The idea
If a circuit has a chain of stages, the signal or voltage can be checked at the midpoint. If it is correct there, the fault is further on; if not, it is before that point. Each measurement rules out about half of the circuit, so the number of tests grows very slowly with complexity.
How it is done
- Define a known input and an output that should be observed.
- Choose an accessible test point near the center of the chain.
- Compare what is measured with the expected value, taken from the schematic or from equipment in good condition.
- If the value is correct, the defect is downstream; if it is incorrect, upstream.
- Repeat on the suspect half until you are left with one stage or one component.
An example in a power supply
If a power supply gives no output, first measure the rectified and filtered voltage on the primary. If it is correct, the problem is in the switching, the transformer or the secondary; the next measurement is then on the secondary winding. Each result reduces the suspect area.
When it works poorly
- Feedback loops: in a circuit with feedback, a defect at any point alters all the others and the midpoint can be misleading. It is best to open the loop or break it in a controlled way.
- Inaccessible test points: if the center cannot be measured, choose the nearest point.
- Multiple or intermittent faults: the binary logic assumes a single stable defect.
- Loads that drag the signal down: a short circuit downstream can make the signal look bad upstream.
How to interpret the result
Bisection locates the area; it does not confirm the component. Once you reach a small stage, move on to component tests with the circuit de-energized. If circuits with hazardous voltages are measured, this must be done by qualified personnel following the applicable regulations.