Recently one of our newer customers started on their first solo board setup. They had previously seen XJTAG demonstrated and had also had their initial board setup done by an XJTAG engineer. When they came to setting a board up for themselves they tried to follow the path that the XJTAG engineer had talked them through – and one of the things they’d been shown was that very early on in the setup we categorise links, zero-ohm resistors, fuses etc as passive connections, even though, being on power/ground nets these items are not accessible from or directly testable by JTAG.
The reasons that we make these categorisations (and recommend that you do too) are:
a) Circuit designers often do not give helpful names to spurs from their power nets. This makes it unclear, when you see test results, whether the net involved in an error is power (rather than N3775640 or something like that).
b)Once XJTAG knows about the real extent of the power and ground nets in the circuit it can make much more sensible suggestions about which resistors are pull resistors (and these are quite likely to be testable once XJTAG knows what they are. It also helps to prevent other mis-categorisations of components later in the project setup.
c) When you come to look at the test coverage of your circuit on the DFT screen, the picture is much more complete if the power and ground nets are accounted for properly.
Anyway – we figured if we set out an explanation of why our experienced engineers take this approach, others can learn from it and get better test coverage as a result!