The worst sort of failure mode is the one that is invisible -- no tests, root cause analysis, or part swap-out can ferret out the root cause.
In the EDN Tales From the Cube installment Time Bomb: The Case of the Invisible Failure Mode, Engineer Walter Lindenbach shares his ghost-hunting experience when he was called in to investigate a peculiar problem with his company's oil-rig-monitorinig device. Time and time again, the transistor that switched the relay coil would fail within a year of installation and the units would be returned for repair.
Though it took some effort, Lindenbach describes his success at finding the failure mode and correcting the problem:
"The relay-coil resistance was about 240Ω (12V dc at 50 mA)--by no means an excessive load for a small transistor. But a little 12V relay coil can produce a 200V spike...by turning the vertical sensitivity [on my scope] down and down again, the peak of the spike was finally visible--at 200V! ....A transistor with a collector-to-emitter breakdown voltage of 80V will not last long under those conditions, and, if you don't protect it from the inductive spike, the transistor will fail after some number of switching operations--in minutes or months."
Subscribe to this blog's feed
Comments (1)
Instead of using a resistor, I would suggest a surge absorber. Adding a resistor limits your current at all times, making it a less than ideal solution.
As the name implies, these parts absorb transients and are the way you would normally deal with large spikes from motors and relays.
Digikey sells them in bulk for a few dollars. Here's the part I used to dissipate the transients from a 2HP motor:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=P7278-ND
There are different sizes available for almost any application.
Of course, this leads to another question - couldn't the circuit use a solid-state relay (SSR) instead of an electro-mechanical relay (EMR)? There's no inductive spike on a SSR. If you're concerned about failures, a CSA or UL approved SSR relay is going to last significantly longer than your $0.05 resistor. You might even be able to find a drop-in replacement.
Posted by themagni | July 30, 2008 4:51 PM