View Single Post
Old 12-01-2012, 02:41 PM   #260
_cy_ OP
Beastly Adventurer
 
Joined: Sep 2011
Location: Tulsa, Oklahoma
Oddometer: 5,306
Quote:
Originally Posted by Twilight Error
Less than 5 years. I keep up on the tech through NASA Techbriefs, my IEEE membership and industry publications. The last batteries I built were for the X-37. They were power for the avionics and control surfaces during re-entry, it didn't burn up when it came home, so I must have done something right.

My problem with the commercial LiIon batteries is cell matching. Every cell, no matter how consistent the manufacturing process, is a little different.
A simple BMS system will shut off charging when the top cell reaches a full charge, leaving every other cell slightly lower, these systems are typically found in power tool batteries. This protects the battery against overcharge, and while LiFeX is more tolerant of overcharge than other variants, it still suffers an effect. Overcharging causes the Lithium to plate out on the anodes. The Li is then unavailable for movement across that membrane, and the capacity of that cell decreases. So the top cell loses capacity, but at the same time, the other cells are chronically undercharged.
A battery that lacks a BMS system will let the cells overcharge, our motorcycles do not have any ability to shunt excess power to ground when the battery reaches full charge. A Lead acid battery, esp. the modern VRLA and AGM units, tolerate it just fine and don't appear to suffer much in the way of dendrites and other uglyness.

If the battery starts with a set of tightly matched cells, it will delay the onset of the overcharge/undercharge cycle, but it won't prevent it. A good BMS system will use a DC-DC converter to regulate the voltage and current fed to the cells and a bleed resistor array to shunt power from the high cells while the low cells continue to charge. Charging will shut off when the low cell reaches full capacity. This is easy to do on a spacecraft designed to use a battery of this type, it is not so easy to retrofit a motorcycle with an 'always-on' charging system. Any BMS used with the charging system we've got on our bikes will need to shunt all excess power to heat, I don't see any of that in the current crop of LiIon batteries.
And none of this addresses the thermal management issues of these batteries. Heat kills LiIon. Cold kills LiIon. We got around the heat problem by attaching the battery to an actively cooled heatsink. Cold was overcome by attaching heaters to the case. All the batteries I built used prismatic cells, which are far easier to manage from a thermal perspective than cylindricals, which are the cells used in every LiIon battery sold for motorcycles. I've seen active cooling systems for cylindricals, but the water jackets are heavy and require their own hardware. They do have a benefit in being able to maintain the battery at a stable temperature, but again, I don't see this feature on the offerings for our market.
Some of The stuff I worked on 5 years ago still hasn't reached the civilian market, it may never get there.


ah... makes sense why you think all that... well dump all that out the window. Li-ion or lithium cobalt required an entirely different set of protections.

li-ion is an inherently unstable chemistry in that it will accept a charge long as it's delivered, until thermal runaway occurs about 4.35v or so. so special protections had to be built in to prevent this from happening.

main disadvantage of LiFePO4 is energy density is about 1/2 of Lithium cobalt. but so much more stable. most importantly 4x 3.0v-3.6v = 13v-13.6v falls within our motorcycle's 12v electrical system. vs lithium cobalt 3.7v-4.2v range doesn't match 12v systems without some type of voltage stepup or down.

now days BMS can be a full blown protection circuit that can isolate entire load. or the most commonly used scheme is to use four individual boards that protects each cell individually. when that particular cell reaches full charge cutoff at say 3.65v. rest of cells that have not reached cutoff will continue to charge until cutoff is reached. that that point excess current is shunted off.

note how each individual cell is protected by a independent BMS. picture shows a true 20 amp hour LiFePO4 battery. originally designed for electric scooters.

_cy_ is online now   Reply With Quote