Don’t be Intrusive!

Ever heard the old saw about the guy (gal) that “went to a fight and a hockey game broke out”?

I’d characterize the ongoing debate on the value and longevity of In-Circuit Test (ICT) as a bit of a brawl…

There have been a number of industry forums held recently which have debated the ongoing role of In-Circuit Test in today’s test strategies. With the increasing complexity and density of modern circuit boards diminishing ICT’s test coverage, and new BIST-based tools emerging to supplant it, it’s healthy to have a discussion on alternatives. As long as no one gets hurt!

I’d profile the debate as follows:

In defense of ICT:

  • It’s what we know.
  • It refuses to die.
  • It would be scary to be without ICT!
  • The equipment is paid for…
  • If it went away, we’d put thousands of machines on the sidelines.
  • Boundary scan will be the savior of ICT!
  • If it went away, we’d put thousands of ICT engineers out of work.
  • It works for me!

And attacking ICT:

  • Circuit density is trending toward eliminating physical test access
    • Backdrilled vias, blind vias, buried vias, HDI (microvias), press fit.
  • Circuit speed is trending toward eliminating physical test access
    • Surface routes can’t tolerate any test points (except, perhaps, bead probes).
    • Critical routes don’t come to the board surface (which prohibits bead probes).
    • No signal routes come to the board surface; surface layers are planes only (which prohibits bead probes).
  • Bead probes were dead on arrival anyway
    •  Signal integrity (SI) and electromagnetic compliance (EMC) concerns killed them…
  • Newer methods such as built-in self-test (BIST) and hybrid functional can’t run on ICT.
  • Maintaining physical access is too much of a design issue.
  • Boundary scan at ICT is too expensive!
  • Boundary scan at ICT is signal integrity prohibitive!

Although everyone seems to agree that the old bed-of-nails is diminishing in value and being replaced with non-intrusive board test technologies, even those who attack ICT for sport concur that it will never totally die. On production lines where it isn’t eliminated, ICT might be moved to a sampling basis, or used for debug & repair.

Alan Sguigna