Big picture
From Mozda Converter
Contents |
Mozda Setup
Michael Mazgaonkar from Mozda Collective in Gujrat has superb renewable energy setup (wind+PV+batteries+mini grid) schematic of which can be found here. This is used to power lighting in schools and other economic uses. He is planning to add Pedal power generator to this setup based on Hugh Piggot's axial-flux-generator plans (originally meant for wind turbines but adapted to pedal power as described here.)
Need for converter
Average human can output 75W consistently on bike. Michael figured that at the current output voltage (48V) conduction losses in coils are about 20W, which is almost 30%! To reduce these, he plans to increase the output voltage to around 150V DC by having more turns per coil. To charge the battery at 12V, he needs a step down converter. It needs to have high efficiency (at least 80%), low cost(max 50$), and high voltage protection at input & output.
Previous design & its drawbacks
Michael had earlier worked in Spring 2006 with student group from UIUC to design a converter for lower voltages. The details of the project can be found here under title DC-DC converter for EWB wind turbine project. Unfortunately, despite satisfying all other tests for DC converter, above converter managed to get just 6-12% efficiency. Thus challenge before us lies in understanding root causes of low efficiency, and improving it. There is surely scope for doing so, as many other MOSFET based converters advertise 85-95% efficiency. We need to figure out how & at what cost.
Immediate task: Feasibility study
Michael is raising funds to buy state of the art but costly converters such as Outback MX60. This will give him the proof of concept about the loss reduction by increasing generator voltage. After that is done, only low cost converters such as the one we plan to build can replicate this efficiency gains in many villages. He needs us to evaluate VERY carefully the feasibility of such device under $50, and if we have capabilities & commitment to construct one. (this is because they ultimately need a working prototype, and till now his association with all student groups has not resulted in a working product ) This pre-analysis has to be done hopefully before Dec 15th. Hence the rush :-)
