The care and feeding of a123-based packs...

[The Mighty Volt]

Oh, by the way.....one idea I did have was putting a 5 wire balance tap on each small sub-pack of batteries {the 5 in parallel} and charging them off a hobby charger. Is this feasible. It seems like a lot of work. Basically I would have 5 batteries per pack, 24 packs, all of them needing charging.

 

[amberwolf]

If you do it as 24 packs of 5 parallel cells, then to manage it you simply run a small wire from each pack's +, as well as one from the "bottom" pack's -, to your management system. If necessary you can use twisted shielded pair wire running from both + and - of each pack, if there is a lot of EMI or other noise on the lines for any reason (common in EVs with brushed motors, probably not so much with smaller brushless stuff). Just ground all the shields only at the management system end.

It saves you running pairs of large gauge wire to each 24s1p pack, as well as a whole lot (23-25 depending on how you do it) of balance wires to each 24s1p pack.

Of course, you could put a separate BMS on each of the 24s1p packs, which could enable "better" cell-level monitoring, but any single weak cell will stop your whole ride much earlier than if all are wired as 1s5p and then in series.

That way, you only have a single large gauge wire for system power to every 1s5p pack, plus at most a shielded twisted pair of small wires for monitoring, and one central BMS that tracks each 5p pack as if it were one much larger cell.

If you are logging data on the packs, and find that one single cell in one pack is causing a problem, you could disconnect and wire around (with a jumper cable, possibly premade to carry with you) the 1s5p pack and lose 3.2V on the whole system, and lose 32Wh. Might have to disconnect all the balance/monitor taps of the cells "above" the one removed, and move them all "down" one so you have no gaps, unless the BMS can be told to "ignore" that one.

If you are running 24s1p packs in parallel, you only have to unplug the offending 24s1p to get back on the road, without losing any voltage or having to rewire anything, but you lose over 150Wh that way.

There are advantages to either one, and disadvantages. I myself think that the advantages of 24x 1s5p would outweigh any disadvantages.

 

[amberwolf]

Oh, by the way.....one idea I did have was putting a 5 wire balance tap on each small sub-pack of batteries {the 5 in parallel} and charging them off a hobby charger. Is this feasible. It seems like a lot of work. Basically I would have 5 batteries per pack, 24 packs, all of them needing charging.

You don't need balance taps on the 1s5p packs. They are all in parallel as if they were one big cell, and if wired in series as a 24s is just charged exactly as if it was a single 24s5p large pack.

The balance taps are still just 23-25 wires (depending on how you do it), one per 1s5p pack +, and a main negative wire.

I'm pretty sure it's *less* work than the 24s1p packs.

 

[The Mighty Volt]

Thanks Amberwolf, appreciate it.

 

[Inkidu]

Great info thanks for contributing GGoodrum

I want to charge my 15 Ah prismatic cells (8s) and I am thinking about this charger.

http://www.meanwell.com/search/pb-360/default.htm

Will this work? cccv charger 28.8 volts / 8 = 3.6 volts float voltage 27.2 / 8 = 3.4

I am planning to make a purchase in the next few days because I only have a single cell charger.

A123 are known for high charge/discharge rates and wondering if this 28.8 @ 12 amps charger will work well?

My problem is that they mention this charger for lead acid batteries only. ???

This might only be to cover themselves. ????

With my limited experience I have been pleased with meanwell and I can purchase them with a warranty in the USA for a decent price.

$96 doesn't seem to be unreasonable for a 28.8v @ 12 amp charger. It is not PFc but I won't be charging more than an hour every week anyways.

The battery pack is for my Black & Decker 24 volt lawnmower.

I have used Meanwell products for my led projects and have been satisfied so I am thinking of going this route. (manufacturer)

Any help would be appreciated.

 

[GGoodrum]

Great info thanks for contributing GGoodrum

I want to charge my 15 Ah prismatic cells (8s) and I am thinking about this charger.

http://www.meanwell.com/search/pb-360/default.htm

Will this work? cccv charger 28.8 volts / 8 = 3.6 volts float voltage 27.2 / 8 = 3.4

I am planning to make a purchase in the next few days because I only have a single cell charger.

A123 are known for high charge/discharge rates and wondering if this 28.8 @ 12 amps charger will work well?

My problem is that they mention this charger for lead acid batteries only. ???

This might only be to cover themselves. ????

With my limited experience I have been pleased with meanwell and I can purchase them with a warranty in the USA for a decent price.

$96 doesn't seem to be unreasonable for a 28.8v @ 12 amp charger. It is not PFc but I won't be charging more than an hour every week anyways.

The battery pack is for my Black & Decker 24 volt lawnmower.

I have used Meanwell products for my led projects and have been satisfied so I am thinking of going this route. (manufacturer)

Any help would be appreciated.

Actually, I think this would be a great charger for your 8s pack. There is nothing wrong with charging to 3.60V per cell, as that will only add to the longevity of the cells. Although the charger will switch to the "float" mode, once the cells are full, it won't really do anything with these a123 cells, if they have the same sort of surface charge retention capability that the M1 versions have, as the voltage in the cells will stay close to 3.60V. Still, probably not a good idea to leave the charger connected, in this "float" mode. If you used a Charge Controller, and an 8-channel HVC board (I have these, but haven't finished the updates to my site to add them yet...), the power would be cut whenever the current in the CV mode dropped to below the adjustable set point, which could be set at about 100mA, for instance. The HVC/CC combo will also keep each cell from going over 3.60V as well, so in effect, individual cell CV modes.

The only thing that changes a MW supply into a MW charger is that the current limit is set to 100%, instead of 130%. What makes this an SLA charger is simply the secondary lower CV mode which is used to keep a float charge on the Pb cells.

-- Gary

 

[Inkidu]

Thanks for the reply GGoodrum

I have not had the cells long but they seem to be finding a float/rest voltage, when connected as a pack, of 3.35 volts

When I take the pack apart and charge separately some of the cell voltages seem to stay a little higher even after some time resting.

I am guessing from your comments that a usual charge and/or resting voltage might be higher that 3.6 volts. ????

Not sure what is happening I have limited experience with this type of stuff but that is why I purchased the cells i.e. gain experience

Learn by doing.

It is encouraging to see the cells wanted to "get together" and find a common ground.

I have not been using any type of balancing system.

I don''t want to start a big debate but what I have been reading about bottom balancing seems to make sense.

Given the abilities of these great cells and without a great need for amps, my pack might work without one.

I purchased a celllog 8s but need to find a small enough crimper for the molex connector and for the separate inline fuses for each cell.

I plan to pretty much baby sit the cells and see what happens.

Less complexity Less to go wrong. Besides being cheaper. I have over a $400 battery in a 13 year old push mower.

Have you ever run each cell down to say 2.8 and then connect together as a pack then charge as a pack?

You would end up with different peak voltages with each cell but the cell voltage would drop off together at the bottom

end instead of just one cell falling.

(bottom balance)

I probably would get more mowing time and power with having each cell at peak voltage but I am more concerned with having a pack for a long time.

I realize this isn't about bikes but feedback might carry over to ebikes.

Thanks for any help.

 

[Anonymous]

Dear Ggoodrum

for a 15s4p (49.5v 10ah) How do you connect the balancing leads? everybody seems to know except me

How do I 'charge' / discharge each battery to know which ones are bad and which ones aren't? (without resorting to a voltmeter on each battery after charging/discharging).

I have purchased 7 dewalt 36v packs (so 70 cells - 7-10 'bad' cells, we will assume I have around 62 good A123 cells ready for action).

With these 62 cells I need to build a 49.5v 9.2ah pack.

if I am building 15s4p for 280amp peak discharges, (later i'd like to build a 500amp peak, 300amp continuous!) 15 x 4 = 60


so therefore I am within reasonable battery state (of mind) to build this.

Would you advise me to build a 16s pack? Would your LVC board only work on 16s/channels? I am still struggling to find a suitable balancing charger solution. I am leaning more and more towards the 'individual lifepo4 chargers' from voltphreaks and a motherboard connector with 20connections.

Great info thanks for contributing GGoodrum

I want to charge my 15 Ah prismatic cells (8s) and I am thinking about this charger.

http://www.meanwell.com/search/pb-360/default.htm

Will this work? cccv charger 28.8 volts / 8 = 3.6 volts float voltage 27.2 / 8 = 3.4

I am planning to make a purchase in the next few days because I only have a single cell charger.

A123 are known for high charge/discharge rates and wondering if this 28.8 @ 12 amps charger will work well?

My problem is that they mention this charger for lead acid batteries only. ???

This might only be to cover themselves. ????

With my limited experience I have been pleased with meanwell and I can purchase them with a warranty in the USA for a decent price.

$96 doesn't seem to be unreasonable for a 28.8v @ 12 amp charger. It is not PFc but I won't be charging more than an hour every week anyways.

The battery pack is for my Black & Decker 24 volt lawnmower.

I have used Meanwell products for my led projects and have been satisfied so I am thinking of going this route. (manufacturer)

Any help would be appreciated.

Actually, I think this would be a great charger for your 8s pack. There is nothing wrong with charging to 3.60V per cell, as that will only add to the longevity of the cells. Although the charger will switch to the "float" mode, once the cells are full, it won't really do anything with these a123 cells, if they have the same sort of surface charge retention capability that the M1 versions have, as the voltage in the cells will stay close to 3.60V. Still, probably not a good idea to leave the charger connected, in this "float" mode. If you used a Charge Controller, and an 8-channel HVC board (I have these, but haven't finished the updates to my site to add them yet...), the power would be cut whenever the current in the CV mode dropped to below the adjustable set point, which could be set at about 100mA, for instance. The HVC/CC combo will also keep each cell from going over 3.60V as well, so in effect, individual cell CV modes.

The only thing that changes a MW supply into a MW charger is that the current limit is set to 100%, instead of 130%. What makes this an SLA charger is simply the secondary lower CV mode which is used to keep a float charge on the Pb cells.

-- Gary

 

[opalstear]

This Thread is useful.. ^^
Thank you~ .
I hope to make Battery pack, battery management system well ,too. >.<

 

[Inkidu]

Just a quick update for anyone that is interested on my particular experience.

I purchased the Meanwell PB-360-24 no PFC ($96) to charge my 8s 15Ah pouch cells for my push mower.(B&D)

http://www.meanwell.com/search/pb-360/default.htm

So far the charger is working fine. I turn down the charger to 28.0 volts because I really don't need all

the Ah that I have now. (It replaces a 17Ah lead acid battery)

My charge time is less than 25 minutes and all my cells only seam to be getting closer together in voltage. (no BMS)

I might of over done it a bit with this project but part of the reasoning, that I have come to believe, is that if you stay in the general

sweet spot of these type of cells there will never be a problem and you can have a battery pack that is light and lasts a very long time.

On the subject of longevity, I place a large Al heat sink under my charger when it is on. It has never got more than even a little warm

but all the same it now doesn't even get warm. I have found most electronics last longer when kept cool.

Hope this helps and thanks for any help. I am calling this project a success albeit an expensive one.

 

[pgt400]

I've been using the new version (v2.5) of the signalab BMS with A123 cells, 16S3P. The charger is a small cheap 48v 3amp agm battery scooter charger. Has been working great. V2.5 has individual smd leds for each channel. They come on at 3.65 volts all within 1 minute of each other. The charger is cut off when any cell hits 3.9 volts. I used an individual charger first to get them balanced. I have about a dozen cycles on them and am very happy with the BMS and cells. The BMS was only $80 built and tested.

 

[scoot]

...

Have you ever run each cell down to say 2.8 and then connect together as a pack then charge as a pack?
...

Actually, I have done just that...

http://endless-sphere.com/forums/viewtopic.php?f=14&t=15247&hilit=jack+pr&start=180#p237749

 

[The Mighty Volt]

Recently I came up with this idea for an A123 pack built into the frame of my MTB

I want to build a pack of in excess of 100+ volts, and I don't want to use a BMS.

Now, before you go off at me for not using one...please consider the merits of the following, if there are any, that is.

I am going to take 10 individual A123, weld them in parallel together with nickel tabs, solder on two 5 wire JST-XH taps, and finish off at the end with a negative and positive wire in the usual fashion. The negative/positive terminals will have Anderson Powerpoles.

I will repeat this process for every 10 cells that I have {I plan to use 300 A123 cells}

The individual packs are bundled up in heat-shrink, then connected in series, say 30 packs, connected in series, finishing with one "spare" positive and one "spare" negative, so to speak.

These two remaining terminals, one positive and one negative, constitute my end terminals. The pack is discharged through these terminals and recharged through these terminals.

When the pack is discharged, I would simply connect the negative/positive terminals to one of those big chargers I see touted around, and let the pack charge away, monitoring the pack voltage periodically.

Then, if I noted a problem within an individual 10 cell unit {basically two 5 cell units, with one balance tap per 5 cells, or two per 10 cell/1 unit} I could use an Accucel charger or a cell-log etc to help balance out that pack.

I.e. charge in bulk up to a certain point, then finish off with the Turnigy, making sure everything is just right.

Is this viable? Are there any noticeable flaws with the theory?

If the theory is sound, then what gauge wire should I use { I am expecting 50A to 45A} and what MAIN charger would I go with.

My chief reasons for not using a BMS are {a} I have no experience of wiring them {b} I will be doing hops and jumps and something might get busted {c} I have been told there is no hard and fast need for them and {d} if there is a problem, then the way my pack is set up, the most cells that could get hurt is 5.

Anyways, thanks for reading through all this.

 

[scoot]

The biggest concern I would have Mighty Volt is the manner in which the 5 or 10 paralleled cell clusters are serially connected to each other to achieve that modular quick replacement approach. Each connector based connection will result in some losses from increased resistance at each connector... that will really add up by the time have that many connections in series. I would use the biggest wires feasable (perhaps 8 or 10 awg with APP 75s) or maybe even multiple 12 awg wires with multiple anderson 45 PPs at each series connection. I would actually be one to use the APP 75s or SB 50s just to have larger/better connector contact surfaces... I swear by them. 8)

BTW, I use 15 amp color coded APPs on the end of my balance taps just to make access to single cells (or clusters) more convenient for measuring, charging and balancing... something to think about. It may prove convenient to manually manage (monitor/balance) your cell clusters in groups of 8s each... like maybe 4 or 5 seperate 8s packs to get your 100+ volts.

You can always update with a BMS when you grow weary of the manual cell mangement approach. I am not using a BMS yet, and it has forced me to learn and understand more comprehensively what is going on... having a versatile hobby charger, a CellLog8s and a CBAII has substantially assisted me in that endeaver.

 

[The Mighty Volt]

The biggest concern I would have Mighty Volt is the manner in which the 5 or 10 paralleled cell clusters are serially connected to each other to achieve that modular quick replacement approach. Each connector based connection will result in some losses from increased resistance at each connector... that will really add up by the time have that many connections in series. I would use the biggest wires feasable (perhaps 8 or 10 awg with APP 75s) or maybe even multiple 12 awg wires with multiple anderson 45 PPs at each series connection. I would actually be one to use the APP 75s or SB 50s just to have larger/better connector contact surfaces... I swear by them. 8)

BTW, I use 15 amp color coded APPs on the end of my balance taps just to make access to single cells (or clusters) more convenient for measuring, charging and balancing... something to think about. It may prove convenient to manually manage (monitor/balance) your cell clusters in groups of 8s each... like maybe 4 or 5 seperate 8s packs to get your 100+ volts.

You can always update with a BMS when you grow weary of the manual cell mangement approach. I am not using a BMS yet, and it has forced me to learn and understand more comprehensively what is going on... having a versatile hobby charger, a CellLog8s and a CBAII has substantially assisted me in that endeaver.

Thanks for that. I will take all of that on board.

I was aware of the resitance problem, better connectors and fatter wires is the best I can hope for in that regard.

Where would the fuse be placed? Between the battery and the controller, or between the controller and the hub? I suspect it is the controller and the battery, right?

I like the idea of a blend of main charger and then cell-log's and other smaller chargers coming in to do a finishing job.

I will be doing a lot of hops and jumps on this bike, space is at a premium, and vibrations and impact at a maximum, hence the desire to go around the need of a BMS

Cheers.

 

[oatnet]

Where would the fuse be placed? Between the battery and the controller, or between the controller and the hub? I suspect it is the controller and the battery, right?

Best place for the fuse is in the middle of the pack, or between two sub-packs. That cuts the maximum voltage in half if you drop a spanner on the pack or have some other manner of short.

-JD

 

[The Mighty Volt]

Where would the fuse be placed? Between the battery and the controller, or between the controller and the hub? I suspect it is the controller and the battery, right?

Best place for the fuse is in the middle of the pack, or between two sub-packs. That cuts the maximum voltage in half if you drop a spanner on the pack or have some other manner of short.

-JD

Thanks for that. At the high voltages {exceeding 100} shorting and accidents take on a whole new dimension.

 

[fechter]

Where would the fuse be placed? Between the battery and the controller, or between the controller and the hub? I suspect it is the controller and the battery, right?

Best place for the fuse is in the middle of the pack, or between two sub-packs. That cuts the maximum voltage in half if you drop a spanner on the pack or have some other manner of short.

-JD

Thanks for that. At the high voltages {exceeding 100} shorting and accidents take on a whole new dimension.

Absolutely!

Here's a good example of what could happen:
(the story of Plasma Boy)
http://www.evdl.org/pages/plasmaboy.html

 

[The Mighty Volt]

Hi Fechter, thanks for your post. Another of the reasons why I am trying to "avoid" a BMS is I want to build my pack to 100v+

Am I correct in saying that the biggest BMS out there only caters for 24s?

Thanks.

 

[fechter]

Not really. I've seen the Ver 2.6 used up to 40s cells with a little modification. I think the same will apply to the ver 4.x whenever it comes out.

If you don't want to use a full BMS, I would highly recommend at least using a HVC/LVC for every cell. We're working on a simple setup that will allow the use of multiple CellLog8 units for large packs. HVC should disable the charger and LVC should either alarm or reduce the throttle. If your cells are always kept in the happy range, neither LVC or HVC will be a common event.

 

[The Mighty Volt]

Not really. I've seen the Ver 2.6 used up to 40s cells with a little modification. I think the same will apply to the ver 4.x whenever it comes out.

If you don't want to use a full BMS, I would highly recommend at least using a HVC/LVC for every cell. We're working on a simple setup that will allow the use of multiple CellLog8 units for large packs. HVC should disable the charger and LVC should either alarm or reduce the throttle. If your cells are always kept in the happy range, neither LVC or HVC will be a common event.

Hi, and thanks again for your feedback

Simply put then, if I have 10 cells in parallel, in a "sub pack", what then is involved in wiring in these HVC/LVC devices?

Thanks.

 

[fechter]

You need to attach small wires to each series connection between sub packs and bring them out to the board. For example, if you have 16S,10P, you would need 17 wires coming out (main pack + and - and one between each series connection). It's handy to use some kind of connector on the tap wires.

 

[The Mighty Volt]

You need to attach small wires to each series connection between sub packs and bring them out to the board. For example, if you have 16S,10P, you would need 17 wires coming out (main pack + and - and one between each series connection). It's handy to use some kind of connector on the tap wires.

Okay then, and if I have 30s 10p, do I need to get two boards, and then link again in series somewhere down the line to maintain my overall voltage target?

Cheers.

 

[ramboman]

I like my iCharger 3010B...

 

[The Mighty Volt]

I like my iCharger 3010B...

I like my beautiful quarter-inch copper braid even more.

Any outstanding issues I have with conductivity should be ironed out [ha ha] once these babies go down onto the existing nickel strip which is currently in place!!