What you've built here is a rudimentary boost converter, using the motor inductance as your inductor, and your battery as your output capacitance.
The next step up here is to build a controllable device that can implement maximum power point tracking; a buck-boost power topology might be what you're looking for here.
Note that MPPT on a wind turbine is more difficult than, say solar, because the rotor speed is another state variable to consider. A naive/greedy algorithm might apply maximum torque to the rotor to extract maximum power for a moment, but this strategy will try to stall the rotor. As power is torque * angular velocity, both terms require consideration.
There are a lot of comments here saying that this isn't a SMPS, then proceed to describe a flyback converter (and the difference vs this unit) as a justification. Just note that the field of power electronics is quite large, and as an example, some power converters don't even have an inductor or transformer (though most do).
As I understood, the definition of a SMPS relates to switching between discrete states (i.e. on/off) as opposed to a linear supply, which dissipates power to regulate.
Source: am a power electronics engineer.
The suggestion to reuse objects, rather than reallocate temporaries (e.g. inside a loop body) was intriguing. Coming from C/C++ where stack allocations are approximately 'free', I tend to scope stack variables as narrowly as possible for readability and to help the compiler break dependencies. This is an interesting paradigm change that I wouldn't have expected from Go.
Expanding on above, Al-based BEOL interconnects were used until the process for making the more-difficult Cu interconnects matured around the turn of the millennium.
Without precautions, Cu will diffuse into Si badly and generate tons of traps. TiN is a typical barrier material, but increasingly exotic materials are now being used in cutting edge nodes.
A similar thing is happening with helium too. The concern over the impending closure of the US National Helium Reserve prompted people to start looking, and the results are already promising:
https://www.bloomberg.com/news/articles/2017-11-07/tanzania-...
I actually totally agree with you. Altium is the least-worst, but that's why I get so annoyed; I'm deliberately choosing to use it, but it's still such an unsatisfying experience.
Altium Designer. I've never had to put up with such an unstable buggy (and expensive) piece of software before. Just reading the changelog makes it apparent that they have no regression testing, and every feature seems to start out as a great idea, but get implemented with one critical function missing or unreliable.
It was an inspection failure from SpaceX's perspective. The vendor mismanufactured the part initially, but the failure rates of the parts were very low (to the order of several out of a batch of 10000 units), hence they weren't caught in batch testing. Also note that metal performance at cryogenic temperatures is often very different. Certain impurities that are largely harmless at room temperature can cause metals to be completely brittle at 90 kelvin. So testing every single strut would be very expensive, though SpaceX probably does that now regardless.