Big series / parallel pack

[jonescg]

Hi Guys,

This post could go somewhere else, but if it's here others might benefit from it.

I have been looking into building a big (~10 kWh) pack for my e-moto and I am wondering what is the best approach. My first instinct is to go for a large capacity LiFePO4 cell like the 20 Ah A123s and make a 120 S arrangement. Failing this, go for a lower capacity cell that has been paralleled up. Inspired by KiM's HOBO cruiser, I was left wondering if it wouldn't be cheaper and a lot easier to go for 18 readily available 6S LiPo bricks, and string them in series, and paralleling three of these strings together. Balancing would be achieved by paralleling each 7-wire lead from the three sets of cells, like the following:

I would then bulk-charge at 453 V with regular balance charges every so often. No BMS. Cause it's just not practical
This unfortunately is quite space inefficient compared to going for a big LiPo stack from individual cells, like the 5000 mAh 40C cells. What do you reckon?

 

[voicecoils]

It's ideal to parallel first and then make series connections to help the cells supply current as equally as possible.

Since it's not paractical to break the 6S1P packs apart and rewire as 1S6P, you can still make 6S3P groups first and then connect them in series. So, you'd have eighteen 22.2v24Ah packs then wired in series for the total 399v24Ah

 

[Jozzer]

The best way I've found for connecting these in large packs is with a 12mmx12mm copper bar, drilled with the required number of holes and populated with 4mm bullet connecters soldered in place. Male connectors at one end, female at the other. Use them to parallel and series the blocks. 2 bars made up for the ends with only male OR female for termination and connection to power wiring.
All packs should be paralleled first indeed.
Balance taps can all plug into a central board, so each cell ends up being paralleled at a cell level, and a pack level before being connected in series.
I make and sell these if you need..

 

[dogman dan]

Well, the first instinct is kinda correct. It just depends on what your amp requirements and budget are. 120 cells has got to be easier to manage than 360.

Not that the mega pack of lipo won't work, just that it could come to be a pita at some point, if you start having to weed out weak packs every month or so later when the pack is older.

My instinct would be to build multiple packs of a more digestible size. Say multiple 24s 15 ah packs. Build enough to have one spare. Then using some method or other for series connecting them into the main 120s pack. Then you can periodicly rotate one pack out of the system for a good scrutiny, weed out the stinker pack from it, and then replace into the main pack.

As the others said, build 15 ah 6s packs, then series connect. A bussbar setup to paralell the sets of three makes sense for later, because you will at some point want to remove and chuck individual packs, not permanently paralelled 15 ah blocks. The bussbar setup seems to me to minimize the amount of wire the current needs to pass through, and allow some really beefy connections for the series between each set of three lipos.

 

[jonescg]

Yeah, I know that paralleling first is the best way to go, I just thought that as long as the balance wires were paralleled it would work just as effectively. However, should one cell fail, the balance wire will take a very high current, so yeah, scratch that one

While I would much rather go for individual cells, the cheapest way to buy LiPo is in these 6S packs. Trouble is, it's such a pain in the arse to wire them all up :x

If I knew how big the individual 5 Ah cells were I could build a pack around them. Indeed, if they made individual 8 Ah cells it would be even better.

But Dogman, you seem to imply these guys will have a high failure rate and need regular weeding out? I'm not so keen on LiPo now...

 

[AussieJester]

CHRiS, there is not a high failure rate of the better quality packs
Dogman uses the lower c rate packs, stick with the nano tech lipo
and you will be good too go for a loooooong time.. Voltage sag will be a thing of the past for you to lol

KiM

 

[jonescg]

(Edited for clarity)

My preference is to go for a pack made from individual 40C 5 Ah Turnigy cells;
http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=18562
unless they have come out with an 8 Ah version?
Are these cells using the same NanoTech formula?

Otherwise the 30C, 6S 8000 mAh guys are from Zippy;
http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=19531
And might not be as hot?

Finally, building a 9.6 kWh pack from 25-50C 6S Nano-Tech cells would mean 90 of them all up, and cost about $8000
http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=11937

 

[AussieJester]

Its been said the lower c rate packs consist of cells that didn't make the grade for the higher
c rate packs CHRiS, thus it is suggested that you buy the highest c rate your pocket will allow.
Anything under 30c your taking a punt...imo. If you can swing the mullafor the nanos, get em.

KiM

 

[jonescg]

OK So this is what a 5p108s (540 cells all up, or 90 x 6S packs) Nano-tech pack would look like:

I haven't drawn the balance wires yes as that looks like a royal pain in the arse. Drawing them is bad enough, wiring them all! Urgh. This pack would store about 10 kWh and weigh in at 70 kg. By my calculations it would stand about 170 mm deep, 260 mm wide and 900 mm high (probably a bit too high, might split it down with an L shape). $8000 worth of cells though. Meh, I've spent that on Agni motors in the last 2 years :lol: ...

 

[AussieJester]

Have you thought about the Lipo that 'EVdragRacer' has been flogging? i'm not suggesting buying from him
but the same are available via other suppliers on flebay...here's the link anywayz CHRiS

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=31687&hilit=evdragracer

I have a beef with the guy for taking credit for others work, have a beef with anyoine that does this actually...
he could be fine to deal with, just a personal opinion of mine, i don't like the guy.... as you were :lol:



KiM

 

[jonescg]

Nah - it would end up costing over $13,000. Even Mavizen's A123s are cheaper, and offer the same energy density!

 

[neptronix]

Its been said the lower c rate packs consist of cells that didn't make the grade for the higher
c rate packs CHRiS, thus it is suggested that you buy the highest c rate your pocket will allow.
Anything under 30c your taking a punt...imo. If you can swing the mullafor the nanos, get em.

KiM

Sorry but that's just rumor, AJ. 25, 30, 35, and 40C+ cells are incrementally thicker and more heavy. Go compare the weight of the packs on hobbycity/hobbyking and you will see.. it's not like they're rejects that get downrated or whatever.

Those 20C cells are fine if used in a quantity that's high enough to provide the desired output. Plus you get more capacity..

No need for nanotech or higher unless you are drag racing and need that extra few pounds off the bike

 

[auraslip]

As far as balance wires getting a large current from a failed cell:

The weak cell would only be slightly less than the parallel cell because it's lower voltage means it would be hit for less power. The whole parallel group would go down.

You can do the math, but current through the balance tap is the the difference in voltage * (Ri + wiring and connecter resistance) It needs to be a LARGE imbalance to make the wires smoke.

You are going to test the pack for weak cells first right? If I was going to take the time and money to put together such a huge pack, I'd probably run each pack through a couple cycles.


That being said, for the price of the pack, you could run 5 cell logs for like 1/500 of the total pack price. Totally worth it IMHO. That way if a cell does go bad it's easy to find.

I'd be more concerned with balancing and wiring harness mess. I'd make parallel boards out of strip boards and headers, and then run the parallel leads to somewhere accessible for the cell logs, and balancing.

Really for a pack this big, I might just make an arduino interface and an LCD to keep track of each individual cell group.
Maybe I'm overthinking the whole balance issue, but with a pack that big.... I wouldn't want to spend a saturday every three months disassembling to balance it.

 

[dogman dan]

I dunno about the failure rate for others, mine has been suprisingly high a year later. New, the packs were quite good, and any bad ones showed themselves right away. But mine are starting to have at least one in 8 packs that have one cell that won't balance so hot anymore. Always about .05 to .08 low. Single cell charge, and they'll drop right back to were they were in an hour or so.

I never tlc'd mine. I kept em charged in storage, so I could hop on and ride anytime. I don't have hundreds of cycles on em, but some were harsher use in races or racing practice. I charge to 4.2v.

What I meant to say was, at some point as the packs get older they would start having more and more of them that don't like to balance so good anymore. I had no problem with treating a few hundred bucks in lipo harshly, but your more expensive pack would be worth a lot more care in dod, and charging to less than 4.2v.

And the higher the c rate the better the quality of the individual cells. At the time I bought mine, I thought I was buying the good stuff, 30-40 c rate turnigy. Later, I bought more that was 20c zippy. I'd love to make the next lipo buy 90c nano, but it's $$$.

All I'm saying, is sooner or later you will have to weed the pack. So building it so that can be done a bit easier makes sense to me.

 

[auraslip]

Good call. "Build with future serviceability in mind" Otherwise pay twice as much for the good stuff...

 

[Doctorbass]

Its been said the lower c rate packs consist of cells that didn't make the grade for the higher
c rate packs CHRiS, thus it is suggested that you buy the highest c rate your pocket will allow.
Anything under 30c your taking a punt...imo. If you can swing the mullafor the nanos, get em.

KiM

Sorry but that's just rumor, AJ. 25, 30, 35, and 40C+ cells are incrementally thicker and more heavy. Go compare the weight of the packs on hobbycity/hobbyking and you will see.. it's not like they're rejects that get downrated or whatever.

Those 20C cells are fine if used in a quantity that's high enough to provide the desired output. Plus you get more capacity..

No need for nanotech or higher unless you are drag racing and need that extra few pounds off the bike

+1

 

[Doctorbass]

As far as balance wires getting a large current from a failed cell:

The weak cell would only be slightly less than the parallel cell because it's lower voltage means it would be hit for less power. The whole parallel group would go down.

You can do the math, but current through the balance tap is the the difference in voltage * (Ri + wiring and connecter resistance) It needs to be a LARGE imbalance to make the wires smoke.

You are going to test the pack for weak cells first right? If I was going to take the time and money to put together such a huge pack, I'd probably run each pack through a couple cycles.


That being said, for the price of the pack, you could run 5 cell logs for like 1/500 of the total pack price. Totally worth it IMHO. That way if a cell does go bad it's easy to find.

I'd be more concerned with balancing and wiring harness mess. I'd make parallel boards out of strip boards and headers, and then run the parallel leads to somewhere accessible for the cell logs, and balancing.



Really for a pack this big, I might just make an arduino interface and an LCD to keep track of each individual cell group.
Maybe I'm overthinking the whole balance issue, but with a pack that big.... I wouldn't want to spend a saturday every three months disassembling to balance it.

+1, It might happen in just 6s group... also the balance wire are AWG24 or 26.. .. it's not enough to take the current of an unbalanced pack in any paralleled 6s..

Doc