Microswitches: Past the Tipping Point

You find them everywhere from 3D printers to jet airliners. They’re the little switches that detect paper jams in your printer, or the big armored switches that sense when the elevator car is on the right floor. They’re microswitches, or more properly miniature snap-action switches, and they’re so common you may never have wondered what’s going on inside them. But the story behind how these switches were invented and the principle of physics at work in the guts of these tiny and useful switches are both pretty interesting.

The Chicken Switch

Microswitches have a long history, at least as long as it takes for the original trademarked “Micro Switch” brand to become a generic term. Back in 1932, a company called Burgess Labs in Wisconsin got an order to build 10,000 electric chicken brooders. Charles Burgess, the head honcho of the company, ordered 10,000 switches from a manufacturer, but was dismayed to find out that the switches were not up to snuff. They wouldn’t consistently switch at the same actuation point, and clearly wouldn’t work for the brooders he was already on the hook for. Dr. Burgess assigned one of his mechanics to develop a better switch. Phillip McGall eventually came up with a snap-action switch with the actuation repeatability required for the chicken brooders, and the order was filled.

Fast forward a couple of years, and McGall had patented his “Snap switch.” Burgess Labs started marketing the switches, which ended up in Rock-Ola jukeboxes and other electromechanical products of the growing consumer electronics industry. Burgess eventually sold the Electronics Division of his company, which was turned into the Micro Switch Corporation and later sold to Honeywell in 1950, where it remains a top brand today.

Snapping Into Action

As Dr. Burgess discovered, a plain spring contact in a switch without snap action has a long, often variable throw and a “mushy” action. The principle behind McGall’s snap-action switch is a tipping point mechanism, where the common contact in his switch is made out of a preloaded spring.

The force of the spring keeps the contact in a stable state up against the normally closed contact. When a small amount of pressure is applied to the common contact through an actuator, the preloaded spring deforms at a characteristic and repeatable point, rapidly moving the contact to another state against the normally closed contact. The important feature is that the force applied through the actuator is not directly moving the common contact between normally open and normally closed. The force to move the common comes from its spring tension alone; the actuator just provides the force to deform the spring until it pops through.

A modern microswitch. By Benjamin D. Esham, via Wikimedia Commons

The tipping point mechanism in microswitches has quite a few advantages over plain contacts. The repeatability of the mechanism is the big one, as is the rapid make-break cycle. Once the tipping point is reached, the contacts can often switch states in just a few milliseconds. This reduces the chance for arcing and minimizes the time the contacts are in an ambiguous state.

Another benefit of microswitches is their low operating force. Even without the mechanical advantage of a long lever arm bearing on the actuator, only a small amount of force is required to pop the spring and toggle the switch. And the design is flexible, allowing switches that can be tiny enough to be built into the smallest stepper motors to industrial limit switches that are designed to survive the harshest conditions and last for millions of cycles.

Modern microswitch mechanisms hardly differ from McGall’s original patent illustrations. Materials have changed, of course, and there have been a few different designs for preloading the common contact — dished springs, for example, or forked springs where a center “tine” is anchored in a different plane than the outside members. But that the basic principle of these simple devices is now over 75 years old and still used is a testament to good design and solid engineering.

Banner image source: eRittenhouse.org, by Allan Mills.

46 thoughts on “Microswitches: Past the Tipping Point

  1. I have had several mice go “bad” as the switch gets so bouncy that one click is counted as two. And my iPad’s toggle switch is crapped out. Mechanical contacts r garbage.

        1. Untrue…

          Not a microswitch…

          And that’s what’s in a majority of mice.

          They do use a snap spring though, but its a disk and can crack. Also the cause of “failure” isn’t even the switch quite often, but the plastic pip under the external button cap has hammered itself flat and can’t push far enough.

          1. ^^^^ Comment with more than one example awaiting moderation….

            But maybe the microswitch is becoming more common, everything I’ve taken apart, and not to get into a dick swinging contest, but that is also dozens, has had the other, but most would have been 5 years old plus, notice on 2 of the mice kicking round my desk, I’m getting a tik-tik which definitely sounds more microswitchey than the tok-tok of the other 2… Was gonna pop them open quick, but risks destruction of the slider feet, so won’t bother.

          2. I’m an electrical engineer and I have had to spec these switches several times. When you spec them you have to specify button-push count expectations. The cheaper units have low counts, like 10K or so. The ones with high counts can be double the cost. In the end, it doesn’t matter what you spec the pencil pusher with the cost-sheet will switch it out for the cheaper (low-click-count) button. Because they don’t care if you have to buy another one, they only want you to buy this one.

          3. this type of switch is not used is mice nowadays, and I dont know if it ever was for main buttons. I only see those on cheap mice as side/dpi buttons.

          4. I’ve seen both tactile buttons and sub-miniature microswitches in mouses.
            Razor DeathAdder = microswitches
            Razor Orochi = tactile buttons (very high end ones, but still)

      1. At least one of my mouses and both of my trackballs use very small (sub-minuature) microswitches without the hinge lever.
        If your mouse uses one of those surface mount tactile buttons, then that really sucks for you. (cheap mouse)

        Examples: Omron SSG-01H-5, Omron D2LS-21, and others.

        1. All my mice use microswitches, except one that uses standard through-hole tactile switches. It’s very small and I believe the main reason it uses normal switches is to save space (its PCB is quite densely populated). Even microswitches (at least the crappy unreliable ones used in mice) are dirt cheap today, and I never bought a mouse for more than $2.

          1. Are there “crappy unreliable” microswitches? I can’t imagine how one wouldn’t be robust.
            The little membrane pieces of crap you find on PCBs of course are crappy and unreliable, but those shouldn’t be confused as microswitches. (tactile button)

    1. Given you had the failure several times, I would question either the quality of mice you are purchasing, or how you’re using it. I’ve one ever had 1 mouse failure from microswitch fatigue, and that was more due to the type of microswitch installed in the mouse, not the quality of mechanical switches in general. If you’re hammering the mouse button like a woodpecker looking for it’s lunch, of course it’s going to fail earlier.

    2. I had a thumb-ball mouse do this. It worsened over a period of time. Every mechanical transducer has a limited number of cycles available to it and software can make it worse by making one switch the entire interface. Speak-and-spell units used to have the keyboards fail in approximate analog to the frequency of the letters in the words. The ‘E’ key was often the first casualty.

    3. I just fixed thiat issue the other weekend with the left button on my aging steelseries Xai mouse. Easy fix, A little bend and works like a champ once again, although I had to pull out the good tweezers to handle the spring, ok the fix was easy, putting it back together was painful. Ugh springs springing…

    4. Re:mice
      Depending on how much you liked those mice, and depending on the wiring of the switch, you could’ve just added a cap (switch common & GND) for debouncing. If they were the typical SPDT you probably could’ve hooked the NC lead to the opposite voltage level so as to recharge or discharge the cap decisively. That kind of assumes the IC input was on the switch’s common, and the NO went to GND or 5V (active low or high rsp.), otherwise you’d need to swap them.

      Re: iPad
      That seems probable.

      Re: ‘Mechanical contacts r garbage’
      All contacts are mechanical. ;)
      Debouncing is a thing, so ordinary that it showed up in the tech college 1st semester digital electronics textbook. I once looked at the dual photointerruptor that watches the mouse wheel and said, “why isn’t the switch just one of these and an opaque bit that drops in?” At least lots of printers do exactly that to sense the head reaching some position…

      1. Mouse buttons don’t use optical interruptors because they require a tactle click when you press them, and to spring back into place. By the time you’ve implemented the mechanics for that, you’ve basically done three-quarters of a mechanical button anyway, so you may as well just use one. Plus buttons are well specified, cheap to fit, and don’t break down all that often.

        I’m glad to see ball mice die. I spent half my time cleaning rollers, and the other half dragging a half-responsive mouse across the table in front of me. How we got Atari mice to last for so many years I’ll never know.

        1. I know it’s overkill, and yes of course something has to click when you click on something. It would be a premium part for insane gamer mice anyway. I bet that dual photointerruptor could be worked into the sidewall of the usual part so that the existing switch still clicked in exactly the same way under the same force, but the contact would cast a shadow on one sensor or the other.

          1. A4tech uses optical switches in their Bloody gaming mice, they label them light strike switches, just a microswitch mechanism with the contacts replaced by a photointerrupter. they also replace the mouse wheel encoder. the switches are only rated at 30 million clicks so not a vast improvement over the mechanical type.

    5. There’s cheap switches designed for where it doesn’t matter, quality switches for industrial applications, mil-spec switches, redundant switches for use in the nuclear industry, and then there’s the switches designed to handle the abuse of dirt, dust, food, and bodily fluids which a typical mobile is subjected to…

  2. Back in the time when mice had balls, I fixed quite a number of bad left button switches with dome springs. I’d carefully cut the four melted over plastic studs to get the metal top and plastic button off. Then I’d do the same to a switch in another mouse that had some other problem.
    Swap the good dome over then use a needle to apply a little super glue to hold the switch together.

    Was a common thing for the domes to crack and lose their snap.

    Many 2600 compatible joysticks had the same problem but were much easier to fix. They often used snap domes (1/4″ or larger diameter) simply mounted atop PCB traces with pieces of heavy packing tape. Get two bad sticks and make one.

    1. Broken wires in the cords right where they exited the mouse body was another common point of failure. One thing I noticed was that the higher priced mice usually did not have a molded on strain relief while cheaper ones, except the absolute cheapest, did. Several of my mice ended up with their cords shortened about 1″, with heat shrink tube added to keep the cord from flexing too sharply where it exited the front.

      If only someone would reproduce the Microsoft Trackball Optical 1.0, the thumb ball model. But with silicon carbide instead of cheap steel support points.

    2. Did you not know how to solder?

      Back in that era 3 button mice had come in, but barely anything made use of middle, or you could emulate it with LR together, so I tended to regard the middle switch as a handy spare.

  3. I haven’t seen a micro switch in the paper path of a printer for a very long time. Normally the paper actuates a lever that has a flag at one end and the flag interrupts a photo interrupter.

    The good thing about micro switches is that they have a high transition speed and a very short bounce period so they are easy to debounce in software as they give quite consistent results until they fall apart.

  4. I have a Logitech FX Ball Mouse that I had to “clean” (accidental bath in spilled Mountain Dew…) and yup, four (yes 4) actual micro switches. I’ve used this mouse for almost 15 years.

  5. I didn’t realise the poultry industry was responsible for microswitches!

    On the turkey farm, they were pretty reliable, I’d replace single phase motors often and had to keep spare 750W/240V motors on hand, just in case, the microswitches that controlled them were very reliable, I only kept a couple on hand, just in case.

    I had a fault that took me an hour or two to figure out, one feedline kept tripping the breaker, the motor was fine, the switch was “fine”, it wasn’t a dead short, it still “switched”.

    I replaced the switch, everything worked again.
    I couldn’t let it go, examining the switch I found one of the poles had arced through the case to the mounting bolt.

  6. Good quality micro-switches surprise me:
    Often rated at 10-million operations.
    Used directly as a limit switch, re-homing repeatability is 0.003″

    Never seen a micro-switch used for paper detection in a printer.
    Mostly I’ve seen opto: photo-interrupter, either direct or with mechanical arm.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s