Using A Capacitor As An Actuator?!

[Dan Berard] has been using capacitors as actuators.

We’ve featured Dan’s awesome self built STM (scanning tunneling microscope) before. These microscopes work by moving a tip with nanometer precision across a surface. While the images he acquired are great, one disadvantage of the actuator he used is its poor rigidity. This limits the system to faster scan speeds.

In his search for a better actuator [Dan] thought he’d try using MLCC capacitors! While not known for their electromechanical properties, you may have encountered capacitors that appear to “sing” (PDF), emitting an audible tone. This is due to the piezoelectric properties of BaTiO3. Effectively the capacitor acts as a weak piezo electric speaker.

Using a 100V drive voltage [Dan] was able to get 300nm of deflection using the capacitor. To extend the range of the actuator he decided to ‘pole the ceramic dielectric’ this involved heating the capacitor above its Curie temperature of 120C. For this he used a transistor to heat the part as an ad-hoc hotplate. This increased the range of the actuator to 800nm, ideal for many STM (and other SPM) systems.

[Dan] is still weighing up his options for his next build, but MLCC capacitors are certainly a cheap and interesting choice.

32 thoughts on “Using A Capacitor As An Actuator?!

      1. There are piezo discs commonly used in humidifiers/vaporizers, rated at tens of watts, which are much thicker than the usual ones. I have no data about their microphonicity but I would expect them to be harder to drive that way too. Also their resonant frequency is well out of the audio spectrum (about 1.6 MHz).

        1. Those look interesting, but I somehow doubt that they would have such a high resonant frequency when used as benders (i.e. clamped only at the edges). Even tiny piezo stacks don’t have resonances that high. Maybe they need to be clamped on both faces? Used this way though, the displacement would be extremely small since there’s only one piezoelectric layer. Buzzers use a flexible brass plate as a sort of motion amplifier, and MLCCs have many layers.

  1. I wonder how accurate it would remain if half a dozen or so of these were used to make a platform move.
    Perhaps there’s a way to use a CCD for extra fine adjustment.

    I wonder how I could measure the force produced by one of these. Maybe it could be used for a cheap home made version of a piezo electronic motor.

    1. Out of curiosity, has anyone found a way to quiet the capacitors that sing?
      I have one motherboard that whistles/sings at different pitches depending on the load on it, and another that would sing as I moved the mouse around.

      I’m curious if something as simple as a dab of electronics safe silicone might be effective…

        1. High frequencies are so sharply directional that your own earlobes may block the sound, and you can only hear the sound or its reflections at certain points in the room while facing certain directions, standing on your left foot and holding your nose.

          There’s actually a device that uses constructive/destructive interference to project sound from two ultrasound speakers. At the focal point of the device, the interference between the two wavefronts causes audible sound. If you put your hand in that spot, it appears like the sound is coming out of your hand, and when you take your hand out the sound vanishes.

    2. That’s probably more likely to be magnetostriction in the inductors used in the switching supplies that are inside of those – piezoelectric effect is definitely a real thing in certain ceramic caps, but coil whine is potentially much much higher energy, and thus louder. It’s common to see SMPS designers soaking large coils in silicone adhesive to dampen vibration (and provide mechanical stability, too).

      1. Considering the number of SMPS’s I’ve seen with capacitors hot glued together (and also with the silicone goop).
        I suspect the capacitors are just another section in the SMPS “choir”.

  2. Very interesting! I’ve had a long burning desire to try and convert some old diesel injectors on an engine I have to electronic control; a piezo stack would be ideal.

    Some of the reading I’ve done suggests that this material only expands with current, regardless of polarity. Dan, have you found this to be true?

    1. An unpoled MLCC will expand, regardless of polarity. This is because the applied voltage temporarily polarizes the ceramic, just as a permanent magnet will align the magnetic dipoles in a piece of iron when brought close. Poling the capacitor though will cause it to contract when reverse voltage is applied, but this can also cause depoling, so the voltage really should be kept positive only.

  3. I’ve accidentally used MLCC’s in the other direction, as microphones…I removed an electret mic from a camera because I did not want to record any sound. Much to my surprise, it still DID pick up sound, however faintly.

  4. I was wondering if anyone is following up on these experiments? Did anyone do a more thorough test of the MLCCs or made a piezomotor? It would be really cool if we could do an Open Hardware piezomotor. Another very nice application could be a spatial light modulator (i.e. a sort of programmable deformable optical mirror) which might be within reach for the deflections at around an optical wavelength. I am sure we can think of many projects that could use these ultracheap and relatively easy to control actuators.
    Placing them on a properly thought out pcb could make a very reproducable and cheap setup with high potential. (Maybe this is exactly what pcbmotors.com did, or did they really go to the trouble and cost of designing their own smd piezo-actuators?)

    1. I’m also interested in any follow-up. I could see them being used for mirror translation in an interferometer. I think the achievable “pixel pitch” may be an issue for a deformable mirror, but not sure. I’d love to see them as slip-stick actuators for nanopositioning and the like, though I think shear mode poling is generally used for that. I’m no expert.

      In general, just some more thorough characterization would get the party started.

Leave a Reply to CRJEEACancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.