Testing the big 15 and 20Ah LiFePO4 cells is tough! *Pics*

[liveforphysics]

Temps on the good (not damaged from abuse) cells stays much cooler. I gotta get some sleep, and I work another 12hr shift here, I'm going to get a 500amp video next. I'm working on getting a current overlay with voltage on the data logger so I can throw up some slick graphs under different test conditions (like temperature).

 

[etard]

So Luke,

what kind of packaging/box/layout would you suggest for a 12s setup? Would bus bars be ideal? Can the tabs be twisted to say a 45 degree angle? Or could soldered cables be used for us mortals?

Thanks in advance

Cell man,

Are you providing any sort of adapter plate or connection dongle that will make these easy to implement? Forgive me if this has already been covered.

 

[cell_man]

Hi Etard,

I'm currently working on a mechanical method for connecting the tabs together and giving a point for the balance cables to be connected to. If you look in the for sale section you should see all the info. I may also look at getting some PCBs made up that will allow the tabs to be soldered. This will be a cheaper solution and maybe a bit simpler too.... The mechanical method is a bit of work to be honest but at least with that it's pretty easy to DIY. If you're running a 1p setup there will be no current flowing through the terminals. You don't even need conducting terminals, as current just flows from tab to tab.

 

[madmadscientist]

Forgive what's probably a newbie type question...as you can see I don't post here a lot.
I've read all the threads on these big LiFePO4 batteries and I want to make sure I understand what people are saying.
These new A123 prismatic pouch batteries appear to be the bee's knee's? Is that correct?

Could someone take pity on this non-expert and explain what that means exactly?
One point i've gathered is that you get about the same amount of energy in half the weight and volume that you would using the A123 cells scavanged from deWalt battery packs?

thanks,

 

[madmadscientist]

Nother question I've got is are these A123 prismatic batts better than other prismatic batts?

The only other Li-Ion prismatic batt that I know of are the batteries made by EnerDel is that correct?

Does anybody have any sort of opinion on the EnerDel prismatic vs. these A123 prismatic cells?

thanks so much,

 

[CroDriver]

20Ah cell:
3.259v resting
2.970v loaded @ 352amps.
Internal resistance 0.821mOhm.
-Luke

So this fresh cells sags only 0,29V at 17C?! Wow.

Luke, what do you guess is a "safe" limit for these cells without damadging them?

I just decided to build in another huge motor and 2000 Amp controller in my car so now I need 3500 battery Amps. I'm not sure if 5 20 Ah cells in parallel would be enough. 3500 Amps would be 35C for a 5p battery pack

 

[RoughRider]

@Cro
take the high voltage DC motors...you will need less than half of the amps

the motors and controllers can take 370V...

 

[CroDriver]

@Cro
take the high voltage DC motors...you will need less than half of the amps

the motors and controllers can take 370V...

Well, that's what I'm doing

Two 330V motors + two 400V/2000Amp controllers

 

[Matthijs]

Maybe a bit offtopic but here you can see how the Dow/Kokam Ultra High Power modules are being build for the white zombie using 30 ah Dow/Kokam cells.

http://photos.plasmaboyracing.com/album07

 

[RoughRider]

@Cro
YOU ARE CRAZY...

lets say 300V*2000Amp * 2 = 1200kW...CRAZY

 

[CroDriver]

@Cro
YOU ARE CRAZY...

lets say 300V*2000Amp * 2 = 1200kW...CRAZY

Ahhmm, thanks...?

5 20 Ah cells and 6 15 Ah test cells will be on my desk in a few days so we'll have more data in a couple of weeks

 

[RoughRider]

i think it(2000Amp) should work with 5p of the 20Ah cells...

400Amp per cells...the cells should take this with a smile...

 

[cell_man]

The attached document was linked to previously. I found it very useful. The information is related to A123 26650s but I think it's a pretty good place to start.

View attachment A123 cell info.pdf

Excessive heat seems to be the biggest killer of cells so I would say it is more a case of keeping the continuous current pull within reasonable limits and manage the temperature. Lots of thermocouples are in order IMO. However, even if sufficient cooling were provided I think it would wise to still leave some margin as internal temperatures will of course be higher. Cro Driver I think you're really pushing the limits of the cells but with all that investment in the cells it would be wise to tread a little carefully as the difference between a very long and happy life or a rather short life for those poor cells, is just a few degrees C. High current peaks for relatively short durations should be no bother, so typical acceleration and braking will give the cells a bit of a break and avoid excessive heating IMO. It would be something like a top speed run where they are under very high load for an extended period where they may get too hot. I would say it is the average current per cell over say a 1-2min period that is most critical with regards to heating. So even if the peaks are very high, as long as they are not sustained for an excessive period you should be ok. I think you will also need to decide a safe peak current and decide upon a safe voltage that the cells will be allowed to sag to under load. If there is some kind of feedback in the system that will keep the load to within safe limits, ie limit the peak load to prevent excessive voltage sag and limit average load in the event of excessive heat, that would be an ideal scenario IMO.

I'd recommend lots of testing initially to see where your safe limit is and then when you install the pack tread a little carefully, increasing in relatively small steps and I'm sure you'll quickly get a feel for where the safe limit is. The good news from the attached doc is that even if cells are driven very hard and are overheated it does not cause instant issues, but is more of an accumulative effect, gradually damaging the cells and shortening their life cycle.

 

[liveforphysics]

@Cro
YOU ARE CRAZY...

lets say 300V*2000Amp * 2 = 1200kW...CRAZY

Ahhmm, thanks...?

5 20 Ah cells and 6 15 Ah test cells will be on my desk in a few days so we'll have more data in a couple of weeks

Still waiting on the fancy scope to come, and I will have graphs of 450amp discharge on the 15Ah cells, and 600amp discharge on the 20Ah cells. They are very much capable of it, and in a drag vehicle like you're proposing, he wouldn't be a concern for them at all.

 

[CroDriver]

I agree Paul, I will have to do lots of testing

I'm even thinking about two separate packs. One for the front motor/controller, one for the rear motor/controller.

One pack would be in a 4p/112s (20Ah cells)

If the cells sag to 2,7V at 25C, the pack voltage would be 302V - 300V will also be the max. controller output voltage (motor input voltage)

I think that this setup would be very good. I still have to test them to decide if 4p is enough or I have to add one or two cells to keep the voltage sag and temperature as low as possible

Btw. What's the exact weight of the cells?

 

[cell_man]

Hi Mate,

the exact weight of the cells is 466-467g for the 20Ah and 400-401g for 15Ah on the cells I've checked. However the 15Ah cells are generally around 16Ah so the energy density is closer than it first appears 141.33Whrs/kg for 20Ah versus 132Whrs/kg for 16Ah.

I think it's gonna be rather challenging running such a big pack of cells to their limits so it will take some time to iron out all the little issues you are bound to come up against.

I'll get those cells sent on Monday so you can get started on the testing a bit sooner :wink:

 

[Doctorbass]

Reading these Energy density calculation i think we might take in acount the connection from one cell to another..

Since these prismatic cells require asophisticated connection method that inclode more material than for the simple cylindrical shape cells, the Wh per kg is not well calculated..

We should add the connection structure weight in the calculations.

Why just calculating the cell weight itsef if for one form factor( ex 15 or 20Ah pouch) you need to add more material to make the connection than for the other cells ( ex 26650, 18650..A123 or the TS?

Cellman, if a 20Ah cell weight 467g but that you need to add 30g of connections to be able to link it to another, i would say that cell weight close to 500g ?

Doc

 

[CroDriver]

Reading these Energy density calculation i think we might take in acount the connection from one cell to another..

Since these prismatic cells require asophisticated connection method that inclode more material than for the simple cylindrical shape cells, the Wh per kg is not well calculated..

We should add the connection structure weight in the calculations.

Why just calculating the cell weight itsef if for one form factor( ex 15 or 20Ah pouch) you need to add more material to make the connection than for the other cells ( ex 26650, 18650..A123 or the TS?

Cellman, if a 20Ah cell weight 467g but that you need to add 30g of connections to be able to link it to another, i would say that cell weight close to 500g ?

Doc

I think that this wouldn't be fair.

It's not easy to make a pack out of Headways and especially of cylindrical A123 cells. Maybe you guys can glue a few of them together for a bike but that's no option for a car or a serious business.

I like this cell's shape better than the cylindrical type.

 

[olaf-lampe]

I agree with Doc, you have to consider also the weight of compression plates. If you want some propper cooling, you also have to put an oversized cooling plate between each cell to spread the heat evenly.
We had a project last year with Kokam cells in a very hot environment.( Indonesia ) We decided to add a watercooled bottomplate which was connected to each of the cell cooling plates. This added 30% of the original weight, but it was worth it.
-Olaf

 

[TylerDurden]

Brain-fart: A large packs might benefit from plates that can pass coolant... Corrugated aluminum panels, perhaps.

Certainly would add to pack size and weight; but the benefits might surpass the penalties.

 

[thomas]

Does these cell need to be compressed like ex. thundersky cells or is okay to just build a box that fit the cells but with maybe 0,5-1 mm of air at the end of a cell stack?

 

[oatnet]

I think that this wouldn't be fair.

It's not easy to make a pack out of Headways and especially of cylindrical A123 cells. Maybe you guys can glue a few of them together for a bike but that's no option for a car or a serious business.

Oh yeah, they'd never run a car like white zombie, or serious drag motorcycle like Killacycle on the cylindrical cells.

 

[CroDriver]

I think that this wouldn't be fair.

It's not easy to make a pack out of Headways and especially of cylindrical A123 cells. Maybe you guys can glue a few of them together for a bike but that's no option for a car or a serious business.

Oh yeah, they'd never run a car like white zombie, or serious drag motorcycle like Killacycle on the cylindrical cells.

They do but it's a PITA to build packs out of them (spot welding thousands of cells)

 

[AussieJester]

They do but it's a PITA to build packs out of them (spot welding thousands of cells)

Agreed...how many cells the Tesla have was it 2000 or 3000? LoL...must suck being on the 'battery detail' at Tesla factory hehe...

I dunno though Doc, not everyone will use the termination method Paul is selling, some will just buy the cells and connect them differently, perhaps a cell only energy density calculation AND one with the termination components added would make all happy?

KiM

 

[voicecoils]

They do but it's a PITA to build packs out of them (spot welding thousands of cells)

Agreed...how many cells the Tesla have was it 2000 or 3000? LoL...must suck being on the 'battery detail' at Tesla factory hehe...

The Tesla Roadster has just a hair under 7k cells.

It's a properly engineered, mass produced energy storage system. Not just a big box of cells tab welded together. I haven't heard any stories of failures either and more then 1,000 roadsters have been delivered.

For DIY purposes, all large format cells are much more appealing. Just look at Thundersky/Sky Energy's popularity, despite some drawbacks in power and energy density.