T-Pain Is On A Boat: Re: [Electric Boats] 18650 cells for propulsion bank?

Wednesday, August 5, 2015

Re: [Electric Boats] 18650 cells for propulsion bank?

I bought 30 of those cells. I also searched out some reviews of them and determined that their actual capacity is around 1.5ah. So effectively 4x the price per cell, or about double the cost per watt of the Panasonic cells.

I am using my pack for dinghy propulsion with a 12v nominal pack (4s7p) connected to a 35lb thrust trolling motor.

I also bought battery cell organizers to make an orderly pack, cell tabbing for soldering together, a cell voltage monitor/alarm ($6), and a 4-cell battery regulator from deal extreme (7$) that can handle 10A continuous/ 30A peak. It will also cut the mosfets in the event of a cell going too low or high. All in, my 12V 6Ah (effective) pack cost me about $65 and lasts me for a couple of trips to/from the boat. But it also only weighs a couple of pounds and is easy to charge from the main propulsion bank through a dc/DC converter.

/Jason

On Aug 4, 2015, at 23:06, Robert Lemke robert-lemke@att.net [electricboats] <electricboats@yahoogroups.com> wrote:

I won't argue because that cell form is doing a great job in all the Tesla EVs out there. Where I will burst your bubble is any claim of a 6 ahr 18650 cell. Panasonic currently produces the best in this format, Tesla uses them, and the highest honest capacity is at 3.4 ahr at a cost of around $6.00 per cell.

Bob

On Tuesday, August 4, 2015 6:13 PM, king_of_neworleans <no_reply@yahoogroups.com> wrote:

Yeah I know, thats a lot of cells to solder together. But I was wondering if anyone had done this already. The reason I ask is because I have seen these Li-ion cells for as little as $1.25/ea, 6ah 3.7v. That's with free shipping from China. Now I know there are going to be a lot of substandard cells in a shipment of these things, but maybe the batting average would be good enough to make this cost effective (but not labor effective lol). But what would be needed would be a good routine for testing these things to weed out the duds and the weakies before soldering or spot-welding them together. It takes 13 in series to make 48v. So a series bundle of 13 paralleled with 49 more just like it would be a 300ah 48v bank, made up of lets see... 650 cells total. That's a lot of testing. So I am thinking some sort of gang/batch testing device and an efficient test routine. Maybe Arduino powered? I don't know. Like most of my projects I am starting out with near zero knowledge.

I am thinking about just getting like 26 of these cells and making a 48v, 12ah battery pack for the E-bike, just to get my hand in the game. That's $32.50 for cells and free shipping... just need some copper and a case and enough reserve cells to replace the inevitable bad cells, and I think I could be all in for well under $100, about 1/3 the cost of a typical e-bike battery.

Or maybe it would be simpler to make up parallel bundles of cells, then connect the sub-units together in a series of 13?

The only reason I am using golf cart batteries is cost. They are simply the cheapest way to go that will work. But 18650 cells would bring the cost down to not much more than the GC batts. Well, apart from a very "busy" BMS setup lol. But still, Im liking the numbers. To get a 220ah bank which is what I have now in flooded lead acid batts, I would need 37 parallel, 13 series, or 481 cells. At $1.25/ea that is $601.25. Lets say a 15% reject rate... still looks good. T he golf cart batts are $85/ea for 6v, 220ah. That's $680 plus tax and core deposit. Even if the 18650 setup cost 50% more, it would still be worth it for the increased lifespan and greater depth of discharge.

Somebody argue with me here, because I think I might be too enthusiastic about this idea.