MEMS Caps that Push Both Ways
by Bryon Moyer
February 21, 2013 at 12:09 PM
The many bands that cellphones must support, coupled with environmental changes that can have a dramatic effect on the effectiveness of a phone’s antenna, have made the concept of antenna “tuning” particularly relevant. We looked at WiSpry’s approach to this some time ago. The concept involves an array of capacitors that can be reconfigured in real time to change the characteristics of the antenna and improve reception.
Most such MEMS capacitor arrays consist of cantilevers – like diving boards – or bridges – basically, cantilevers supported on both ends. (Which… makes them not cantilevers… yes, I understand…) But there’s a new company, just on the heels of a funding round, that has proposed a different way to build the capacitor.
DelfMEMS (I was so expecting them to be associated with Delft, the Netherlands, but no… they’re French) points to stiction as a particular problem for these traditional structures. Stiction is the tendency of a variety of influences – atomic forces, residual gunks, etc. – to cause a micro- or nanoscale structure to stick when you don’t want it to. So, for instance, if you push a cantilever down until it touches a base surface, it may stick when you remove the force you used to push it down.
Some companies actually use this as a way of holding down the cantilever, but in general, the issue comes when trying to release it. Assuming you’re not intentionally using stiction as your hold-down mechanism, DelfMEMS says that the traditional approach to avoiding stiction issues when opening the contact is to rely on mechanical force sufficient to overcome the stiction. And this mechanical force is typically provided by using a really stiff beam structure, located relatively far from the contact surface. That way it’s pulling away hard.
The downside of that is that you need a very high voltage – on the order of 50-100 V – to bend the thing when you want to actuate it.
DelfMEMS has a different structure. Viewed from the side, you can think of it like a bridge, except that, instead of being anchored at the end of the beam, there are two pillars placed in from the ends, and the bridge rests on them. This means that the beam can flex both so that the middle bows down, with the ends raising up, and so that the middle bows up, with the ends flexing down. The key is that they put actuating electrodes both in the middle, to pull the middle down and make contact, and under the ends, to pull the ends down, which raises the middle and breaks the contact. In other words, you get electrical help in both directions. And that reduces the voltages needed to less than 20 V.
This appears to be the essence of what they bring to the table, although they talk about other details in a whitepaper on their website. You can find out more about them and their recent funding round in their release.