WHC 2: penta-vibrating grip
posted Feb 1st 2007, 23:09 under arduino, tactile, vibration, WHC
For this second Weekly Hardware Challenge (WHM), I decided to explore tactile signal resolution: is it possible to easily identify the source of a stimulus? Could you have a tactile display, composed of 'haptic pixels' that can convey information adequately? I read a good number of papers about this topic and, I thought it would be interesting to try it out even if my hardware resources and capabilities are very limited.
Two years ago, I played a bit with mobile phone vibrating motors. They are relatively cheap, small and friendly to use. The resulting vibration can be surprisingly strong. Two form factors are available on the market: disk-like shape and long cylinder ones. The ones I have on hand are the disk type.
After testing one disk hooked to an Arduino board,I decided to arrange five of them in a grip/block shape to check how well they could create a 5 tactile pixels array.
The result is very deceiving, at least concerning the resolution and identifying the source of the vibration. It is almost impossible to feel from which disk(s) the vibration is coming. The signal is not localized at all. You can feel the vibration across the full block. I guess the wooden block is very good at propagating the signal throughout the whole grip.
To be honest, I knew in advance that the vibration would be spread to some extend to the support block. But I didn't expected it to be this much. Nevertheless, I learn a lot building this quick model. The nature of the vibrating motor, a spinning offset weight, make them poor signal encoder. The rotating element takes some time (about 100 ms) to start moving, and has a clear falloff (150-300 ms) after the current is turned off. It means that the system is not capable of producing very sharp and precise vibrating signals.
I think I will try one iteration with this model eventually: split the block in sections and try to isolate them with foam or other damping material. I would be interesting also to try with other types of actuators, either with tiny cylindrical motors or with mechanical elements of some sort. I found an old article from the University of Toronto that suggests using a cheap relay as a tactile actuator. They supposely are very good at producing sharp 'tic' when the mechanical part switches inside. I'll look into it for sure.