6

We have a test that checks whether or a push button works or not, and as of now the test is done manually by hand. My question is how to do this automatically?

We have considered using a solenoid to push the button but have some hesitations on whether this might damage the somewhat delicate button or the PCB itself. The button is similar to this but mounted perpendicular to the PCB.

enter image description here

Some solutions I thought of:

  • Using a small solenoid with a damper so it doesn't accelerate quite as aggressive

  • Adding a spring between the button and the solenoid

  • Using a small linear actuator

  • Using a servmotor to actuate the button

I would prefer if the "buttonpusher" would revert back when the power is removed so I am somewhat biased against using a linear actuator or the servomotor.

Heneer
  • 489
  • 7
  • 12
  • 1
    I assume the PCB must be moved into and out of the test jig. Although it presents other (power/timing) constraints, this movement could be used to test the switch against a static 'finger'. – amI Jun 28 '18 at 17:40
  • How long do you need the push to last? – joojaa Jun 29 '18 at 04:54
  • The board is first placed in a jig, then a lid is closed around it as a safety measurement. As of now the button is pressed for about 1 second, however I don't think having a different length is an issue. – Heneer Jun 29 '18 at 08:19

5 Answers5

7

Instead of using the actuator to push the button, arrange something with a limited force to push it, like a weight or a sprung lever. Then use your solenoid to pull it back. This allows you to use any solenoid that's powerful enough, but limit the force to your delicate device under test.

Neil_UK
  • 158,152
  • 3
  • 173
  • 387
  • 1
    The requirement is to stay off on power failures. – Dorian Jun 28 '18 at 17:17
  • @Dorian I read that as a preference, hinted at by his use of the word 'prefer'. However, I didn't elaborate *how* the solenoid could pull it back. There are two ways. The first is for the solenoid to pull it back directly. The second is to iterate the idea, so a second spring pulls the actuator back, and the solenoid defeats that spring. – Neil_UK Jun 28 '18 at 17:44
  • It's fair, actually the best answer is in the question. – Dorian Jun 28 '18 at 18:08
5

Use a solenoid with a dashpot to slow it and a rubber tip to contact the button.

3

You can also use an eccentric attachment (a crank) on a shaft of a regular rotational motor. You'll have to ensure that it stops its rotation in a specific sector, but that is trivial to implement

Orson Maxwell
  • 109
  • 1
  • 4
  • Sort of like this. - https://lifehacker.com/5758231/hack-an-air-freshener-into-an-remote-camera-trigger – KalleMP Jun 28 '18 at 22:01
2

If you only need to quickly press the button and then release you can use a voice-coil actuator that will return to rest with power off.

The easiest voice coil actuator to find is a BIG speaker and just glue a pin with a guide to the dome and make sure you do not leave DC current (series capacitor perhaps) running for long enough to overheat the voice-coil.

This will not be a compact (or very cheap) solution but it should be robust and reliable with near infinite operation capability.

KalleMP
  • 4,039
  • 1
  • 14
  • 32
  • 1
    Yeah -- any old junkyard car subwoofer will do I reckon, so it's probably not too expensive if you do some digging around ;) – ThreePhaseEel Jun 29 '18 at 00:35
  • 1
    hard drives use voice coils too, but optimised for pivoting motion, old hard drives may be more plentiful than large loudspeakers. – Jasen Слава Україні Jun 29 '18 at 04:22
  • @Jasen I knew I had seen more voice coils out there. The hard drive one is much more compact, might work with a simple lever pivoted near the bearing. – KalleMP Jun 29 '18 at 08:35
  • @ThreePhaseEel Yes, there is one next to my storage unit out in the rain, the condition of the paper cone does not matter much if there is still some left to provide the required guidance. – KalleMP Jun 29 '18 at 08:36
0

Use a solenoid with a damper attached.

I'm not a mechanical engineer but I think you can figure something using a door damper like the one in this video and a lever to scale the force needed.

Dorian
  • 2,546
  • 2
  • 11
  • 20