Integrated cellular transceivers: challenging traditional test philosophies

In today's high volume manufacturing of cellular transceivers, cost of test is paramount. With more and more integration of mixed signal and RF components onto the same die, traditional test philosophies are being challenged. Should this highly integrated transceiver be considered as a series of blocks, which need to be tested individually? Would it make more sense to treat the whole die as a radio and test it accordingly? In general Mixed Signal engineering techniques treat the chip as a series of blocks, and RF engineering techniques treat the chip as a radio. The optimal test flow lies somewhere in between. This paper will illustrate, with examples, the tradeoffs between treating a system on a chip or SOC as a radio versus a series of functional blocks. Both approaches have their merits and this paper will give an example test list breakdown for each case. Two real world block diagrams of cellular receivers will be used as examples and the test breakdown needed to cover functionality of all the blocks using both philosophies will be discussed. Each approach will be analyzed to determine its effectiveness and efficiency in final test. To illustrate the differences, two optimized test flows will be presented which contain elements of both approaches and the reasoning behind the tests chosen will be discussed, providing the reader with more insight into writing efficient system tests for radio transceivers. This paper will also highlight the difference between characterization and final test, as well as provide some insight into the differences between mixed signal and RF test.