2

Which components are the most prone to mechanical failure, via shock or vibration? For example, I know that a speaker coil will generally be more delicate than a fixed resistor, and a IC chip will generally be more fragile than a simple transistor.

Has anyone done a general analysis of mechanical properties and durability of electronic components? Are there any good rules of thumb for this?

  • Many of the problems are avoided by "potting" the board and its components. However, subsequent repair is then replacement, also heat production has to be considered. – Solar Mike Jul 16 '22 at 06:34
  • What happened to make list questions on topic again? – pipe Jul 18 '22 at 05:20

6 Answers6

5

I would like to add:

  • Crystals can break internally on mechanical shock. Sometimes they produce just wrong frequencies or fail.
  • Big electrolytic SMD capacitors or inductors can fall off the board.
  • BGA chips can lose contact, more related to temperature stress
Jens
  • 5,598
  • 2
  • 7
  • 28
4

Usually components themselves don't get damaged by physical abuse such as shock and vibration, the connections to them (and sometimes the PCB boards, if they are brittle) are what fails. Sometime the PCB bends and the connections to otherwise safe parts fails, or they even pop off. Wires and other things can fatigue and break with vibration.

So massive or loose components are a direct risk to themselves and to anything else in the vicinity.

Proper testing of products designed to withstand harsh conditions may including the product being mounted on a electromagnetic shaker and exposed to vibrations of various frequencies, directions and amplitudes. For example, the cheap seats on a launch get exposed to a lot of vibration for a period of time.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
4

Assuming there has been no electrical overstress of the system, it's almost always a connector or a connection, such as a solder joint.

In the case of connectors it could be that a connector's pins were damaged somewhere along the line. Or maybe the connector wasn't masked off properly when the conformal coating was applied. Or the connector had a manufacturing defect.

In the case of solder joints, it could be a cold solder joint or a solder joint that was improperly made (lack of solder filet under the heal of the pin for instance).

This is the purpose behind doing ESS (Environmental Stress Screening) of boards, assemblies, or systems. Part of this test suite is vibration and shock tests, what some call shake, rattle & roll. These tests are to help weed out such connection problems.

SteveSh
  • 9,672
  • 2
  • 14
  • 31
2

The printed circuit board and its mountings. Connectors with wiring harnesses.

jp314
  • 18,395
  • 17
  • 46
2
  • Radial leaded electrolytic capacitors. It's entertaining to watch these part wobble about on a vibration table with a strobe). Many open frame power supplies now put a blob of adhesive on tall capacitors to adjacent mechanically stable parts. In the old days, we had to do the the gluing.
  • Anything with socket connections (connectors without latches, ICs in sockets). I've had parts fall out of sockets during hammer tests.
  • PCB mounting points can fracture or break. Seen this on cheap consumer electronics.
  • Electrical connections via screws can become problematic as seen on a certain manufacturer of modern TV sets.
  • Mechanical brackets can break causing excess stress on the PCB.
  • Soldered wire connections to a PCB are prone to breaking if the wire/harness isn't secured properly.
qrk
  • 7,474
  • 1
  • 5
  • 20
2

Speaker coils are made of copper wire, which is quite flexible and shock-resistant.

Conversely, ceramic capacitors, for an example of a component on a PCB, are made of fragile material which is not intended to be flexed. But they are attached rigidly, by solder, to a flexible board of resin-bonded fibreglass. The board and the copper tracks on it can flex without damage; the rigid components mounted to it not so much.

You can research how ceramic SMD capacitors should be aligned to minimise stresses that contribute to their failure.

However, for smacking a skateboard the wrong way on a kerb, I'd suspect an imperfect solder joint: pushed one way it could connect sufficiently, but with a bit of movement in a different direction it might no longer make good electrical contact. Another hit in another direction could move it back into contact temporarily.

Andrew Morton
  • 2,389
  • 1
  • 17
  • 26