DIY Electric Car Forums banner

how to get a balancing for a 64S pack?

8K views 35 replies 7 participants last post by  nimblemotors 
#1 ·
If you build a 64S pack out of bare cells yourself, how do you get about balancing that?

Any commercially available chargers/balancers don't support anywhere near that amount, what solutions are commonly used for this by DIY battery builders here?
 
#3 ·
Yes, 64 cells in series, like in "sixty four".

Nominal voltage of one cell is 3.3V.

64 * 3.3V = 211.2V

At full discharge to 0, the voltage of one cell is 2.0V, so that would make 128.0V at full discharge.

What's nice with such setup is that you can use pretty much any worldwide (with wide 100V-250V AC range) switch-mode power supply directly with the battery and it will work both for a fully charged battery and all the way down up to fully discharged.

That's why I want to use such setup - to be able to plug in most standard power supply's of electronic devices directly to the battery.

And for a 22kW charge from a standard European three-phase 400V 32A socket, only 100A of current goes between the charger and the battery - which is easier to manage then 400A for a 48V battery for example.
 
#7 ·
I had made separate balancers for 40..100Ah cell with 1.7A shunting current (actually any amp value is available). They have been designed to fit WB (Thunder sky) cells. But I think they should fit other cells, too. I need to know the size of your cells to make useful advice.
I'm sorry for advertisement, if not allowed.
 
#8 ·
I need to know the size of your cells to make useful advice.
That will be only 3Ah (3 times 18650 1Ah parallel). Or maybe 4.

But I might be adding more cells in parallel as I gather more money, so the battery will grow, and the balancing circuit should be expandable too.

In parallel I don't need any balancing, they will self-balance, yes?
 
#11 ·
So you got a 64S 3P out of 18650 cells, correct?
I think (no experience) it could be a little tricky to wire them up to a balancing system and let every connection be in place for a long time ... I will follow your thread to see it :)
Good luck.

What kind of cells did you chosen? The NCR-18650A for example should be chargeed at 885mA, so three parallel would be 2,66A, not 4Amps ...
 
#12 ·
Yes, 64S3P. Cells are A123 (cheap B-grade) of course (why does anyone buy anything else?), and each can be charged at 5A, so three in parallel makes 15A.

I am seriously thinking that instead of buying an expensive charger and complicated balancer, to hook up 64 different DC-DC converters - one for each group of 3 parallel cells. Like the guy here does: http://electricporsche.rwaudio.com/2010/11/electric-porsche-charger-design/

Do you think that's a good idea in my situation?
 
#13 ·
The theoretical idea is great. I thought about something similar a while ago, but the practical issue is a little bit "ugly" to me :)

So many connections and electric components that could be fail ... and if you count every dollar (euro) you ivested, it could be easyer to get a charger of the shelf.
 
#14 ·
That would create a very balanced pack, its a great idea, BUT, you still need LVC. If any cell goes too low, you need to know about it.

If one "charger" fails, you wouldn't know if it was only the DC-DC converter route. So maybe some cell-logs and that would be a great solution.

(note: I have a similar setup waiting to be put together, 32 3.7V DC-DC converters with 48V input and a couple 48V AC-DC power supplies)

Still a lot of work for that many cells in series.
 
#19 ·
Miernik,

Mine are Vicor, custom model number and sold out. They were $10 each shipped. Found on ebay, seller has none left.

The MAIN reason people don't use A123, is they're expensive for a large pack with any sort of real range, and a REAL PAIN IN THE ASS to put a pack together. You cannot compress the ends, because it interferes with the pressure seal. You should NOT solder them, because it can lead to damage, so the only way you should put them together, is with a tab welder, which is an added expense.

Another HUGE reason, is that most of the "A123" B cells are unknown as to their authenticity. There are many conterfiet cells on the market, so its a crapshoot. Every chinese company is trying to make a buck.

They're also not very energy dense and can be heavy for a large pack, and have lots of connections. Its also very hard to replace a single cell if one goes bad.

Headway's are good for energy, decent for power and a hell of a lot easier to put together. I think rwaudio used them only for his 12V aux battery.
 
#20 ·
The MAIN reason people don't use A123, is they're expensive for a large pack with any sort of real range, and a REAL PAIN IN THE ASS to put a pack together. You cannot compress the ends, because it interferes with the pressure seal.
Wait, do you mean that this: http://www.tppacks.com/proddetail.asp?prod=Battery-Pack-Kit is not a good method to making an A123 pack?

I planned on doing something like that, will it damage the cell by pressure?

As for expense: if you divide the cell price not only by the capacity, but also by the number of cycles it will live, aren't A123s the cheapest cells there are? Especially B-grade 18650 1Ah for 2.05 USD a piece from A123RC.

I do plan to test a few thousand cycles on a few cells before buying a large amount of them, and I hope they'll still have about 50% capacity after 3000 cycles. If one 4C charge/discharge cycle takes half and hour, then after 1500 hours = 62.5 days, so after 2 months continuous test I will know what remains of them after 3000 cycles. Or did anyone do such a test already?
 
#22 ·
Nope, definately not a good method at all. Look up how an A123 cell is made. The ends are made to release if the cell goes pop. If you put any pressure on it, you can damage the cell. Kinda like opening a pop can, it tears a hole in the A123 cell. A123 specifically tells people NOT to assemble like that. Also, if you put pressure on it, it can't vent properly, so instead of blowing outward on the end, it will burst in the center, damaging other cells.

A123 are definately NOT the cheapest cells out there. And A123RC hasn't been proven to be selling legit cells, yet.


Compare cost for a 5kwh pack (in order of high to low power cells). These are just some examples of quick price-checks online.

A123:
1563 1Ah cells: 1563 * 2.05 = $3203.125
680 M1 2.3Ah cells: 680 * 76 per 10 cells = $5168
78 20Ah cells: 78 * 55 = $4290
plus tab welding, tab welder, nickel bussbar and that assumes they're legit

Headway:
156 10Ah cells: 156 * 18.50 = $2886
98 16Ah cells: 98 * 31.50 = $3087
plus: Buss bar and plastic holders

CALB:
16 100Ah cells: 16 * 137.50 = $2200 (about the same for all others)
plus bussbar

I sure as hell wouldn't want to put together a pack with anything more than a couple hundred cells.
 
#23 ·
Compare cost for a 5kwh pack (in order of high to low power cells). These are just some examples of quick price-checks online.

A123:
1563 1Ah cells: 1563 * 2.05 = $3203
plus tab welding, tab welder, nickel bussbar and that assumes they're legit

Headway:
156 10Ah cells: 156 * 18.50 = $2886
plus: Buss bar and plastic holders

CALB:
16 100Ah cells: 16 * 137.50 = $2200 (about the same for all others)
plus bussbar
Nice numbers, but now divide each one by the number of cycles of 4C charge + 4C discharge before the capacity drops below 50%, and what will be the result? Or even at 2C or 1C.

Maybe I buy a few of each brand and cycle them all for 2 months continuously. Or can you predict the result?

If Clabs would work out cheaper then A123 at such 4C cycle-life comparison test, I'd take them. But I doubt it.
 
#24 · (Edited)
4C charge, thats nice, but do this for me. 4C * 2.3Ah = 9.2A. Multiply that by the number of cells in parallel 9.2*3 = 27.6A. Now, multiply that times the volts (64 * 3.2V = 204.8) which is 204.8 * 27.6A = 5652.48W. Do you actually have a feed that large (47A @120V, 24A @240V)? Great if you do. Do you have a charger that big?

Discharge of 4C ain't no big thing. CALB are fine at 4C contuinuous. So are headway. Not sure what the CALB's do for peak, but headway will do 10C for over 30S, and will do 5C all day.

A123's drop quite a bit under a 10C load. I recently tested an A123 M1 at 5C, and it dropped to ~2.9V. At 20C (46A) it dropped to 2.25V, then as it heated up, it went to 2.5V... but it got to 150F. Headway 38120S cells at 5C dropped to 2.875V, never went above 127F.

The capacity of the cells was as follows:
A123 went for 640s, or 0.1777h at 11.5A, so capacity was roughly 2.044Ah (1C rating is 2.3Ah).
Headway went for 672s or 0.1866h at 50A, so capacity was roughly 9.33Ah (1C rating is 10Ah).

So, A123 is 0.255Ah lower at 5C, headway is 0.67. not a huge difference. I realize that's about 2.6 times the Ah rating loss in headway, but its not that far off. Not bad considering the cost and effort needed to build a pack. For the difference in cost, you could put a cell or two more cell in parallel and get the same results.

Those were just my preliminary tests with a couple cells, constant current and I monitored temp and voltage. Not as accurate as it is now, but its the data I've got.
 

Attachments

#25 ·
I'm with Travis on this one, unless you have equipment to properly weld the cells into packs it would be a massive effort with the 1ah or 2.3ah A123's, the 20ah pouches bring the cell count low enough that it is getting practical even if it's not easy. This is what drew me to the pouch cell.

I can see 4C peaks on discharge, but not continuous in anything practical. If you were racing the discharges would be higher, if it's your daily driver I would want a bit more than 15 minutes of driving time.

As for charging my headway 10ah cells are rated for 6c maximum, but I've only ever been able to do that one cell at a time on the bench. The pack I had planned for my Porsche would require a power generating station of my own. To expand on Travis's example with the exact pack I have planned for my Porsche.
768 10ah cells in a 96S8P configuration, lets use the 6C charge rate because that's what the data sheet says. I would need the following wall plug:
8 parallel cells x 10ah x 6c = 480 Amps
96 series cells x 3.5v = 336v (I could charge higher but I need to stay under the limit of the Soliton1)
So I need 161kw, given 90% charger efficiency that's 1475A @ 120V or 737A @ 240V

So unless you are building a pack for a pedal bike you are going to run into limits from the wall plug before you need to take advantage of high charging rates. My pack for example will be charging at 0.1C when I plug into a 240v line.

Perhaps if you explain more about the intended use and the type of conversion that you are doing we could give you more practical answers.

I don't want to discourage you from A123, but they are really only the best choice in certain applications.
 
#27 ·
At ~300grams each, you could fit about 32 of them into a 10kg package, 32Wh each cell, is about 1024kw. Lets Just say that's 102.4V nominal (3.2V each), so 1024/102.4=10A would charge this pack in 1 hour. The batteries can take 6C, which is 60A. If you did 60A and 102.4V, that'd be ~6kw. If you HAD a 6kw charger, you could charge in about 10 minutes.

BUT, you stated the limit is 828Wh from the wall, and charging at 102.4V, you're not limited by the batteries, you're limited by the wall. They'd get the same amount of energy from the wall as the A123 batteries.

Feel free to keep tossing questions my way. I test/sell/build battery packs on the side..... specializing in Lithium Ion.




Now, concerning the "module" idea....

You want to charge individual modules. You want to take them out of series (64s3p right). Are you just going to charge each one up as needed?

That will create a very unbalanced pack, especially if one "module" is charged, and another is not fully charged. Your pack is only as "full" as the "most empty" group.

(I have experience with modular pack design if you feel like discussing that as well)
 
#29 ·
You want to charge individual modules. You want to take them out of series (64s3p right). Are you just going to charge each one up as needed?

That will create a very unbalanced pack, especially if one "module" is charged, and another is not fully charged. Your pack is only as "full" as the "most empty" group.
That would be a problem only if the modules would be connected in series, but I'll connect them in parallel. When I'll connect a full module to an empty module in parallel, after a bit of time the charge will flow from the full battery to the full module to the empty module and both will become 50% full.

Each module will be 64s3p, if I connect three modules together that'll make 64s9p. Batteries connected in parallel self-balance, yes?

Concerning Headways, did you read this: http://www.zeva.com.au/tech/headway/
It suggests Headways don't perform too well.

Googling a bit I found many sources cite max charge of Headways at 2C, and some at 6C. I wonder why?

At 6C they would be better then A123s! And if it would be so, then everyone would be excited about Headways on forums and not about A123 as it is now.

And what about cycle life? B-grade A123s can be bought at 2.05$ per Ah shipped, and what is the cheapest price you can get Headways (shipped) now? In quantities of up to 100Ah at a time?
 
#30 ·
Whats this even going into anyway? 200+ volts and 9Ah? Is it going into a drag bicycle? Even at 20C with those batteries, that's only 180A and it'l sag so bad you'd be lucky to get 150VDC out of them under that load.

That would be a problem only if the modules would be connected in series, but I'll connect them in parallel. When I'll connect a full module to an empty module in parallel, after a bit of time the charge will flow from the full battery to the full module to the empty module and both will become 50% full.
That would really complicate the design. You have to buss the first battery of the first module to the first battery of the second module. Then buss the second to the second, and so on. Just putting the "pack" in parallel with the other "pack" will not self balance. The packs will rest at the same voltage, but the cells will be all over the place.

I'd kinda tend to agree with RWAudio on this one.


Each module will be 64s3p, if I connect three modules together that'll make 64s9p. Batteries connected in parallel self-balance, yes?
so you want to connect 192 togther and have 3 modules? good luck. Its going to be a large task. I'm assuming you're using the 18650's. I've welded packs of 120 cells a piece, maybe 6 packs, and that was a REAL pain. Realize, you have 2 connections. that's 1280 welds, and you usually have to have 2 welding points, so 2560 welds. Forget about soldering.
Concerning Headways, did you read this: http://www.zeva.com.au/tech/headway/
It suggests Headways don't perform too well.
You realize that test was done 3 years ago, right? Those cells have come a LONG way since then, and I've tested them. I mean, its ok for companies to continuously improve products isn't it? Well good news, they have.

Googling a bit I found many sources cite max charge of Headways at 2C, and some at 6C. I wonder why?
2C was 2 years ago. 6C is their current spec. Sometimes thats just how things are with China. The datasheets direct from headway state 6C max charge current on the newest cells. Some of the info is outdated because the "resellers" are lazy. I'll put that on my to do list of things to test.

At 6C they would be better then A123s! And if it would be so, then everyone would be excited about Headways on forums and not about A123 as it is now.
They're not BETTER for power, they're better for cost, ease of assembly, form factor, energy density and the ability to get legit cells with a WARRANTY.

As far as people on the forums being excited about A123, Maybe in RC and the bicycle forums, but from what I have witnessed in the last year or two around here are more interested in TS, CALB, GBS and Headway.

And what about cycle life? B-grade A123s can be bought at 2.05$ per Ah shipped, and what is the cheapest price you can get Headways (shipped) now? In quantities of up to 100Ah at a time?
As with all batteries, lifecycle depends on their use. If they're used no more than 20% DOD, you can get well over 2000 cycles.

As far as cost.... $2.05 an Ah shipped..... 100Ah is only 10 cells... so thats not really a very "large" quantity. Now, lets say you want the equivalent of 64S9p of the 1Ah cells, so about 64 cells of the 10Ah, thats something we stock, but its not a very large quantity of cells. Shipping wouldn't be that bad. I bet on 64 cells of 10Ah, they'd run about $18-19 + shipping, so lets say (worst case) $1216 + maybe $35 shipping + $25 hazmat = $1276. Now, compare that to buying the same thing 64s10p (equivalent pack), at 2.05. It works out to be just slightly more, $1312. Except you get no warranty, have to figure out how to weld 640 batteries, don't know if they're legit or not, and did I mention, its 640 frickin cells.


Headways are down below $1.80/Ah in quantity and TS and CALB are around $1.25/Ah in quantity. Shipping depends on how many you buy, cost goes down the more you order.
 
#32 ·
Whats this even going into anyway? 200+ volts and 9Ah? Is it going into a drag bicycle? Even at 20C with those batteries, that's only 180A and it'l sag so bad you'd be lucky to get 150VDC out of them under that load.
No, its a 2035 kg van.

Even 180A x 150V that is 27kW power and about 5kW, maybe max 10kW is enough to keep the van rolling at 30 km/h. I will only use batteries for city driving, and the speed limit is 30 km/h in most places here anyway. The diesel engine will stay, this is supposed to be a hybrid. The batteries are for short trips in the city.

Also the 9Ah is what I want to put at first, will allow me to travel several km at small city speed on batteries. I will increase the battery size once I get it all working and see that it all makes sense. It'll take me a year or two to convert anyway, so no sense buying more batteries earlier and just have them age.

But these 9Ah will also allow me to use a computer in the van so that will be an important purpose of the batteries even before I get any electric motor attached to it. I would be cool that this will start to get sold: http://www.poulsenhybrid.com/ so I could convert it to a hybrid quite easily then.

Anyway, I'll now be silent for some time, as I have to concentrate on work otherwise I'll never have money to buy more batteries.
 
#33 ·
Ok, now more details of the actual build come out..... Temporary pack, not a long range and .....It's a hybrid. And it sounds doable if you can get the mechanicals worked out.

You want to put a 1.8kwh pack of A123 cells, and draw 10kw continuous off it, so lets say ~5C (~45-50A) which would maybe drop to 2.9V and give you ~10kw. Thats doable, but from my testing I did, it'l only give about 10 minutes or so of runtime, if that.

The devil is in the details, it'd have helped if the details were more up front.

You've got lots to work out, but it sounds like a fun project, good luck!
 
#35 ·
well, good luck on your pack then. If they're so great then get em, but have fun assembling all of those cells, you'll need a tab welder.

I'm not going to argue the SOC, DOD and cycle life of A123 or Headway, mostly because the data isn't out there, and what info is out there, you can research it yourself.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top