I was given the task of redesigning the arming timer in the AGM-130 air-to-ground missile. The purpose of the timer is to delay arming the missile warhead until it has had plenty of time to clear the airplane. The timer originally consisted of a UJT (unijunction transistor) and an SCR, which applied 28V DC to the trigger electronics a few seconds after launch. Because the UJT was no longer available, I redesigned the timer circuit using a MIL-approved version of the 555 timer chip.
Once I modified the circuit board, I gave it to a technician to test. He ran only the affected test. He did not test the entire circuit card. When the technician gave me the card back with the test printout showing that the circuit was working well, I released the design. My employer had a big batch of 7-layer circuit boards made at a cost of about $180K (I think). When the finished circuit boards were built and tested, they failed one of the other tests.
A contract engineer troubleshot the circuit board and found that my circuit had caused a sneak path that interfered with one of the other circuits, causing it to fail its test. The contract engineer added a diode to block the sneak path, as my circuit was otherwise OK. Rather than throw the boards away, they added the diodes manually and added "white wire" jumpers to the boards.
The obvious lesson here is to test everything to make sure a circuit change does not affect something else.
This entry was submitted by Andrew Morris and edited by Rob Spiegel.
Andrew Morris is a retired electronics engineer who spent most of his career designing automatic test equipment for the defense industry. The last five years of his career were spent designing electronics for automatic dispensers for the paper industry. He has a BSEET degree from Virginia Tech.
Tell us your experience in solving a knotty engineering problem. Send stories to Rob Spiegel for Sherlock Ohms.