Heh. So, I just realized that I forgot to tag the last entry which would have put the previous Process post appear in the MS:I blog area. Dumb. No wonder Yury didn't see it.

I'm thinking about the Savonius Turbine in particular because of projects like
PicoTurbine.

At 300 windings per stator, the basic version of the PT should be able to produce between two and three volts, according to the authors. The plans have a pretty great discussion of the possible power coming out of the turbine.

In particular, I was looking at their high-power turbine variation, which had 6 fixed magnets, and their technical notes. This version should produce about a 1W of energy, they say, almost certainly through an
increase in voltage. With 8 stators and windings, I'd anticipate being able to create even more. I could then step down the voltage with a transformer or regulator, which would

• put the voltage in the typical 5V range
• increase the maximum possible amperage

If the PicoTurbine folks are right, then 1W/5V = 200 mA, which is a decent amount for most of the kinds of applications I'm thinking about.

Another possible application is the horizonal-axis wind turbine out of a Pringles can. This is a more familiar approach and the one that seems to have been adopted by large wind farms. There are a few problems with this, though, for what I want to do:

• HAWTs require some kind of vane to direct the turbine into the wind. This will:
  • Make it difficult to place on the side of walls or in any other position that would expose the horizontal axis to gravity
  • Add another axle to the design, increasing the complexity and cost, and lowering the reliability
• The large blades require more room in which to spin, meaning the application would almost certainly be limited to the rooftops and would only be able to stick to walls with a long armature extending from the side, which is structurally unreliable and difficult to manufacture cheaply and consistently.

So, sticking with the VAWT, most likely the Savonius, what kinds of applications can be run off of it? One example could be based on the GM862 Cellular Quad Band Module. This is a compact unit that performs all the functions of a cell phone, using a SIM card, without the screen, keypad, etc. According to the specifications, it needs between 3 and 4 volts to run, with about 17mA current requirement in standby more. Its operating current is up to 250mA, but this could be mitigated by using software that runs only periodically, remaining in low-power or standby more for most of the time, and by designed the hardware such that the turbine trickle charges a battery, leaving the cell module to draw max power of the battery itself and not the turbine.

For simpler LED applications, the power provided by the turbine should be more than sufficient for several LEDs. A typical red LED has a 2.4V drop and uses a max of 40mA. With a 5V source of 200mA, you could arrange a matrix of about 10 LEDs with five parallel lines of two LEDs in series. Minus a couple of LEDs and throw in a low-power IC, and you could have interesting timing circuits that blink different lights on or off. One or several switches could even be added. In fact, there's no particular reason why a Pleech could not be installed at a height near the top of a building, with a thin-gauge power line or conductive paint running down a building to enable an interactive piece at street level.