Hacker News Reader

Comment by jatone | Hacker News Reader

depends on a lot of things, including how the battery charging is laid out in the battery itself. if you have 100 cells and can charge them simultaneously vs 100 cells where you have to charge sequentially you're going to see huge differences in time.

every piece of the system impacts this including the cable. if the cable is a bottle neck you need to fix that before any optimizations downstream can be unlocked.

View on HN

My car already runs its thousands of cells mostly in parallel, because the manufacturer isn’t stupid.

I am usually battery-limited while charging instead of cable limited. I plugged into a 250kW charger on Friday and my car peaked at 132kW because even with a preconditioned and mostly-empty battery it can still only take so much.

EDIT: I overstated things a little. I’m not usually battery-limited: I’m usually limited by the fact that my garage can only provide like 10kW of power. A better cable isn’t doing much there without major(ly expensive) upgrades to the wiring in the walls.

If electrical service enters through the garage, the additional wiring is probably minimal (though Cu prices are very high!), however:

(Calculations for common US garages)

NEC calls for 240V/200A service for an average-sized residence. Wire downsizing is allowed to 83% of service rating, so 166A.

So if my math is reasonable, in an average US single-family residence, you have max 39.84kW safely available for charging, with no other demand in the house.

(missing factor: what is the efficiency of AC-DC conversion in these chargers?)

Even at best, this is still only 30% of your battery limits, but might be more than 10kW for many customers.

5C is a number indicating how much current you can out into the cell. In this case, 5 times the nominal capacity or 12 minutes. This is on a cell basis, so no matter how you do it, you cannot charge faster unless you have cells with a chemistry that can handle more. In that case you'd already have a 10C or 20C battery if the cooling and electronics could handle it.

That is not correct. In order to fully charge a cell in 5 minutes, you need to charge that cell at 12C. That is just not possible with lithium batteries. 5C is pushing it. There is nothing to optimize, you would have to change the battery chemistry.

Everything else you mentioned is irrelevant, we are talking about all cells being charged at the same time which is really the only way that makes sense. There is no point in waiting to charge cells.

those are implementation details around the particular batteries. which is fine. nothing I said is absolving us of physical requirements. my point was entirely the system itself as a whole.

there are a number of ways to pluck the chicken. smaller more numerous cells -> same power, faster charging due to concurrency.

point was if your cable can't supply the power the entire point is irrelevant.

> smaller more numerous cells -> same power, faster charging due to concurrency

Not really, because smaller cells -> less capacity -> lower current for same C rating.

One 1000 mAh battery charging at 5C means 5A charging current, while two 500 mAh batteries in parallel leads to 2 * 2.5A = 5A.

And AFAIK typically a larger battery scales better, that is, less dead space for connectors/housing etc so more mAh per volume.

facepalm the assumption being made in my statement is that the individual cell can support the same charging rate.

my understanding was it was the chemistry of the cells that impacts the charging rate, not the physical size. happy to be wrong on that though.

scaling has mean dimensions. whats best for energy density may not necessarily be best for charging rates.

You don't seem to understand what "charging rate" means. There is nothing magical about subdividing lithium cells. No matter the size of the cell, you can't insert energy at a higher rate per unit of volume.

Reply on news.ycombinator.com