiple failures, although with diverging paths the symptoms can be misleading. If two units which take inputs from a common source both fail, we would tend to suspect the wrong block. If multiple faults are present, it is unlikely to be a coincidence. One failure may have triggered further damage. A good place to start is to check the power supply. If that has failed, then none of the units requiring power will work. If the power rails have gone over voltage, for example, allowing unregulated voltage to reach the rest of the system, then there is likely to be further damage. Fix this first, then set about repairing the consequential damage. If several blocks in the system are faulty, it would be helpful if they could all be tested individually. This can be difficult if suitable signals need to be present for a unit to be tested properly. In these circumstances, it can be worthwhile to build a circuit to generate suitable test signals to permit a unit to be tested on its own. There is also the possibility of board swapping, transferring a circuit board from a working system to a faulty one to see if this cures the fault. This technique needs to be used with care. It could endanger working boards, and increases the amount of dismantling involved. Also, it is not appropriate to use units from equipment that is to be sold as new.

SQA Version 1

33

Developed by COLEG

Electronic Fault Finding

DG3J 35

Catastrophic failure A catastrophic failure is defined as one which is sudden and complete. There is no intermediate stage where the equipment operates with reduced performance, and no warning signals that it is about to fail. A gradual failure would allow us to make repairs before the equipment fails completely, and would prevent failure in service, with possibly expensive consequences. One advantage of catastrophic failure is that it is better than an intermittent fault. Most of us have experience of a recurring problem which disappears when the equipment is taken for repair. It can be difficult to track down a fault when there are no symptoms present. The normal approach is to identify a likely candidate for the source, then stress it in the hope of triggering the fault. For example, a loose connector may work most of the time. Wiggling it while the equipment is running may make the problem reappear. There exist sprays which can be used to cool down components, which sometimes helps identify those which have worked loose in their sockets. An alternative is to heat them up with a hair dryer. Flexing cables can help track down loose wires.

Manufacturing faults These should be identified and corrected by the manufacturer before the unit leaves the factory. Failing that, they should be covered by the warranty. However, there is always a chance that a fault may not lead to a problem until after many years of service. Also, the manufacturer still has the problem of fault finding and repairing the equipment at the factory.

SQA Version 1

34

Developed by COLEG

Electronic Fault Finding

DG3J 35

Table 1.1 Typical manufacturing faults. Fault Loose cables, integrated circuits and connectors Comment Connectors are often designed with locking levers to prevent them coming loose. Also, ICs are usually soldered direct to the PCB instead of using sockets. Heating and cooling can cause them to work loose over time otherwise. Automatic soldering should eliminate this. However, it is sometimes necessary to solder manually, especially during a repair, and poor solder joints can work to begin with, only to fail later. Loose drops of solder and strands of wire can float around inside an enclosure, occasionally causing a short. Bad soldering can also short IC pins, or cause open circuits. Damage to a PCB can cause cracking. Gaps can appear in very narrow tracks during etching, especially at sharp corners. If the mechanical design of the equipment is poor it can mean that joins are under continuous stress. A circuit board may be seated at a slight angle in its slot, or wires may be not quite long enough so that they bend at a sharp angle right next to the connector. The slightest knock to the unit may cause something to break, or it may eventually fail through metal fatigue. These will overheat. If this is bad enough for them to smoke or catch fire the problem will be fixed quickly. However, a less dramatic failure will be a gradual deterioration in performance over time. The component may show discolouration or swelling or leaking. Diodes, transistors and integrated circuits all need to be inserted the correct way round. IC sockets have the correct polarity marked on them, but there is always the possibility that the socket has been soldered in the wrong way. Applying a reverse voltage to an electrolytic capacitor eliminates the dielectric, turning it into a short. CMOS devices are easily damaged by static electricity. Soldering in components can also cause heat damage. If the leads on devices such as resistors and capacitors are bent too close to the component, the casing can crack, allowing dampness to enter and degrade the performance. IC pins may be bent or broken during insertion.

Faulty soldering

Shorts and open circuits on circuit boards

Stressed joins and wires

Under-rated components

Incorrect polarity

Damage to components

SQA Version 1

35

Developed by COLEG

Electronic Fault Finding

DG3J 35

Other possible faults are if the wrong value of component is used, or inserted in the wrong place. If automated machinery is used, it is conceivable that a whole batch of boards could be manufactured. Such faults should be picked up on the production line. A batch of faulty components may have been supplied. This has happened with batteries and capacitors in recent years. It can prove extremely costly to the manufacturer if large numbers of the product enter the market place then fail later.

1.1 If you can, give details of three component failures you have encountered yourself. The author could list a tantalum capacitor which was soldered in with the wrong polarity and caught fire, soldered battery connectors which broke free when dismantling the unit, and a power supply with a wire connected to the wrong end of a resistor so that it overheated.

SQA Version 1

36

Developed by COLEG

Electronic Fault Finding

DG3J 35

Summary of this section

We have covered a number of basic logical approaches to fault finding. Block diagrams have been used to represent various system configurations, and we saw how to approach fault finding with more complex layouts, such as those featuring divergence and convergence. We also considered possible manufacturing faults and considered some examples in general terms. The next section covers practical fault finding to block level.

SQA Version 1

37

Developed by COLEG

Electronic Fault Finding

DG3J 35

1.1 1. Define: a. b. 2. sequential fault location methods systematic fault location methods.

If the receiver unit fails in the system below, describe how you would find this fault using: a. b. c. the input-to-output method the output-to-input method the half-split method.

front panel

diode matrix

microcontroller

power amp

IR LED

receiver

microcontroller

motor driver

3.

Define diverging and converging systems with reference to the system below. State the consequences of failure in: a. b. c. d. Unit A Unit D Unit G Units B and E simultaneously.
A B D F C G E

SQA Version 1

38

Developed by COLEG

Electronic Fault Finding

DG3J 35

Answers to SAQs
SAQ 1.1 1. An acceptable answer is: test the output of C (correct) test the output of E (correct) test the output of F (incorrect) conclude that F is faulty.

Since the number of blocks is not a whole power of two, other alternatives are just as efficient. For example, an alternative would be: 2. test the output of D (correct) test the output of F (incorrect) test the output of E (correct) conclude that F is faulty.

An acceptable answer is: test the output of D (incorrect) test the output of B (incorrect) test the output of A (correct) conclude that B is faulty.

3.

An acceptable answer is: test the output of C (correct) test the output of E (correct) test the output of F (correct) conclude that G is faulty.

SQA Version 1

39

Developed by COLEG

Electronic Fault Finding

DG3J 35

SAQ 1.2 1. Correct output A, B, C, D Incorrect output E, F, G, H 2. Correct output S, T, U, V, X Incorrect output W, Y, Z 3. Correct output A, B, C Incorrect output D, E, F, G, H, J

SQA Version 1

40

Developed by COLEG

Electronic Fault Finding

DG3J 35

Section 2: Implementing a fault location strategy

SQA Version 1

41

Developed by COLEG

Electronic Fault Finding

DG3J 35

SQA Version 1

42

Developed by COLEG

Electronic Fault Finding

DG3J 35

Introduction to this section

What this section is about This section covers the process of implementing a fault location strategy in an electronic system.

Outcomes, aims and objectives At the end of this section you will be able to: identify risks and use safe working practices identify fault symptoms in terms of system operation interpret fault symptoms using test equipment and/or diagnostic aids locate faulty circuits using system documentation and test equipment.

Approximate study time 10 hours.

Other resources required A selection of analogue and digital systems is necessary to give practical experience. The unit specification states that various electronic systems may be used for fault finding. A commercial/industrial electronic system would be particularly appropriate but it is recognised that this will not always be possible. Other suitable systems could include: a programmable logic controller operating external equipment a desk-top computer driving a multiapplications board a microprocessor driving a multiapplications board an alarm system a multistage AF or RF amplifier a static inverter.

It would be best to have a complex system that can be subdivided into clearly separate stages. In order for the candidate to practice fault finding, several systems need to be available to test. A wide variety of systems would be helpful.

SQA Version 1

43

Developed by COLEG

Electronic Fault Finding

DG3J 35

Assessment information for this section

How you will be assessed Candidates will submit evidence to satisfy this section in the form of a report which documents a practical fault-finding activity.

When and where you will be assessed The assessment will normally take place at the end of this section. Centres are recommended to develop and use appropriate checklists to monitor the candidates fault-finding activities and provide a check on the authenticity of the report.

What you have to achieve The report should include: the identification of risks appropriate safe working practices used a description of the fault location strategy implemented details of tests carried out and the test equipment and/or diagnostic aids used sketches of appropriate block diagrams reference to the documentation used a contemporaneous log of the fault-finding activities.

Opportunities for reassessment If the report is incomplete or erroneous, it is likely that you will be asked to correct and resubmit it. It may be necessary to perform further tests in the laboratory, adding the results to the report as required. In extreme cases you may be asked to fault find on a different system and submit a new report.

SQA Version 1

44

Developed by COLEG

Electronic Fault Finding

DG3J 35

Introduction
This section is of a more practical nature than Section 1. You need to demonstrate the ability to implement a fault-finding strategy. An outline of the process will be presented here, and later you will need to perform an assessed exercise. Obviously, experience with repairing equipment will be very helpful. You can learn more by practical experience than by reading theory in a book, but you need to be aware of the basic principles. We have an additional problem in that these notes cannot suit your circumstances exactly. We could cover a particular piece of equipment in detail but it is likely that you will meet different equipment instead. We will cover the general principles, with some examples for illustration. This will be more useful in the long term as you can then apply these principles to a wide variety of fault-finding problems. Remember that if you only needed to be familiar with one type of system, there would be no need to study fault-finding principles as in theory your employer could supply a repair strategy which covered all likely problems. If you already have some fault-finding experience, these notes should help you to put your skills in context. They will allow you to extend your skills to other applications and increase your employability. At this level it is not necessary to have an in-depth knowledge of how the system works. Familiarity with the block diagram is enough. Also, research has shown that as technicians become better at fault finding they tend to forget most of the theory they learned. Finally, the practical exercises do not require you to fix the fault, only to identify it. Here the emphasis is on choosing a good strategy for finding the fault. In real life, you would fix the problem, and then check that the repair has been effective. There might be more faults still to be dealt with.

SQA Version 1

45

Developed by COLEG

Electronic Fault Finding

DG3J 35

Safety
Before you start any practical fault finding, we need to consider risks and safe working practices. We can classify safety as: primary safety, which protects the user from injury, and secondary safety, which protects the equipment from damage.

Cynics suggest that there is also tertiary safety, which protects the manufacturer from litigation. This explains the labels which read No user-serviceable parts inside and warn the owner not to open equipment. If you do dismantle the apparatus and then injure yourself, you cannot sue the manufacturer for damages. In our circumstances, we should be qualified personnel, competent to investigate faults and trusted to use safe working practices. Be aware, however, that if you do not follow the correct procedures and are then injured, your employer may not support you. This may mean loss of pay while off work, but you may also lose your job or be prosecuted for endangering others. Primary safety The obvious risk when working on electrical equipment is one of electrocution. However, electronic devices usually run off low voltages, such as 5 or 12 V. In some cases this is supplied by an external transformer. If there is an internal power supply, it may be a sealed unit. Provided the power supply is working correctly, there is no need to have access to mains electricity. In any case, the power supply may be designed to be impossible to dismantle, so that it has to be replaced if faulty. However, bear in mind that low voltages can still be harmful, particularly if large currents are drawn. For example, batteries can be dangerous if shorted. NiCads have a low internal resistance, so shorting them can cause wires to melt. Lead acid batteries generate hydrogen when sourcing a large current. This can explode, showering the surrounding area with acid. Lithium polymer batteries are damaged by being driven into deep discharge, and can also explode or catch fire if mistreated. Cathode ray tubes as found in televisions, monitors and oscilloscopes require a highvoltage supply, thousands of volts, to operate. This voltage can be stored in capacitors and still be lethal after the supply has been switched off. Be aware that some designs of television have a chassis which floats at high voltage. There are also mechanical risks to be considered, such as moving parts and sharp edges. Take care to avoid motors, gears and levers. Video recorders in particular have complex mechanisms for feeding tape to the heads. Be aware that watches, jewellery, clothing and long hair can be trapped. Also take care when removing PCBs and connectors. These are often stiff and require force to take out. If a component comes undone suddenly it would be easy to injure oneself on sharp edges.

SQA Version 1

46

Developed by COLEG

Electronic Fault Finding

DG3J 35

There can also be chemical risks, although these are not so common with electrical equipment. Avoid chemicals leaking from damaged components, as they are often toxic. Your employer should have procedures for safely disposing of hazardous materials.

Secondary safety Testing circuits can endanger them. Clearly, it will make your job harder if you accidentally introduce new faults while attempting to repair old ones. Also, we are trying to complete repairs quickly and efficiently, not create unnecessary work. Much equipment, especially consumer goods, is designed to be reliable and cheap to build, with little thought given to making it easy to dismantle. One method is to solder ICs direct to the PCB instead of using sockets. They cannot work loose with time, but can then be very difficult to remove. Parts and connecting cables can be jammed in tightly removing one unit can risk damage to the cables. Plugs are often clamped onto the wire directly without solder. Pulling the connector out by the wires risks damage. It is easy to accidentally short two tracks together on a circuit board or multiway connector when attaching a meter probe. Shorting out the supply, even briefly, can destroy sensitive semiconductor devices. Static is also a hazard. Humans can easily be charged to thousands of volts. If you touch an IC, this voltage is discharged in a tiny fraction of a second. It causes you no more than discomfort, but can destroy the device. Use a static discharge wristband if handling static-sensitive devices (they are usually labelled as such in their packaging). In some cases it is enough to touch the chassis of the equipment while it is earthed. If you need to inject a test signal, it will normally be necessary to disconnect the existing signal to avoid a conflict. However, some circuits can be damaged if they do not have a load. Audio amplifiers are an example they should not be operated with the loudspeakers disconnected. Basically, a current generator with an open circuit output has the same risk of damage as a shorted voltage source. With no load, all the power is dissipated internally, by current flowing through the equivalent output resistance. Unused current sources should be shorted, but check this very carefully.

SQA Version 1

47

Developed by COLEG

Electronic Fault Finding

DG3J 35

Also, some equipment does not operate without a load. Devices with feedback were mentioned at the end of the last section. Switching mode power supplies are another example. One method sometimes used when testing them is to substitute a light bulb for the load.

Safe working practices

First of all, you should be familiar with the correct use of test equipment as covered in a separate unit, Electronic Testing Skills. Check the manufacturers documentation for safety advice and other information which may help identify hazards. Switch the power off while removing and replacing the cover screws could easily fall inside and cause a short circuit. Also, it may be necessary to have the power on while performing tests but there is no point endangering yourself when unscrewing the lid. It may be possible to use an isolation transformer to reduce the risk of electrocution. If you have one hand inside a unit with exposed mains terminals, make sure the other hand is nowhere near earthed metal. Do not use a static discharge wristband in these circumstances. Watch out for moving parts and keep fingers, hair and test probes out of the way. Computers can still be partially live while not running. If you want to use a static discharge wristband clipped to the chassis, leave the computer plugged into the mains but turned off with the switch on the PSU so it is isolated from the live and neutral but still earthed. Just remember to turn it back on again, as most users never use this switch. In a workshop shared with other people, take care turning on and off equipment in case you accidentally turn on the mains supply to a unit someone else is working on.

SQA Version 1

48

Developed by COLEG

Electronic Fault Finding

DG3J 35

Fault location strategy

The strategy we will be using is based on the objectives for this section. The plan is to: identify risks and use safe working practices identify fault symptoms in terms of system operation interpret fault symptoms using test equipment and/or diagnostic aids locate faulty circuits using system documentation and test equipment.

When you attempt the practical assessment, this all needs to be documented. It is a good idea to do this in a real-life application also. The advantages of documentation include: it will save time if you encounter a similar problem later if the repair turns out to be unsuccessful, you will need to refer to this record colleagues can benefit from your expertise you may forget what you have done, especially if the system is complex or numerous tests are required your employer may need this information for billing purposes lastly, in the event of an accident or a complaint from the customer, you will need a record of what procedures were followed.

What you need to do is: Identify potential risks. These could include high voltages, hazardous chemicals, damage to static-sensitive devices and moving parts. If the equipment needs to be tested while live, identify which parts are at mains voltages. List safe working practices used. Equipment should be disconnected from the mains where possible. Otherwise, consider some way of ensuring you cannot accidentally touch live contacts. There should be a procedure for safe disposal of hazardous chemicals. Describe the fault location strategy implemented. This may be one of the strategies covered in Section 1, such as input to output, or may have to take into account diverging or converging paths. Detail the tests carried out and the test equipment and/or diagnostic aids used. This may be vital for later reference when you measure voltages and signals, write down the results so you can refer to them later. You will forget otherwise. Sketch appropriate block diagrams. The units documentation may already contain block diagrams, but make up your own based on what is important to you while testing.

SQA Version 1

49

Developed by COLEG

Electronic Fault Finding

DG3J 35

Refer to the documentation used. Much time can be wasted by searching through manuals to find information whose location you have forgotten. Log fault-finding activities. Do this as you go along, rather than waiting to the end, by which time you will have forgotten important details.

This information is all required as evidence that the assessment has been performed properly, but it is also for your own benefit, and that of any of your colleagues and employer in the future.

Worked example
Suppose we have a motor controller used in an industrial plant. It uses a microcontroller to operate a conveyor belt motor at a fixed speed, set by a dial. If necessary, a modified program can be downloaded from a computer, through a USB link. A tachogenerator connected mechanically to the motor measures the actual speed, and a software algorithm uses that information to adjust the speed. There is also a separate relay for switching the motor on and off, and a safety cut-out in case the motor overheats. This is independent of the microcontroller, as it is good engineering practice to avoid relying on software for safety-critical tasks. One of our first tasks will be to acquire a block diagram for the system. It is possible that units from different suppliers were used, or various parts were hand built specifically for this application. For these and other reasons, a suitable block diagram might not exist. In any case, it is often best to sketch your own diagram to help make it clear in your mind how the system works. That way you can include information you need for fault finding, and miss out extraneous detail. (You will be required to submit diagrams as part of the assessment.) Figure 2.1 shows a block diagram for this motor controller. It may seem rather complicated, but it helps to see a system of realistic complexity too simple a system makes the task trivial.

SQA Version 1

50

Developed by COLEG

Electronic Fault Finding

DG3J 35

MAINS plug-top transformer power supply +12V -12V 5V speed control USB microcontroller board

MAINS transformer/ rectifier

relay

PC

power transistor electric motor temp sensor warning light

amplifier

tachogenerator

Figure 2.1 Motor speed controller. The unit has failed in operation. The symptoms are that the motor does not run at all, and adjusting the speed control has no effect. A primary concern is that of safety. There are two obvious hazards here. Firstly, this apparatus is mains powered, so there is the possibility of electrocution. Secondly, there are moving parts, so we could be injured by the motor or anything connected to it, such as the conveyor belt, gearbox and transmission. The block diagram shows separate supplies for the controller and motor. It should be possible to test the system with the supply to the motor disconnected. It pays to be thorough here in case someone else accidentally turns the power back on while you are working. Take out the plug, tape over the switch, add a big sign with DO NOT TURN ON in big writing whatever it takes to ensure your own safety. If the motor is disabled, then we can perform our tests without running the conveyor belt and affecting the rest of the plant. The motor will only need to be operational if we find that everything else works. The controller runs off a mains-powered transformer, but this is a sealed unit with +5 V and 12 V outputs, so this should not be a hazard. Microcontrollers are static-sensitive devices, so a static discharge wrist band might be necessary if we have to remove any component from the microcontroller board, otherwise it will not be required.
SQA Version 1 51 Developed by COLEG

Electronic Fault Finding

DG3J 35

We already have a description of the fault, and in this case there is not much else we can record. Check that the overheat warning light is not on you could also test the bulb in case that has failed, but then we would have two faults, including the source of the overheating. With a more complex system there will be other indicators to check, and we could check which functions work correctly and which ones do not. Otherwise, the next stage is to start making some measurements. We could apply the input to output method described in Section 1. This clearly is not a linear system, but it can be treated as: the low-voltage power supply the speed control, microcontroller and power transistor the tachogenerator and its amplifier the over-temperature warning system the high-voltage side with the transformer and motor.

Each of these can be dealt with separately. Note that this system also has feedback; we need to bear this in mind but hopefully it will not cause us too much difficulty. It is often a good idea to start off by testing the power supply. It is something which can be done quickly and easily, plus power supply faults can cause consequential damage. You want to eliminate this possibility immediately. If the power supply is providing +5 V and 12 V outputs correctly, the next stage would be to test the microcontroller side. Record what you do as you proceed, including any measurements. A log might read: Tested power supply: +5.01 V, +11.96 V, 12.01 V. Tested speed control: 0.01 V at minimum, 4.97 V at maximum, 2.5 V in the middle. Tested microcontroller output with speed set to half (Figure 2.2).

SQA Version 1

52

Developed by COLEG

Electronic Fault Finding

DG3J 35

5.7ms 5V 0V 1.05ms

Figure 2.2 Microcontroller output at half speed. Tested microcontroller output to relay steady 5 V.

Now the microcontroller is clearly trying to run the motor. This is an example of pulse width modulation, where motor speed is controlled by turning the power on and off very quickly (100 Hz10 kHz is a typical speed.) If the power is on half the time on average, the motor runs at half speed. The next task is to check the power transistor. This presents a problem as it cannot be easily tested without the high-voltage supply, and we are trying to avoid turning that on. Therefore, test other parts of the system first, and only return to this if everything else checks out. Next we check the tachogenerator. This should produce an output voltage proportional to the speed of rotation. With the motor stationary this will be 0 V. Check that the outputs of the tachogenerator and its amplifier are both zero a fault here could cause the system to conclude that the motor is already running too quickly, and switch off the power. The over-temperature system would be tested next. It has two outputs: one which turns on the warning lamp and one which triggers the relay to cut off the power. Alternatively, the relay may be wired in such a way that a broken wire from the temperature sensor causes it to drop out check the documentation for details. Lastly, if none of these tests shows suspect results, we would have to reconnect the high-voltage side and try running the motor. Proceed carefully, avoiding any exposed terminals and moving parts. Take care when connecting the voltmeter probes. The fault-finding log might continue: Tested transformer/rectifier output 75 V dc. Tested voltage at relay input and output 75 V dc. Tested connection between power transistor and motor with CRO steady 0 V.

We seem to have identified the problem the power transistor is faulty.

SQA Version 1

53

Developed by COLEG

Electronic Fault Finding

DG3J 35

2.1 If the fault symptoms are that the motor ran at full speed while the speed control setting has little effect, how would you set about testing the system? If the source of the problem is that the tachogenerator output has gone open circuit, what signals would you expect to see?

2.1 The unit specification recommends that you are given repeated opportunities to fault find, locating faults to block level in a wide range of electrical circuits and equipment. The centre will supply you with details of fault finding activities at this point. No detailed information can be here as it depends on the equipment which is available at your centre. Various electronic systems may be used for fault finding. A commercial/industrial electronic system would be particularly appropriate but it is recognised that this will not always be possible. Other suitable systems could include: a programmable logic controller operating external equipment a desk-top computer driving a multiapplications board a microprocessor driving a multiapplications board an alarm system a multistage AF or RF amplifier a static inverter.

While working on the practical exercises, it will be helpful to refer to Section 1 again. Theory is always easier to understand once you have some experience to relate it to.

SQA Version 1

54

Developed by COLEG

Electronic Fault Finding

DG3J 35

Summary of this section

So far we have only narrowed down the location of the fault to one unit. This may be one piece of apparatus in a large system, or a single PCB in a piece of equipment, or one section of the PCB. The next stage is to fault find to component level. This is covered in Section 3.

SQA Version 1

55

Developed by COLEG

Electronic Fault Finding

DG3J 35

2.1 Fault find one system and submit a report. This need not be as thorough as the report for the assessment but will need to include: a description of the fault location strategy implemented details of tests carried out and the test equipment and/or diagnostic aids used sketches of appropriate block diagrams a log of the fault-finding activities.

SQA Version 1

56

Developed by COLEG

Electronic Fault Finding

DG3J 35

Answers to SAQs
SAQ 2.1 As before, we would attempt to test the system with the high-voltage side disconnected. Check the outputs from the speed control, microcontroller board, tachogenerator and amplifier. There is little point in testing the over-temperature warning system. If the tachogenerator output has gone open circuit that would explain the symptoms the feedback indicates that the motor is stationary, so the controller is applying maximum power. We would expect to see a nearly 100% duty cycle at the output from the microcontroller on a CRO. The speed control would be working normally, but the tachogenerator output would be 0 V instead of a voltage corresponding to the motor speed. The amplifier would have 0 V at the output, which is incorrect, but this is because the input is faulty.

SQA Version 1

57

Developed by COLEG

Electronic Fault Finding

DG3J 35

SQA Version 1

58

Developed by COLEG

Electronic Fault Finding

DG3J 35

Section 3: Locating faults to component level

SQA Version 1

59

Developed by COLEG

Electronic Fault Finding

DG3J 35

SQA Version 1

60

Developed by COLEG

Electronic Fault Finding

DG3J 35

Introduction to this section

What this section is about This section covers the process of locating faults to component level in digital and analogue circuits.

Outcomes, aims and objectives At the end of this section you will be able to: identify risks and use safe working practices identify fault symptoms in terms of system operation select a suitable fault location method locate a fault to component level on an analogue system locate a fault to component level on a digital system use appropriate test equipment correctly use a circuit diagram.

Approximate study time 25 hours.

Other resources required The systems used in Section 2 could be tested here. It is now necessary to have access to an individual PCB or unit so that fault finding to component level can be performed. In order for the candidate to practice fault finding, several systems need to be available to test. A wide variety of systems would be helpful. This section contains a couple of examples of simple circuits. It may be practical to build multiple versions of each design, with different faults on them. A simulation package could also be used to simulate a working circuit, then implement various shorts or open circuits to confirm the effect of these faults. Simulations are not recommended for the practical assessments.

SQA Version 1

61

Developed by COLEG

Electronic Fault Finding

DG3J 35

Assessment information for this section

How you will be assessed Candidates will submit evidence to satisfy this section in the form of two reports, one for fault finding on an analogue circuit and one on a digital circuit.

When and where you will be assessed The assessment will normally take place at the end of this section. Centres are recommended to develop and use appropriate checklists to monitor the candidates fault-finding activities and provide a check on the authenticity of the report.

What you have to achieve The reports should include: the identification of risks appropriate safe working practices used a description of the fault location strategy implemented details of tests carried out and the test equipment used circuit diagrams used reference to the documentation used a contemporaneous log of the fault-finding activities.

Opportunities for reassessment If one of the reports is incomplete or erroneous, it is likely that you will be asked to correct and resubmit it. It may be necessary to perform further tests in the laboratory, adding the results to the report as required. In extreme cases you may be asked to fault find on a different system and submit a new report.

SQA Version 1

62

Developed by COLEG

Electronic Fault Finding

DG3J 35

Introduction
In the previous section we used fault-finding techniques to identify the faulty unit. This is known as fault finding to board level if it stops at identifying the damaged PCB. Sometimes the faulty board will be discarded and replaced with a working one. It may be passed on to someone else to repair if this is considered financially worthwhile. If it is covered by the warranty, it will be returned to the manufacturer. However, you may need to perform the repair yourself if no replacement unit is available. Also, you may be asked to fix the fault in a unit someone else has identified as malfunctioning. In this section, we will continue on downwards, fault finding to component level. A key factor in fault finding at this level is the ability to understand how the circuit is supposed to work. Exact details are not necessary, for example the designer will have calculated resistor and capacitor values, and picked suitable components. (They may not be critical anyway, or the designer may have simply adapted someone elses design.) With experience, you can tell from a schematic diagram which components are likely to have a bearing on the fault, and identify roughly what they are for. Hopefully, the documentation will include an outline explanation of the circuit. As before, we will tend to concentrate on general principles. Your centre will need to have some circuits for you to fault find on, but details of these cannot be given here. Your tutor will have more information. We will begin with two simple circuits, one analogue and one digital, to show basic principles. Later you will have some more complex circuits to fault find on, and finally the assessment. This will also feature an analogue and a digital circuit.

SQA Version 1

63

Developed by COLEG

Electronic Fault Finding

DG3J 35

Analogue example
+9V R1 100k C1 + 10F input R2 22k Rc 3.9k Q1 BC184L b c 10F e + Ce 47F gnd output +

C2

Re 1k

Figure 3.1 Common emitter amplifier. Figure 3.1 shows a single transistor common emitter amplifier. It should be familiar. Basically, the transistor needs to have a dc bias to work. The base needs to be about 0.7 V above the emitter, and the dc level of the collector is half-way between the supply and the emitter voltage. This maximises the signal amplitude that can be output without clipping. Capacitors C1 and C2 are the coupling capacitors. They permit an ac signal to enter and leave the circuit, while removing the dc bias. Figure 3.2 shows the signals which should be present at various points in the circuit when there is a sinusoidal input.
V 1.7V 0.9V time V
b

V 5V

Ve

output voltage time

(a) Signals at base and emitter

(b) Signals at collector and output

Figure 3.2 Typical signals at key nodes in circuit.

SQA Version 1

64

Developed by COLEG

Electronic Fault Finding

DG3J 35

Resistor Re is present to permit thermal runaway. The value is usually chosen to make Ve 10% of the supply voltage. As the transistor heats up, it conducts better and the emitter voltage rises. This causes the emitter current to reduce, acting as a negative feedback effect and stabilising the circuit. Without Re the transistor could become too hot and be damaged. Ce effectively disables the feedback for ac signals, as otherwise the gain would be much reduced. Now, if we are testing this circuit, there are only three points or nodes in the circuit which can be usefully measured, namely the base, emitter and collector of the transistor. The input, output and power rails are external connections. We assume that these are intact, although it is a good idea to check. There is a chance that a working unit has been handed in for repair, when actually one of the connecting cables was at fault. Also, if the power rails are shorted somewhere on the board, you would want to find out quickly before further damage is done. (This can be easily checked by measuring the resistance between the power rails with no power or signal supply connected.) Table 3.1 lists the voltages at the transistor terminals, with no input signal connected. These are the dc levels. The values shown as expected are calculated roughly based on our understanding of the circuit. The measured values are taken from an actual circuit built in the laboratory. Table 3.1 DC voltages for common emitter amplifier. Terminal Emitter Base Collector Measured (V) 0.91 1.56 5.43 Expected (V) 0.9 1.6 5

If we test the circuit and find substantially different values then we have evidence of a problem. The actual values give an indication of the source of the fault. We should be able to work backwards and work out what effect a possible fault might have. If the measurements agree with this we have probably found the fault. Be aware that some faults can give similar symptoms. All dc faults can be classified as: faults on the PCB shorts and open circuits resistors faults usually open circuit (shorts are rare) transistor faults short or open circuit on collectorbase or emitterbase junction.

SQA Version 1

65

Developed by COLEG

Electronic Fault Finding

DG3J 35

Capacitor faults are considered separately, as a capacitor appears as an open circuit to dc anyway. A dc fault will cause the transistor to be either fully off or saturated (fully on). If the transistor is fully off, the emitter will be at 0 V and the collector at the supply voltage. If the transistor is saturated, the emitter and collector will be at approximately the same voltage. This can be estimated by treating Re and Rc as a voltage divider, giving about 1.8 V. The base cannot normally be more than about 0.7 V above the emitter; the transistor will be damaged otherwise.

Resistor faults We will consider open circuits only. If R1 is open circuit, R2 will pull the base down to zero and turn the transistor off. Hence the emitter will be at zero, and the collector at the supply voltage. If R2 is open circuit, R1 will pull the base up and saturate the circuit. Hence the emitter and collector will be at 1.8 V, and the base at 0.7 V above this. If Rc is open circuit, no collector current can flow, and the base and emitter currents will be the same. The emitter current will then be much less than normal, approximately zero. If Re is open circuit, the transistor cannot conduct so the collector will be up at the supply voltage. The base will sit at whatever voltage the R1/R2 bias resistors give.

Table 3.2 shows actual measured voltages from a real circuit in the lab. Note that connecting a voltmeter to components which are open circuit can have an effect. This explains why the collector is not at 0 V when Rc is open circuit. Similarly, connecting the meter across an open circuit emitter resistor effectively replaces it with a large resistance.

SQA Version 1

66

Developed by COLEG

Electronic Fault Finding

DG3J 35

Table 3.2 Voltage measurements with resistor faults. Fault R1 open circuit R2 open circuit Rc open circuit Re open circuit Emitter voltage (V) 0.0 1.9 0.0 1.2 Base voltage (V) 0.0 2.5 0.7 1.6 Collector voltage (V) 9.0 1.9 0.1 9.0

3.1 Confirm that Re = Rc = 1.8 V if the transistor is saturated. Calculate also the base voltage if the base current is zero. This should agree with the symptoms with Re open circuit.

3.1 Calculate the emitter, base, and collector voltages if a fault on the PCB shorts out each of the resistors.

3.2 If time permits and you have access to a simulation program such as Electronics Workbench, try simulating the circuit of Figure 3.1. Confirm the symptoms of each of the resistor faults listed in Table 3.1 and check your calculations for SAQ 3.1.

SQA Version 1

67

Developed by COLEG

Electronic Fault Finding

DG3J 35

Transistor faults The transistor can fail by internal damage. The baseemitter or basecollector junctions can appear as open circuits or short circuits. Similar symptoms can appear if there is a short or open circuit where the transistor is soldered to the PCB. Note that there is no emittercollector junction in a transistor, although the terminals could still be shorted externally. The effects these various faults have on the amplifier circuit can be worked out the same way as with resistor faults, so these faults can be traced by checking the dc voltages with no signal connected. If the basecollector junction is open circuit, then the transistor cannot conduct. There will be virtually no emitter current, so the emitter will be at 0 V with the base sitting at 0.7 V. If the basecollector voltage is shorted, then the base will be pulled up, saturating the transistor. If the baseemitter junction is open circuited, the transistor cannot conduct, but this time the emitter will sit at 0 V when measured with the meter. If the baseemitter junction is shorted, they will be at the same voltage, forcing the transistor off. The collector will be pulled up to the supply voltage. If there is a short between the emitter and collector, they will both be at 1.8 V, with Re and Rc acting as a potential divider. The base will then sit at 1.6 V, not loading the bias resistors.

Table 3.3 lists some actual measurements taken by disabling a working circuit. Note that a short between the base and emitter gives similar symptoms to R1 being open circuit. If these voltages are measured and there are no obvious signs of a short or component damage, then remove the transistor. Check that the R1/R2 potential divider is giving the correct base voltage of 1.6 V; if it does, the transistor was at fault.

SQA Version 1

68

Developed by COLEG

Electronic Fault Finding

DG3J 35

Table 3.3 Symptoms of transistor faults. Fault Emitter voltage (V) 0.0 1.7 0.0 0.1 1.8 Base voltage (V) 0.7 2.3 1.6 0.1 1.6 Collector voltage (V) 9.0 2.3 9.0 9 1.8

Basecollector junction open circuit Basecollector junction short circuit Baseemitter junction open circuit Baseemitter junction short circuit Short circuit between emitter and collector

Capacitor faults Open circuits in the capacitors do not affect the dc operation of the amplifier, so the base, collector and emitter voltages will all appear correct with no signal connected. A test signal is required to fault find in this case. C1 and C2 are coupling capacitors, effectively allowing ac signals to pass while eliminating dc offset. If either capacitor goes open circuit, then the ac component of the signal will disappear at some point. Use an oscilloscope to test for this problem. If C1 is open circuit, there will be no ac signal at the base of the transistor or anywhere else in the circuit. If C2 is faulty, there will be an amplified signal at the collector, but none at the output. Figure 3.3 shows these symptoms.
V 1.6V
V

V 5V

time

output = 0V time

(a) C1 open circuit

(b) C2 open circuit

Figure 3.3 Symptoms of open circuit coupling capacitors.

SQA Version 1

69

Developed by COLEG

Electronic Fault Finding

DG3J 35

As described previously, the emitter capacitor means that Re can eliminate thermal runaway without affecting the ac gain. If the gain is much reduced, it indicates that Ce is open circuit. The gain of this circuit is controlled by the gain of the transistor, quoted as or hfe in the manufacturers data sheets. This is of the order of 100 for general purpose transistors such as the BC184L. If Ce goes open circuit, the gain will drop to around 5. Capacitors are more likely to be shorted than to go open circuit. Their operation requires the presence of an extremely thin insulating layer, the dielectric, between the two conducting plates. In the case of an electrolytic capacitor, this takes the form of a coating of aluminium oxide. Internal damage can lead to a gap in the dielectric, allowing current to pass. For a capacitor to go open circuit, one of the leads would need to break.

3.2 What effect will a short in one of the capacitors have? How would you test for such a fault?

SQA Version 1

70

Developed by COLEG

Electronic Fault Finding

DG3J 35

Digital circuits
Some things need to be taken into consideration when testing digital circuits. A few are described here. Open circuit inputs tend to float high on digital gates. That is to say, they act as if they are at logic 1, not logic 0 as one might expect. This is not always the case, however. Shorting a digital output to ground or the supply is likely to damage it, as one of the output transistors will effectively short out the power supply. AND and OR gates have a peculiar property when it comes to fault finding. Under certain circumstances, a fault at one input does not affect the output. For example, if one input to an AND gate is at zero, the output will be zero, even if the other input is wrong. With an OR gate, if one input is high the output is high, regardless of any fault in the other input or inputs. NAND and NOR gates have similar properties. You can use this effect to your advantage when testing a combinational logic circuit. It means you can work out what the output of a gate should be without checking all the inputs. Fault finding a digital circuit means identifying a device whose output is inconsistent with its inputs. A logic probe can be used for testing, although it is quite common to use an oscilloscope since most engineers will have one near at hand. One method would be to check each signal in turn as the circuit runs, watching out for signals which never change. These are likely to be faulty. It is easier to check logic levels while they are static. A circuit which runs off a clock, with signals which are continuously changing, is much harder to analyse. One cannot check an AND or OR gate with a two-channel oscilloscope as the output and inputs all need to be displayed simultaneously. A digital storage scope is necessary if we need to examine a signal which does not repeat continuously, or repeats too slowly to allow a steady trace on a normal CRO.

SQA Version 1

71

Developed by COLEG

Electronic Fault Finding

DG3J 35

A logic analyser would be useful for a digital circuit, as it allows multiple signals to be sampled and displayed as parallel traces. They tend to be more complex to set up, and are often used only if simpler methods do not allow the fault to be found.

Digital example
+5V Green +5V TP1 +5V
14

+5V Amber +5V Red

IC1 74LS93 IC3a

12 9 8 11 1 2 1 2 4 5 3 5 A

R IC2b Q

IC3b 270
6 3 4

TP2 Clock TP3 Gnd

A
1

IC2a

270 IC3c
6

B
2 3

Q Q

B C D

270

R01 R02

Figure 3.4 Sample digital circuit. The circuit in Figure 3.4 simulates a set of traffic lights. It is quite a simple circuit (it could be even simpler), but it is a good example to begin with. A 3-bit binary counter cycles through eight states. The sequence should run as shown in Table 3.4. In order to represent realistic timing, the lights spend more time at red and green than changing through amber. Table 3.4 Output sequence. State QD 0 0 0 0 1 1 1 1 QC 0 0 1 1 0 0 1 1 QB 0 1 0 1 0 1 0 1 Decimal 0 1 2 3 4 5 6 7 Traffic light displayed Red Red Red Red + amber Green Green Green Amber

SQA Version 1

72

Developed by COLEG

Electronic Fault Finding

DG3J 35

The LEDs are wired to the digital logic so that a logic 0 turns the LED on. The combinational logic used is: red comes on when QD is low, so the red LED is connected directly to QD. amber comes on when QB and QC are both high, so the logic is QB QC .

green green LED should be on whenever red and amber are both off, so it is connected to Re d Amber .

The integrated circuit used is a 74LS93. This is a 1-bit counter plus a 3-bit counter. They can be connected to form a 4-bit counter if required. We are just using the 3-bit part, so there are some spare pins. Inputs R01 and R02 are reset inputs. The counters are reset whenever R01 and R02 are both taken high. On this circuit they are wired to ground as this feature is not required.

+5V Clock Gnd

TP1 TP2 TP3

IC1

IC2

IC3

Figure 3.5 Traffic lights simulator.

Figure 3.5 shows the circuit board and its connections. The power and clock signal are supplied from external circuits. An ordinary switch cannot be used for the clock input as switch bouncing will cause the traffic lights to skip parts of the sequence at random. The most obvious way to test this circuit is to use the input-to-output method. Note that it has both converging and diverging paths. The AND gates have two inputs and one output, so signals are converging. The output from the left hand AND gate goes to two inverters, so the signals are diverging also. A logic analyser would be useful for a circuit such as this. We could record the 3-bit number generated by the counter, along with a selection of other signals, and check for one which is incorrect. Figure 3.6 shows some sample waveforms.

SQA Version 1

73

Developed by COLEG

Electronic Fault Finding

DG3J 35

Figure 3.6 Sample output from logic analyser.

3.3

Suggest a test strategy for each of the following faults: 1. 2. 3. 4. 5. amber permanently off green permanently on green permanently on, amber permanently off amber comes on for two extra cycles when green should be on circuit does not cycle through states.

3.3

The unit specification recommends that you are given repeated opportunities to fault find, locating faults to component level in a wide range of electrical circuits and equipment. The centre will supply you with details of fault finding activities at this point. No detailed information can be here as it depends on the equipment which is available at your centre. It is acceptable to use the same faulty systems as in Section 2, locating the faults to component level for Section 3. The bulk of the time allocation for this unit is allocated to this section, in order to give candidates the opportunity to locate many different faults prior to proceeding to the assessment.

SQA Version 1

74

Developed by COLEG

Electronic Fault Finding

DG3J 35

Summary of this section

You should now be in a position where you are able to fault find to component level. This section covered some examples of fault finding both analogue and digital circuits. Applying these principles to real equipment can be challenging, but expertise comes with experience. Familiarity with particular items of equipment makes repair easier, but you should be able to apply these techniques to a variety of electronic circuits.

SQA Version 1

75

Developed by COLEG

Electronic Fault Finding

DG3J 35

3.1

Fault find one analogue system and one digital system and submit a report. This need not be as thorough as the reports for the assessment, but will need to include: a description of the fault location strategy implemented details of tests carried out and the test equipment and/or diagnostic aids used circuit diagrams a log of the fault-finding activities.

SQA Version 1

76

Developed by COLEG

Electronic Fault Finding

DG3J 35

Answers to SAQs
SAQ 3.1

Symptoms of shorted resistor.

Fault R1 short circuit R2 short circuit Rc short circuit Re short circuit Emitter voltage (V) Base voltage (V) Collector voltage (V)

8.3 0.0 0.9 0.0

9.0 0.0 1.6 0.7

8.3 9.0 9.0 0.0

SAQ 3.2 C1 shorted

Signal supply will pull down the base by an amount dependent on the sources output resistance. This will eliminate the dc bias, as seen in the diagram below, and prevent the transistor conducting. Connect a suitable signal and check the base voltage with a CRO. Note that the oscilloscope needs to be set to dc coupling, otherwise the trace will look exactly the same as the correct signal.
V 1.6V C shorted 1 time

correct signal

The effect of a short on C1.

SQA Version 1

77

Developed by COLEG

Electronic Fault Finding

DG3J 35

C2 shorted

The output will have a dc bias of about 5 V. Check the output with a CRO, remembering to use dc coupling. As seen in the diagram below, the signal will be the same on both sides of the capacitor, a clear sign of a short.
V
V

5V

5V

time

time

a) Collector voltage

b) Output voltage
Symptoms of a short in C2.

Ce shorted

This will have the same effect as a short in Re. If there is no obvious sign of the source of the problem, replace the capacitor first as they are more susceptible to shorts. Resistors are more likely to go open circuit.

SAQ 3.3

1. 2. 3. 4. 5.

As green LED is operating normally, IC2a must be working correctly therefore check circuit for amber, namely IC3c and its connection to the LED. As amber LED is operating normally, IC2a must be working correctly therefore check circuit for green, namely IC3a, IC2b, IC3b, and the connection to the LED. IC2a is the obvious candidate for the fault as the green and amber LEDs both depend on it. (This is an example of divergence.) IC2a is again the obvious candidate for the fault. This is very probably a clock problem. Check the clock input to IC1.

SQA Version 1

78

Developed by COLEG

Electronic Fault Finding

DG3J 35

Glossary
CMOS

Complementary Metal Oxide Semiconductor. A type of integrated circuit which consumes very little power, but is susceptible to static damage." Failure mode in which multiple independent systems fail simultaneously because of an external factor affecting all of them. System structure in which a unit has multiple inputs. System structure in which a unit has multiple outputs. Testing method in which the suspect area of the system is tested in the middle. The process is repeated, narrowing down the search until the faulty component is found. Testing method in which the first unit in a chain is tested, then the next and so on until a faulty signal is detected. Rechargeable battery using lithium salt electrolyte held in a polymer casing. Useful for applications where weight and charge density are important, such as mobile telephones. Rechargeable battery using plates of nickel and cadmium to generate voltage. Testing strategy in which multiple tests are performed effectively simultaneously, and the fault is determined by looking up the test results in a database. Automated test equipment uses non-sequential methods. Fault testing by checking units at random. Testing method in which the last unit in a chain is tested, then its predecessor and so on until a correct signal is detected. Power supply unit. A device which converts mains voltage to one or more low-voltage supplies, usually smoothed dc. Often a sealed unit for safety reasons. Testing strategy in which tests are done in a sequence, one after the other. The result of each determines what test is done next. A logical fault-finding strategy. Device used to determine speed of rotation by generating a proportional voltage. Basically, an electric motor run in reverse.

Common mode failure

Converging paths Diverging paths Half-split method

Input-to-output method LiPoly

NiCad Non-sequential testing

Non-systematic testing Output-to-input method

PSU

Sequential testing

Systematic testing Tachogenerator

SQA Version 1

79

Developed by COLEG