IPC-D-356A (or ET vs. CAM)

I was looking over some statistics from [www.solidigm.com] and noticed that I'm getting alot of traffic from people "googling" for IPC-D-356A (or IPC356A, or IPC-356A, or similar). Electrical test in general, and netlist formats in particular are a difficult subject with most of the PCB CAM-types that I typically hang out with. Often in a printed circuit fabrication facility, the electrical test department is kind of "off on their own"...sometimes better described as "...in their own little world". Oddly, most electrical test engineers I know describe CAM departments the same way.

Frontline and other CAM vendors have made significant pushes into the world of electrical test fixture generation, but that is still largely a world unto itself.

Some fabricators argue (with some merit) that separation of CAM and ET inherently adds another check into the process. This additional check comes (as it often does) in the form of duplication of effort. Specifically, in this process, CAM reads in the gerber data (or whatever format the designer happens to send) from the CAD system along with netlist data if it is present. That same "original customer data" is then passed along to electrical test, where they read it into their (typically disparate) systems. Using two different softwares to interpret the incoming data will, in theory, catch more issues. It could be argued, of course, that this process can also create more issues.

Anyway, here are some interesting links on the topics of netlist formats and electrical testing.

This article is a rather dated, but still informative regarding IPC 356 formats.
[ ipcd356_testtransfer.html ]

This is actually a chapter from a book. The book is on bare board PCB design, the chapter is on electrical test.

[ bbchpt2.html ]