Basic Troubleshooting Strategies

A very powerful method for discerning cause-and-effect relationships is scientific method. One commonly accepted algorithm (series of steps) for scientific method is the following:

1. Observation
2. Formulate an hypothesis (an educated guess)
3. Predict a unique consequence of that hypothesis
4. Test the prediction by experiment
5. If test fails, go back to step #2. If test passes, hypothesis is provisionally confirmed.

This methodology is also very useful in technical troubleshooting, since troubleshooting is fundamentally a determination of cause for an observed effect. Read the following description of an experienced troubleshooter diagnosing an mechanical noise problem in a bicycle, and match the troubleshooter’s steps to those five steps previously described for scientific method:

One day a bicyclist called a mechanic friend of his over the telephone, and describes a problem with his bicycle. The bicycle is making a rhythmic “clicking” sound as it is pedaled, but the bicyclist is not very mechanically inclined, and cannot determine the cause of the noise. The mechanic considered some of the options. Being a rhythmic noise, it was probably being caused by one of the bicycle’s rotating objects. This includes the wheels, crank, and chain, which all rotate at different speeds. After a bit of thought, the mechanic asked his bicyclist friend a question. “Does the pace of the clicking increase as you ride faster?” The bicyclist answered, “Yes, it does.” “If you shift into a higher gear so that your crank is turning slower for the same road speed, does the pace of the clicking change?” asked the mechanic. The bicyclist admitted he didn’t know the answer to this question, as he hadn’t thought to pay attention to this detail. After riding the bike once again to test the mechanic’s idea, the bicyclist reported back. “No, the pace of the clicking does not change when I shift gears. It only changes with changes in road speed.” Upon hearing this, the mechanic knew the general location of the problem, and continued his troubleshooting over the telephone with further questions for the bicyclist.

Where is the clicking sound coming from on this bicycle, based on the information presented here? How do you (and the mechanic) know?

A very powerful method for discerning cause-and-effect relationships is scientific method. One commonly accepted algorithm (series of steps) for scientific method is the following:

1. Observation
2. Formulate an hypothesis (an educated guess)
3. Predict a unique consequence of that hypothesis
4. Test the prediction by experiment
5. If test fails, go back to step #2. If test passes, hypothesis is provisionally confirmed.

This methodology is also very useful in technical troubleshooting, since troubleshooting is fundamentally a determination of cause for an observed effect. Read the following description of an experienced troubleshooter diagnosing an automotive electrical problem, and match the troubleshooter’s steps to those five steps previously described for scientific method:

One day a car owner approached a mechanic friend of theirs with a problem. The battery in this car seemed to be dying, requiring frequent jump-starts from other vehicles, or the application of a battery charger overnight, to be able to start reliably. “What could be the problem?” asked the car owner to the mechanic. The mechanic considered some of the options. One possibility was that a parasitic load was draining the battery of its charge when the car was shut off. Another possibility was that the car’s charging system (the engine-driven generator and its associated circuitry) was faulty and not charging the battery when the engine was running. A third possibility was that the battery itself was defective, and unable to hold a charge. “Let’s check the battery voltage with the engine stopped, and with the engine running,” said the mechanic. The two walked over to the car and opened the hood, then the mechanic connected a voltmeter to the battery’s terminals. It read 11.3 volts DC. This was a 12-volt (nominal) battery. “Start the car,” said the mechanic, still watching the voltmeter. As the electric starting motor labored to turn the engine, the voltmeter’s reading sagged to 9 volts. Once the engine started and the electric starter disengaged, the voltmeter rebounded to 11.2 volts. “That’s the problem!” shouted the mechanic. With that, the owner stopped the car’s engine.

Explain which of the three hypotheses was confirmed by the voltmeter’s reading, and how the mechanic was able to know this.

Troubleshooting a system of any kind requires scientific thinking: sound deductive reasoning from effect to cause, and cause to effect. One of the principles frequently applied in science is Ockham’s Razor, named after Sir William of Ockham (1284-1350). In Ockham’s own words, the principle is as follows:

Applied to troubleshooting electric circuits, one could re-phrase Ockham’s Razor as such:

Justify the use of Ockham’s Razor in troubleshooting circuits. Why should we first consider single faults to account for the problems the circuit is having rather than considering interesting combinations of faults which would account for the same problems?