Rod Hower via EV wrote: >>> You probably shouldn't use microcontrollers because they have >>> millions of transistors with the potential of failure...
David Roden wrote: >> I don't think it's quite the same. The microcontroller has lots >> of semiconductors, but they're all formed at one go on one >> substrate. OTOH, the lithium cells are individual units, >> manufactured individually, with individually welded connections. >> >> I too was skeptical about the Tesla ant colony battery construction >> -- which IIRC actually was used in earlier EVs with much less >> publicity and far lower production numbers. I believe the T-Zero >> Roadster was one of them. Paul Dove wrote > It’s called sarcasm! (smiles) so true... but it's hard to recognize sarcasm when folks aren't familiar with the actual situation. Putting lots of parts on the same chip means the reliability of each part is closely related to the rest. If one part is good, they're all good. If one transistor is weak, or one resistor has a resistance too low, they are ALL are weak or low resistance. And when one part fails, they all likely to fail. Same for batteries. Yes, a big cell is really a lot of small cells inside. It may have multiple plates wired in parallel, or one big plate folded or rolled into a cylinder (any piece of which would have been a fine cell in its own right). All these little cells were manufactured at the same time, and are "identical twins". Then they all get put in one big case, which seals the whole lot of them. This means they will all be kept together, at the same temperature, and experience the same charging and discharging regimen. Contrast that with individual cells. When they started mass-producing cheap 18650 lithium cells for laptops, many people independently came up with the idea of using thousands of them to build an EV pack. The initial attempts were failures, because there were too many differences between cells. Lots of failures and fires. Alan Cocconi is the first person I heard of that succeeded with them in his tZero. It required carefully matched cells, and a BMS to individually monitor them. The tZero inspired the Tesla Roadster, and led to their subsequent EVs. It only works if you get every single detail right. That's expensive. You can afford it for luxury cars where there is enough money to do it right. But I have serious doubts that it can be scaled to mass-produce cheap EVs. They'll get beat by the first company to figure out the best way to use far smaller numbers of much bigger cells. Complex solutions always come first. Simple solutions take longer to perfect; but usually win out in the end. -- Excellence does not require perfection. -- Henry James -- Lee A. Hart http://www.sunrise-ev.com _______________________________________________ UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub http://lists.evdl.org/listinfo.cgi/ev-evdl.org Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)
