We are in section 5.4, Troubleshooting Strategies. Observation and equipment history are important in troubleshooting complex circuits and there are a couple of things that can be looked at. First of all, we'll mention logbooks. Logbooks often can provide a history of component failure. It provides the ability to track specific high failure items and often if there is a history with a given system and it's been well documented it will be found that there are very specific items that frequently fail. If a technician has access to this kind of information frequently, he'll be able to go in and rather than spend many, many hours diagnosing a problem. He can look at, well this is the high failure items and just go specifically to those items and see—well are those defaults at this point?
Also a technician experience. Those who have prepared specific systems to get a sense of what fails and are able to quickly solve those faults and anyone who has worked on a specific system for any length of time you begin to be able to observe those weaknesses that are in the system. You begin to develop a knack for quickly identifying the problem vice a person who has never worked on the system and they have to spend a great deal of time identifying a fault.
Troubleshooting fences are used to identify suspected components in a faulty circuit. This term fence, the idea is that you define boundaries concerning where the fault is located. There are lots of ways to describe this. Your techs use this term called fences and the idea on this over here, we see these blue brackets that are showing an entire system. The idea of fences is that you isolate the problem areas and in this case, down below here now we have bracketed in two areas that potentially are the failure items. Through testing, we've been able to rule out these other items as being part of the problem. Now we have isolated it down to this item and that's the idea of using fences.
Equal probability troubleshooting is a technique that allows you to make a few initial observations and then move troubleshooting fences to include only one or two suspected components. Well, how do you do that? Within the lifetime of a system, certain components tend to fail more often than others and that's what we had talked about on the previous slide. Using this information the probability of a failure being a repeat event can lead to faster diagnosis and repair and as was mentioned in the previous page, logbooks, technicians experience will aid in being able to quickly diagnose a problem.
This technique allows you to locate defects with a few measurements. Essentially the technique involves splitting the circuit path into halves and troubleshooting from the middle of the first half then the second half and so on. The idea is here if you have the circuit and there's many potential test points, you measure at the halfway point. If the signal is good at this point, and you have to know what is a good signal and what is a bad signal, but if you measure here and you have a good reading you know, the problem is probably over here. If the signal is bad, then you would go back and look at this area and then again, you do more tests. This is essentially, what we had talked about, I believe it was in chapter one, where we talked about the idea of divide and conquer.
Again, these were mentioned in chapter one and we would again mention them here. Signal tracing and signal injection are useful techniques when used with the split half troubleshooting method that was from the previous slide. Instead of tracing or injecting the signal from the beginning point of a circuit, the signal is introduced and traced from midpoints within the circuit. If the output is good when a good signal is injected at the midpoint, then the second half of the circuit can be ruled out. The fault must be in the first of the circuit, and the idea in this particular case, and we just looked at this one. The idea would be to inject a known good signal at this point where the signal that should be here and then to evaluate is the signal the signal processed and do we have the correct signal here. If we do not then we know something is afoul in this area.
Once the system has been analyzed and the source of the defect found the suspected defective component should be substituted to confirm its status. You find a component that you suspect is bad then you replace it. The component that you have replaced at any rate needs to be disposed of. You do not want components that are either suspected to be defective or are defective in with components that are no good. Substituted components that do not fix the system should be discarded to prevent their reintroduction into the system. Installing and un-installing components can damage them and this is even in the event that you have components that are good, you install them and you take them from the system those components should not be reused. Semiconductors are relatively inexpensive and they can be damaged by putting them into a circuit, taking them out of a circuit, so it's better to just dispose of those and not to put those in with your known good components.
And this is the idea that sometimes you can test components while they are in the circuit. In some cases, the components of a circuit may be tested while still in the circuit. Open some shorts, may be detected with an ohmmeter while the component is still in the circuit. Specialized test equipment is also available for a specific component in circuit testing and this is a common practice where you actually measure the components in circuits. It's often a tricky thing though because when you measure a process specific component, often times that component is in parallel with other components and so the reading that you would expect with that out of the circuit and what is in-circuit is often very different. You have to really know what you are doing in order to do this effectively.
This wraps up our look at troubleshooting. We finished off talking about testing components while they are in the circuit. We've looked at circuit substitution, briefly looked at signal tracing and injection, split-half troubleshooting. This was the divide and conquer from the first chapter, troubleshooting strategies and we started here on this page, we talked about defenses and that observation and equipment history are important in troubleshooting complex circuits. We looked at logbooks, technician experience. All of these things will contribute to being a good troubleshooter. Probably one of the major things that make a person a good troubleshooter is just simply experience, being exposed to a system, working on it and your troubleshooting skills will grow from that experience.
Video Lectures created by Tim Fiegenbaum at North Seattle Community College.
In Partnership with NXP Semiconductors
by Dale Wilson