I see a wide range of opinions and suggestions on this topic, particularly on technical forums, ranging from complete nonsense, pure fiction and guess work to well thought out logical suggestions. My first manager back in the early sixties thought that most good engineers could find faults with just a test meter. Somewhere between these extremes lies a good working solution.
Any approach to fault finding will be determined by the type of equipment and fault and whether the unit is a prototype or a production item. In the former case the design may not be correct whereas in the latter case the assumption has to be that the underlying design is correct.
I have never believed in changing random components in the hope that one can find a fault. It is better to analyse the symptoms in order to localise the fault.
Testing old and often valves based equipment that has not had power applied for many years is a special case -
ALWAYS check the supply lines for any remaining voltage from charged capacitors from a previous use and discharge with a resistor, not a screwdriver. Then do a visual inspection first with the unit disconnected from any power source. Investigate any wiring, component, connector or printed circuit area that looks as though it has overheated, expanded, leaked or has an appearance that looks suspicious.
If you can safely apply power then a logical process can be implemented working back from the section that is working correctly. If you cannot apply power because of a catastrophic fault (the magic smoke has escaped) then the faulty component(s) must be found using a test meter or similar and fixed or replaced before power can be applied.
Most established commercial designs will have a maintenance manual with test instructions and data and a series of test points that may be used to measure voltages that are defined in the manual. If you do not have a manual then think about the operation of each stage and what should be expected.
A lot of fault finding is just basic common sense and does not require sophisticated test equipment. If a valve or transistor is not conducting when it should be then there will no voltage drop across its cathode or emitter resistor and that is very easy to find. Any bipolar transistor requires a (small) voltage on the base to make it conduct -
Audio amplifiers can be roughly tested by setting the audio gain to maximum and touching the input line with an insulated screwdriver -
The most difficult area of fault finding is inside closed loops because when you open the loop to test each individual stage the circuit will run in an uncontrolled manner and the gain may be significantly increased bringing other problems. Phase locked loops and regulated power supplies are typical examples of this class of circuit.
More to follow….