Choosing an In Circuit Tester

• Low cost tester
  
• Large tester capable to handling large number of nodes
  
• Flying probe tester where probes move to the nodes to be tested This negates the need for a bed of nails test fixture but is slower
  
• Combinational tester capable of testing boards using a variety of test techniques
  

Selection considerations

The choice of the correct type of tester depends upon a variety of different elements. These include the component technology used. Very densely packed surface mount boards will need a different type of tester to one that might use conventional leaded components. The production volume and throughput required will also have a major impact on the choice. Also the mix of board types and whether there are frequent changes of the type of board to be tested are important factors. Further information that is essential is the anticipated process yield and fault spectrum. While these two factors cannot be determined from real figures for boards or assemblies that are being designed, they can nevertheless be estimated quite accurately from existing boards that are in production.

Collecting information

Determining the yield of a process is relatively straightforward. It is simply the number of boards passing successfully through the process divided by the total number entering. Collecting these figures is a simple number collecting process. Once collected, these metrics give a good indication of the process. The higher the yield, then the better the process. If yields are low then this indicates that higher levels of testing are required.

Batch size is another important issue. Small batches require frequent changes of fixture, reloading of programmes, and the like. Small batches result in a reduced throughput, but this is often necessitated by the requirement for the product, and the availability of the equipment. If batch sizes are small, then a machine tailored to suit small batches needs to be chosen. Often a roving probe machine may be suitable because fixturing costs are less, although they run more slowly than machines that use a bed of nails. This is because the probes physically move from one node to another to undertake the test.

A knowledge of the fault spectrum is important in determining which In Circuit Tester ( ICT ) to use. Unfortunately this is not always easy, but it may be possible to gain data from similar boards that are currently being produced. Adjustments can be estimated according to any knowledge of the board to enter production. With a knowledge of the typical fault spectrum anticipated, a suitable tester can be chosen. For example if most of the faults found are short and open circuits, then the more expensive device testing facilities may not be justified. It may be more cost effective to perform a different type of test to locate the other tests.

Choices

It is necessary to adopt "production orientated" approach to testing. When developing a test strategy there is often a tendency to want to test everything. This approach is particularly expensive as test time and investment in the ICT machine and the ICT programme increase to enable this over-comprehensive test to be undertaken.

Instead, look at the throughput to see the level of machine required, and the fault spectrum to judge the facilities required in the machine to perform the tests that are needed. Also develop the test ICT test programme to locate the likely tests. There is no point in spending time testing for faults that never occur.

By adopting an approach that meets the needs to the requirement, considerable amounts of money can be saved. However it is also necessary to ensure that the whole test strategy fits together so that each stage of testing performs the required tests most efficiently. In this way a quality product can be assured at the lowest cost.