New A123 Prismatic 20Ah cells 480g and 30C

First, to clear up one point, I have used 15s a123 26650-based packs with 12s RC controllers for at least four years now, going back to my early electric Raptor RC helicopters. I think the absolute voltage that a Phoenix craps out at is around 53-54V.

As for the pack configurations, what I'd like to see is basically just a big "Turnigy"/RC-styled sub-pack, with 6 or 8 20Ah cells in series, shrink-wrapped together, with large main power leads (8 AWG?) with either big Andersons or 8mm bullets, and a balancer plug pigtail. For the latter, RC packs typically use JST-XH connectors 22-gauge wires for the balancer pigtails. For this big boy "RC"-style pack, I think at least 18-gauge wires and a suitable connector with pins that can handle 2A, or so.

Richard and I are testing a major new update/revision to our LiFePO4/LiPo BMS unit. I will post more about this in the day, or so, in the BMS thread, but basically, there will be multiple 6 or 8-channel sections that will be stackable, along with one more board with the charger control logic, which is all new. Here's what one current section looks like:

View attachment 6s LiPO BMS-v3.9a.png

Each channel can support shunt currents up to 2A, and those familiar with the existing BMS series will note this has fewer parts per channel. There will be active cooling via fans that will mount on the sides of the stack of boards.

Anyway, more about this later, but this new BMS is modular, and able to support much higher voltage setups (400V+...) from a single control board. The modularity will match up well with the "RC"-type sub-pack configurations I described above.

-- Gary
 
Outstanding Gary!
 
Gary,

Sorry but I haven't done so much reading up on your BMS design but judging by the the interest that it has developed from knowledgeable people on ES, I would have absolutely no qualms about using them with these packs. Maybe we can have a chat about getting some of these boards to try with these packs and if that is what people want, that can be an option and I'm happy to go along with their wishes. Alternatively the packs could just be assembled and prewired to accept 1 or as many is required of your BMS boards. We'll have to have a chat soon :)

Realistically cells could be leaving China in about 10 - 14 days. Packs would take a little longer but I'd do my best to pull out all the stops to get some assembled packs despatched not long after.

I'm happy to go along with 2 stacks of 6 or 8 pieces. If end to end that would result in a pack as follows:
39.6V 12S 15Ah - 40.8 * 150 * 422 - 4.8kg
52.8V 12S 15Ah - 42 * 166 * 454 - 5.76kg
39.6V 16S 20Ah - 54.4 * 150 * 422 - 6.4kg
52.8V 16S 20Ah - 56 * 166 * 454 - 7.68kg

There's still going to be a BMS required and some packing/case so it will work out a bit bigger and heavier, but it's still pretty impressive for such a safe chemistry with the ability to deliver over 5000W per kilo in the case of the 20Ah according to A123s specs.

liveforphysics said:
That chargery 1010B charger is my personal favorite, and I've got 4 of them. :)

The cell-log is also a very slick piece of equipment. Gary, myself, and others have found them to be an outstanding product for the minimal price they charge.
Click to expand...

Liveforphysics,

if you reckon the 1010B is the 1 to go for, it's good enough for me. There's no need to go for the 1010B+ I think?? I think 2 of them is enough for my requirements at the moment unless you reckon more would be useful. 2 off of the celllog 8 should be enough too unless you reckon otherwise. What sort of $s do they come in at and if you've got the contact details of anyone there that is good to deal with let me know ;) If they're as cheap as you say I might just chuck a few more in just in case :)
 
Hi Lynchy,

This might be a better choice (ask Luke) if you are doing 8s or less or multiples of 8s:
http://www.hobbycity.com/hobbycity/...ct_Name=iCharger_208B_350W_8s_Balance/Charger
Capable of handling up to 350W or 20A charge rates. Lipoly, LiFe and Lilo compatible as well as able to charge up to 8 cell packs.
The iCharger is famous for its ultra-accurate balancing feature. Each unit is calibrated at the factory prior to shipping!
This is a quality precision charger.
Click to expand...

If heat won't be a problem with stacked cells on an Ebike why did A123 build packs with heat sinks between the cells? Wouldn't an EV with a lot of cells in parallel draw less per cell?

There are rumors of the new model of Castle Controllers are going to be available in 16s and 20s lipo. If and when that happens the 15s limit for RC based systems will go away. Until that happens a convenient way to configure the sub-packs Gary suggested as 15s would be nice. Maybe offer 6s, 7s and 8s sub-packs?
 
MitchJi said:
Hi Lynchy,

This might be a better choice (ask Luke) if you are doing 8s or less or multiples of 8s:
http://www.hobbycity.com/hobbycity/...ct_Name=iCharger_208B_350W_8s_Balance/Charger
Capable of handling up to 350W or 20A charge rates. Lipoly, LiFe and Lilo compatible as well as able to charge up to 8 cell packs.
The iCharger is famous for its ultra-accurate balancing feature. Each unit is calibrated at the factory prior to shipping!
This is a quality precision charger.
Click to expand...

If heat won't be a problem with stacked cells on an Ebike why did A123 build packs with heat sinks between the cells? Wouldn't an EV with a lot of cells in parallel draw less per cell?

There are rumors of the new model of Castle Controllers are going to be available in 16s and 20s lipo. If and when that happens the 15s limit for RC based systems will go away. Until that happens a convenient way to configure the sub-packs Gary suggested as 15s would be nice. Maybe offer 6s, 7s and 8s sub-packs?
Click to expand...

I would go 1010b+ if they are in a similar price range and you want to charge in 10cell increments.

Mitch brings up a good point, if you're planning to be charing at 8s or smaller cell groups, going with the 208B would be a better choice. It also has a neat feature to connect it up to a 600w external loadbank and discharge graphs and cell testing with it.

Maybe 5s packs would work out best? Making lots of different pack sizes would increase costs.
 
MitchJi said:
If heat won't be a problem with stacked cells on an Ebike why did A123 build packs with heat sinks between the cells?
Click to expand...
It might be for adding heat as well as removing it. It got down to -9 F here the other day. :shock:
 
With LiFePO4, it really would be good to stick with multiples of 4 or 6 cells, because it will match up with a ton of existing chargers and CC/CV supplies much better, in my opinion. For charging purposes, 4 LiFePO4 cells = 6 SLA cells. The latter usually get charged to about 2.40-2.45V per cell, so for a 36V SLA setup, the charger will have its CV set point at around 43-44V. LiFePO4 cells normally like to get charged to 3.65V so a 12-cell pack would have the charger set to 43.8V. Same thing happens at the other common SLA configurations, like 12V, 24V, 48V and 72V. Again, for charging purposes, these match up perfectly with 4, 8, 16 and 24-cell LiFePO4 setups, respectively.

Anyway, what I'm thinking is if you had to pick just one "big boy" RC pack size for these 20Ah a123 cells, I'd go with 8s. That will cover the most bases, I think. People who are going to be interested in these are going to put them in higher performance setups, IMO, which I think will start at 48V. Two 8s sub-packs would do 48V, and three of them would work for 72V. For the real performance junkies, four would work in a monster 96V setup. That would take a 120V supply/charger, though, and I think fresh off the charger this voltage might be a bit too much for methy's 100V/100A controllers. There's lots of e-motorcycles that could use this configuration, with a 120V Kelly controller.

I also like 6-cells as a sub-pack size as well. You can still do a "standard" 24s/80V setup, using four 6s packs, and it lets you do a 30s/100V pack that should play nice with the 100V controllers, by using five 6s packs in series. Three of these would also probably work in most 48V setups, if the controller can take about 65V fresh off the charger.

Anyway, that's my two cents. :)

-- Gary
 
Gary,

Now that I see the artwork for the "new" model BMS... very nice but it raises a question or two:

1.) Since you will be wanting to do current throttling durring charging directly to teh power supply, how will that be incorporated?
2.) I don't see a common voltage reference output, so I assume when connected to multiple packs in series each channel is adjusted for max voltage and acts independently of the others so no need for adjacent balance boards to know about neighbor packs voltages / balance at a cell or pack level?

I guess I answered question #2 - sometimes I just think too quick (or act too quick depending on perspective) and in slowing down to type it up in "normal" speed, things become clearer. Alot of wasted typing really = )_

Please book me for 4 - 12 of the new units... I can build them here if you can supply them in kit form, even if you just want to sell the PCBs (to help get bootstrapped) then I'm happy to oblige. I have customer/associate who will be going between 15-24S (haven't decided yet, still engineering) either 5 or 6S sub packs and I have 2 of my own rides which I would like to use them on.

That said... do you have any plans of incorporating intelligence? An MCU to control the supply, interact with the BMS units (enable/disable balance), active cutout and monitoring of ah/wh with possible serial output or logging? Have any plans now that I ask? Wouldn't be too difficult, it's the part of the solution I've been working on besides the meanwells and "remote" controlling them and it wouldn't be too difficult to make this up and write the code.

Great job on the boards!

-Mike
 
Hi Gary,

GGoodrum said:
With LiFePO4, it really would be good to stick with multiples of 4 or 6 cells, because it will match up with a ton of existing chargers and CC/CV supplies much better, in my opinion.

-- Gary
Click to expand...
I completely agree, except that right now, for RC setups using the Castle esc's it would be really nice to have a convenient way to get to 15s.
 
mwkeefer said:
1.) Since you will be wanting to do current throttling durring charging directly to teh power supply, how will that be incorporated?
Click to expand...

The control board is not shown, because we are still testing a bunch of the new features, but it has an updated PWM-based throttling circuit which seems to work much better at controlling the cell voltages, without cutting the current as much as the old scheme did. I'm also testing having the throttling start at a bit higher voltage than the shunts and the charger/supply are set to, which also will reduce the amount of balancing/shunt time.

Other changes/new features include a new end-of-charge detection scheme, an auto power detection function to only enable the charger control logic when the charger is connected and on, and an adjustable failsafe timer that will shut the charge current off after a period of time after the shunts come on and the cells are balancing, should the end-of-charge signal not get set for some reason.

mwkeefer said:
2.) I don't see a common voltage reference output, so I assume when connected to multiple packs in series each channel is adjusted for max voltage and acts independently of the others so no need for adjacent balance boards to know about neighbor packs voltages / balance at a cell or pack level?
Click to expand...

This is a major difference between a BMS and the typical RC balancers. The latter balances cells to the level of the lowest cell, while a BMS like this one balances to a set voltage, which is 4.15V, in this case.

mwkeefer said:
That said... do you have any plans of incorporating intelligence? An MCU to control the supply, interact with the BMS units (enable/disable balance), active cutout and monitoring of ah/wh with possible serial output or logging? Have any plans now that I ask? Wouldn't be too difficult, it's the part of the solution I've been working on besides the meanwells and "remote" controlling them and it wouldn't be too difficult to make this up and write the code.
Click to expand...

No, not at all, and for a very good reason. What makes this work so well is its simplicity and non-digital reliability. I spent a career chasing timing, EMI and heat-related issues with microbased systems in jets, attack helicopters and special ops versions of C-130s and other aircraft. My background was software and then systems engineering. Sorry, but I would prefer not to retrace any of those paths. :roll: I'm a card-carrying believer in the KISS principle. :mrgreen:

In one application of this we are working on, for a motorcycle, they plan on having a fancy vehicle management system that will do cell-level reporting and data-logging, but they've had issues just even trying to do a simple USB A/D board to measure cell voltages. They decided they needed both, a reliable, but basic system that will protect the cells from over-discharge and over-charging, and also balance the cells, if needed, and then the fancy computer-based system that will do the data logging, etc., in parallel with the basic BMS. For ebikes, simple is still better, in my opinion, and if data logging is required, you can always run some of those CellLog widgets that HC sells, in parallel.

-- Gary
 
Gary,

You are beginning to sound like a fellow boeing employee / contractor?

I hear the KISS (man would you and my dad get along, maybe you can explain why the complexity of an RC system with chain and gears can be so much better than hubs for me?) and I don't disagree... guess I'm used to doing things in software and having hardware based failsafes for ... well safety.

BTW - the cell log8s can be connected multiple at a time to USB but you have to sever the VCC line on the USB bus to each one and let them power from the pack... the ground and signal wires for USB don't short this way and you can (I have you should be able to) cascade 3 of them together to monitor cell level voltages of up to 24s. Trying to connect them without severing the VCC from USB will result in fried / melted USB connectors (yep, I'm a dumb ass and had to try it) and / or a blown USB port / hub.

Can't wait to see the new units!!!

-Mike
 
mwkeefer said:
Gary,

You are beginning to sound like a fellow boeing employee / contractor?
Click to expand...

No, it was Lockheed. :)

mwkeefer said:
I hear the KISS (man would you and my dad get along, maybe you can explain why the complexity of an RC system with chain and gears can be so much better than hubs for me?) and I don't disagree... guess I'm used to doing things in software and having hardware based failsafes for ... well safety.
Click to expand...

Hubs are good as well :) The non-hub setups are good for lightweight, high performance, but they're not very stealthy.
 
Mmm!

Cells are in transit making there way from China to Luke's house. :)

I haven't been this excited about a battery getting delivered since buying my first LiPo pack, and I've never before been excited about LiFePO4 batteries getting delivered.

This battery type opens doors for the EV revolution. Never before has this sort of power, power density, and energy density been available in a safe battery chemistry, AND it has a crazy high claimed cycle life number!

The batteries might be excited to come to the USA, but they're going to wish they were somewhere else when I finish the break-in cycles, and start my testing on them. :shock: :twisted: :twisted: 8)
 
sounds like this got people's blood roiling,better than drugs!I'll be reading all about it even if I don't have a clue what all the charts and graphs mean once you guys dig in.Keep me busy until spring breakup.Foot of snow,these will be the only wheels I'll see turning till then.
 
liveforphysics said:
Cells are in transit making there way from China to Luke's house.
Click to expand...

Thanks for paving the way for the rest of us who are watching closely. If any of the product markings give an indication of build date, I'd like to know.

Cheers!
 
Hey Luke?

Whats your address?

Oh yea... that gruff looking guy with dark glasses sitting on an eBike in front of your house, just ignore him... it's a figment of your imagination (he he he) - guess that would make me the Grich who stole xmas eh?

Seriously though - can't wait!

-Mike
 
Hi,

Luke is correct on the China connection:
http://www.greencarcongress.com/200...battery-packs-in-china-atbs-to-be-p.html#more
A123 Systems and SAIC Motor Form Joint Venture for Hybrid and EV Battery Packs in China; ATBS to Be Preferential Supplier for All HEV and EVs Manufactured by SAIC and Subsidiaries
17 December 2009

A123 Systems is forming a joint venture with SAIC Motor Co. Ltd, a leading automaker in China. The focus of the joint venture is to develop, manufacture and sell complete vehicle traction battery systems for use in hybrid electric and pure electric passenger vehicles and heavy duty truck and bus applications in China.

The new venture, called Shanghai Advanced Traction Battery Systems Co. (ATBS), will seek to develop business throughout the entire Chinese transportation market and position A123 to strategically gain market share. ATBS will also be the preferential supplier of complete energy storage systems for all hybrid electric and pure electric vehicles manufactured by SAIC Motor and its wholly-owned subsidiaries.

A123 Systems will supply the advanced automotive battery cells and work with SAIC Motor to design and develop the integrated battery systems for ATBS.

The venture’s ownership will be held 51% by SAIC Motor and 49% by A123 Systems, with the management duties of the venture being shared equally between the parties.

In concert with this agreement, ATBS has been awarded a contract to supply battery systems for SAIC Motor’s plug-in hybrid vehicle program. This program seeks to develop a plug-in hybrid vehicle that meets the growing demand for alternative-energy transportation in China.

SAIC Motor is also developing a broad overall portfolio of “new energy” vehicles. SAIC Motor previously disclosed plans for a hybrid Roewe 750 sedan and a plug-in hybrid version of the Roewe 550, which could cut fuel consumption by 20% and 50% respectively, in addition to plans for introducing electric vehicles on the market in 2012. The Roewe 550 and 750 will utilize A123 battery cells.
Click to expand...
 
Hi all,
Wanted to kick off by saying that this is an amazing forum - lots of info, interesting topics, and positive and courteous people - great.

Have watched and read about the new a123 prismatics for a while and would now really want to get some firmer info on who I can source some from? I think I read about an RC store in spain that will be supplying them and that Lynchy and Liveforphysics will also be sourcing them - is that right?

Any help or advice will be much appreicated - Thanks
 
forestbootneck said:
Any help or advice will be much appreicated - Thanks
Click to expand...

http://www.endless-sphere.com/forums/viewtopic.php?f=9&t=14832
 
You must log in or register to reply here.

Similar threads

oatnet
NEW a123 20ah 30c cells, agnuism packmaker kits NEW
Replies
0
Views
542
oatnet
oatnet
N
Advice for Electro Streetfighter build - CBR900RR conversion
Replies
1
Views
908
SlowCo
SlowCo
LockH
Phylion Prismatic Cells
Replies
1
Views
609
Punx0r
P
T
A123 prismatic cells - THE solution?
2 3
Replies
65
Views
9,304
999zip999
999zip999
T
My new (finally!) assembled E-bike, with pictures!
Replies
9
Views
3,897
thunderstorm80
T
Back
Top