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

[JB_EB]

Hi all, sorry for the long post, but I'm looking for some advice for a temporary charging solution for my pack while I wait for Gary to produce the next batch of kits of the new BMS. I have only recently got hold of some a123 cells and I don't want to kill any cells through mistreatment from lack of knowledge. I have also bought a new 27V - 30VDC, 5.6A Power Supply specifically for use with the new BMS, so I don't want to have to buy more stuff if I can help it.

My existing 24V, 10AH LiCo battery pack has only lasted 14 months. I think the peak current draw of about 20A from my Cyclone 370W motor has caused the internal resistance to increase to the point where the pack can't deliver more than about 8A and that is with big voltage drop.
I need a solution now, so I can continue to commute to work...

I have a commute to work of 6.5km each way and I can charge at work. The 10AH LiCo pack (when new) would give me a range (until BMS cutoff) of about 20km. So, each direction of my commute produces a DOD of about 30%. Therefore, I'm hoping I can just use an existing power supply or SLA charger to charge the a123 pack for my commute for a month or so without killing any cells.

The cells were extracted from "warranty return" De Walt packs. All the cells I kept had resting voltages of over 2.95V. Most were at 3.31V. I have individually charged cells to a common 3.31V and have made up a pack of 8s5p (~26V, 11.5AH). I have individual wires coming from each 5p block of cells. (Other posts from Gary et al indicate that cells that have not been discharged below about 2.5V should have full capacity.)

My plan is:
Charge pack up to 3.6 x 8 = 28.8V.
I can use either:
a. The new 5.6A power supply I bought to use with the BMS (27V but I can adjust up to 28.8 or 30V in the future). This supply will overcurrent limit to about 5.6A.
b. An existing 2A 24V SLA charger (CCCV).

During charge, monitor the pack voltage with a multimeter. Also, monitor individual 5p block voltages. I will also use a simple mains timer to turn off charger within a set time based on initial monitoring.

Every weekend, I can "manually" charge individual 5p blocks with a variable power supply (~3A) to keep them in balance if required.

Given the short commute, do you people think this will be a safe way to do things for a month or so until I get the full BMS / LVC kit?

Any advice greatly appreciated.

Thanks,

JohnB

 

[dnmun]

how did you get these warranty return a123 cells? there is a repair facility here in portland too. i would like to know what angle to use to get them to dump their batteries in my basket for recycling. did you sell them on some 'green' PR perspective or good references to the local city authorities for their environmental responsability, yada yada? do you work there?

i wonder if you already have some reversed cells in your parallel cells, what voltage is across the 5Pcells when you terminate your charge? does the resting voltage drop by the same amount for each 5P set?

 

[JB_EB]

dnmun - It was a pure fluke. I'd been having a browse of this forum and learnt a bit about all the work that the people here had been doing on De Walt cells. Then a chance meeting of a friend of a friend at a BBQ led to 5 warranty return packs landing in my lap. Of the 50 cells, I got 43 cells that had voltages greater than 2.85V. All my work todate, has been on the individual cells. I went through and measued each cell and rejected all cells with less than 2.5V. These cells had been sitting for a while in a box of "dead packs" so my assumption was that these cells were ok. I then manually charged up the lower voltage cells to a common 3.31V using a single variable power supply. I then made up 8s strings and then soldered cross-wires to connect the cells into 5p blocks. I haven't connected the pack to a load yet. I was waiting to build up the BMS. However, the delays in gettng the BMS mean I just have to do something now. Hence the request.

The cells I got, had been removed from the plastic case ready for recycling so they couldn't be resold, but this was fine for my application. If there is a repair facility in your city, why don't you try calling them and asking to speak to the person in charge of battery warranty returns. If you can get to the individual, you might be lucky and score some cells. Good luck.
JohnB

 

[GGoodrum]

JB, you should be fine with that supply. I'd set it to about 29.5V, or just about 3.7V per block of cells. ince you aren't drainig the cells down very much, I seriously doubt they will get too far apart. I still have a 16s5p a123 setup that is still going strong and I've never balanced this pack at all. Can't, as there's no taps to the cell junctions.

-- Gary

 

[dnmun]

jb, i should be so lucky. however, i will give it a try, never know if i get lucky, but i will take some cash in case some palms need greasing. just kidding, just barely. you should convince your friend to secretly distribute any more of those he has access to. you know you owe him, big time.
dennis

 

[ebinary]

jb, i should be so lucky. however, i will give it a try, never know if i get lucky, but i will take some cash in case some palms need greasing. just kidding, just barely. you should convince your friend to secretly distribute any more of those he has access to. you know you owe him, big time.
dennis

Portand is where I sent my near-new pack that had blinky lights on the charger (i'm in Bend). After a month, I'm still waiting for a replacement which it going to come directly from the factory. So it doesn't seem like they fix these things in Portland, because I'm sure one cell just needed a little boost to get the pack charging again.

Anyhoo, if you get an Oregon source, let me know - I want in!

Eric

 

[JB_EB]

Gary - Thanks for your reply.
I've used my existing LiCo 2A charger and am monitoring the pack voltage manually (visually??) with a multimeter while charging. After two round trips, with charging at work and home, I am seeing almost identical voltages each time which is a relief.
Straight off the charger I see 28.9V (~3.6V/cell). This drops to 28.6V overnight. After the ride to work (6.5km) the pack reads 26.6V (~3.3V/cell). After about 90 minutes charging, the charger LED turns green and the pack reads 28.9 to 29.0V. I see almost identical readings after the return trip.
All these readings look well within the specs.
On the weekend, I will open up the box and measure individual cell voltages before and after charging to check balance.
These a123 cells rock! Now I know what all the fuss is about.

I am now confident to use the a123 cells without a BMS for a short time and this gets me back on the road.
Thanks for all your help people.
Regards,
JohnB

 

[ohyea]

ok, im new here, and i have taken allot in reading 21 pgs of this thread in one sitting! Someone please tell me what the next runner up is to the a123 cells? i know they will take up more physical space, but what the 2nd best right now? this thread is a year old so something new has hit the market? thansk

 

[GGoodrum]

ok, im new here, and i have taken allot in reading 21 pgs of this thread in one sitting! Someone please tell me what the next runner up is to the a123 cells? i know they will take up more physical space, but what the 2nd best right now? this thread is a year old so something new has hit the market? thansk

In terms of performance, the a123s are probably still "king-of-the-hill". I would say the 10C PSI/LiFeBatt cells are next in line. It really depends on your application. If you need a lot of range, you can still get pretty good performance from packs using less expensive prismatic-based cells, like the popular Ping packs. That's because the amount of max current that can "comfortably" be pulled out of a pack is a function of the packs capacity times the "C" rating. For example, if you have a 25Ah pack that uses cells with a 2C rating, that means you can hit 50A peaks for a few seconds, without too many problems. With most stock controllers, you can't go too much more than about 35-40A, so this is fine. With higher "C"-rated cells, however, you can get much more power out of them, even with low capacities. A 10Ah PSI/LiFeBatt cell is rated at about 10C, so it is good for 100A bursts, without pulling the the voltage down so far that the LVC function starts tripping. The a123 cells are rated for about 30C, so even a "single-P" 2.3Ah configuration can put out about 70A. In a typical 4p/9.2Ah configuration, that goes up to a whopping 280A.

One characteristic I have seen is that it doesn't matter what a company rates their cells in terms of "burst" current, because in an ebike application, the cell's "continuous" current rating is about the max the cell can provide befor the voltage really starts to drop off. Better still is to not even push it that far.

Anyway, the advantage of the higher performance cells, like the a123s, is that you can pretty much tailor the pack capacity to whatever your range requirement is, which keeps the weight down. If you are looking to commute, perhaps, and need a bit of range, you can easily get by with some of the less expensive offerings. In the future, we will have more, and more options, but right now, not much has changed from last year. Ping is on his second generation, though, which is a good thing.

-- Gary

 

[ohyea]

ok, thanks, i was just wondering if the k2 or any of the cells out of other power tool packs are in the same ballpark or if they are a waste of time. it seems dewalt new packs are running 140/pack now? thanks

 

[johnrobholmes]

k2 cells are pretty good as well. I have some here. They do not offer BMS though, and I can't comment on the overall quality of the cells yet since I have only been using them for about a month. The price is good for the stated discharge rates and capacity. I bought four 16ah 12v packs, which have 4s5p construction each. I want to rip em apart and put balance taps on them.


I take care of my a123 packs with a lack of care. I make them in 6s chunks with a balance tap. I then wire them in 12s or 18s bricks for my bike. In the 12s configuration I use a 36v lead acid charger for simple plug and play charging. It puts them at 90% charged. Typically I will do this for a week and then check each pack with a balance charger at the end of the week. So far so good, all cells are staying within .02v of each other. I typically use 50 to 75% of pack capacity on a run.

 

[slayer]

Gay, Seeing how you are into bikes, what do you think about stretching a cut inner tube over your pack stick's body for more protection? You could even watertight seal this with that boot goo you light with a flame and drip on it?
Jeff K. "Deep Cycle" project

That's an interesting idea. I might give that a try. My favorite covering is the heavy rubberized shrink tubing that Batteryspace used to sell, but no longer does now. Here's what the last bits of this stuff that I have looks like:

I have searched high and low for another source for this stuff, which shrinks up so that it ends up being about 1/16" thick. I'll give a free a123 4s4p kit to the first person who can tell me where to buy some more of this.

-- Gary

Hi Gary look at what i found i don t know if you still need it but anyway i was thinking about you when i saw this on ebay
and i know you don t have any 4s4p kit available on your site so you don t really need to give me a free kit ...loll ... plus i dont think it is thick enough


http://cgi.ebay.com/24-Feet-of-5-Battery-Shrink-Wrap-A123-Lipo-NiCad-Nimh_W0QQitemZ330282948203QQcmdZViewItem?hash=item330282948203&_trkparms=72%3A1234%7C39%3A1%7C66%3A2%7C65%3A12%7C240%3A1318&_trksid=p3286.c0.m14

 

[GGoodrum]

That's good stuff, but it isn't wide enough for my needs. The rubberized stuff shown above was made by 3M, but it has been discontinued.

I'm working on a new way to build a123 packs, similar to the "stick" construction method, but not necessarily in a stick format. I'm using the same G10 plates, with nickel-plated springs for contact to the cell terminals. Each plate has four springs, with screws protruding up through the plate. Two of these plates are used with 4-cell clusters. The two plates and the four cells are bolted together, using a threaded spacer, and what you end up with is a block of four cells with "studs", for each contact. Here's what it looks like:

These 4-cell blocks can be oriented, and wired, in a variety of ways. For instance, I am building an 8s6p configured pack that I will use as "booster" pack for my Townie. The blocks are oriented in a 2 x 6 configuration, hot-glued together. I'm using copper strapping (available at Lowes and Home Depot...) to connect the cells. This pack will get an 8-channel version of the new BMS board

-- Gary

 

[slayer]

this is a great idea you have there...i actually broke my 48 volts a123 in 4 times 12 volts packs to use them in 4 12 volts parallel for winter, i bought a 1750 watts inverter to use with my electric snowblower and also a heater for the car or 4 series for biking in summer. but i use a 5 p configuration so i could not use your system.

 

[geoff57]

hi all
can anyone help me I have a ping lifepo pack which is great for range at a low price but there is no burst power in a ping pack thats not what they were designd for, when I went down to the clash of the titans in the uk jozzer sugested to me about how to solve my problem by making a hybrid pack by putting an a123cell in parallel with each block of parallel ping cells.
The help I am looking for is where to get the cells from I need 24 cells and I am in the UK. I have been seeing other dewalt battery voltages about with the nano name on are these the ones to go for.what sort of priceshould I be looking for?
building this new hybrid pack is a winter project so time is not short.
any help most appreciated and needed

Geoff

 

[ejonesss]

what do you mean by "no burst power "?

is it when you hit the throttle either a sluggish startup or the lvc kicks in?

hi all
can anyone help me I have a ping lifepo pack which is great for range at a low price but there is no burst power in a ping pack thats not what they were designd for, when I went down to the clash of the titans in the uk jozzer sugested to me about how to solve my problem by making a hybrid pack by putting an a123cell in parallel with each block of parallel ping cells.
The help I am looking for is where to get the cells from I need 24 cells and I am in the UK. I have been seeing other dewalt battery voltages about with the nano name on are these the ones to go for.what sort of priceshould I be looking for?
building this new hybrid pack is a winter project so time is not short.
any help most appreciated and needed

Geoff

 

[Battboy]

He probably means he's not getting at least 5C out of the cells ?

Don

 

[geoff57]

He probably means he's not getting at least 5C out of the cells ?

Don

don for all the times you put down ping batteries this time you hit the nail on the head.
I never believed pings claims of 5c I only try to run at max of 1.5c and at that it works fine my motor has changed the old motor a kollmorgon could never pull more than 30A the puma is a much bigger beast, once up to crusing speed ping cells would be able to handle the beast the A123s are going to be used to take up the battery killing draw of startup throttle guning and for me the few hills I have where I live jozzer has one of his KMX trikes with a 24s 1p A123 battery pack with a xtlyte 53xx in a 20 " wheel and at goodwood race track it was clocked a 30mph,and to quote another KMX from his signiture
KMX Trike, Puma hubmotor,24s1p A123 batteries. 3KW power. Total weight 21KG. Range 4 miles (can't have it all!)
I have a KMX like jozzer with the same puma what I am going to do is to get the range by using the ping cells I already have not as good as lifebatts or just A123 but a lot cheaper.
A hybrid battery pack would be a nice idea to try see what the results are.

 

[Battboy]

Hi Viper - If you mean by "Hybrid" - mixing different brands of cells be careful and expect balancing issues ? It's generally best to stick with all one brand of cells and preferably a brand that practices "cell matching" @ the factory floor level. Unless you want to balance every cell in your pack repeatedly, personally I wouldn't recommend a "Hybrid Mix". Likely you may end up with a "mutt" for a pack that will require more training than you would ever want to have to do.

Best,

Don Harmon

 

[GGoodrum]

Actually, I think this will work fine, but I would use a couple of Schottky diodes with each pack, and I would charge them separately. The diodes will keep one pack from draining down the other. Another way would maybe be to put the a123 pack in series with a "boost" switch, that you turn on only when you need the extra power, and then turn it off when the load subsides.

 

[geoff57]

Actually, I think this will work fine, but I would use a couple of Schottky diodes with each pack, and I would charge them separately. The diodes will keep one pack from draining down the other. Another way would maybe be to put the a123 pack in series with a "boost" switch, that you turn on only when you need the extra power, and then turn it off when the load subsides.

hi
I was going to add a single cell onto the end of each parallel block then charge as if it was two 36v battery packs are you saying to make up a 24s 1p pack of A123 cells and link them in with the 24s 4p ping pack I have (each ping cell is 5 ah) in parallel, with a switch to kick in only when needed say from a full stop, quick getaway at traffic lights and on hills. A bit like nitros oxide in road racing cars :wink:
would be an easyer build just haveto get a charger system for the A123pack.

 

[GGoodrum]

hi
I was going to add a single cell onto the end of each parallel block then charge as if it was two 36v battery packs are you saying to make up a 24s 1p pack of A123 cells and link them in with the 24s 4p ping pack I have (each ping cell is 5 ah) in parallel, with a switch to kick in only when needed say from a full stop, quick getaway at traffic lights and on hills. A bit like nitros oxide in road racing cars :wink:
would be an easyer build just haveto get a charger system for the A123pack.

For discharging, you could probably do what you are talking about, which is to simply add a single 2.3Ah a123 cell in parallel with the four Ping 5Ah cells, but for charging, I'm not sure how that would work. I'd have to think about it. I mean, there's no problem with charging with the same voltages, because the Ping cells and the a123 cells are still both basically LiFePO4-based chemistries, and I know by paralleling the cells, the Ping and a123 cells will all equalize to the same level, but I'm not sure what happens when you try charging them like that. It might be fine, even with the Ping BMS, but I don't know for sure.

The other way would be to keep the a123 cells separate, in their own serial string, and then connect them in parallel with the Ping pack, at the pack level, either using a pair of Schottky diodes, or with the "boost" switch. If you have two 36V/20Ah Ping packs in series, you could configure the a123s as two 12-cell sub-packs as well. You could use the same charger(s) you are using for the Ping packs, just not at the same time, maybe. The other issue with doing separate a123 packs is that you'd have to be careful not to let them get discharged to far, and you'd need to keep an eye on how well the cells stay balanced. A separate BMS would work, but I'm not sure it is wrth it for just a single string setup like this.

The more I think about it, the more I think I'd give adding an a123 cell to each block of paralelled 5Ah Ping cells a try. I think all that will happen is that it will add to the capacity of the block, but will also lower the combined internal resistance of the block. In effect, what you are doing is raising the "C" rating of each block of cells, to something that is somewhere in between the 1.5-2C of the Ping cells and the 30C+ rating of the a123 cells. With four 5Ah Ping cells in parallel, you can "safely" get about 20-30A peak draws before the voltage starts to really sag. With a single a123 cell, you can easily pull 60-70A peaks, with probably a lot less voltage sag. Putting the two together, I'm guessing will maybe pull the combined C-rating up to maybe 4-5C? Just and educated guess. Somebody smarter than me can probably calculate this, but I don't have anywhere near enough caffine in me yet this morning to even attempt it. Anyway, even if it were only about 3C, that would give you somethng like 60-65A peaks, which I'm guessing is more than enough. A BMC/Puma running at 72V will only need about 40-45A to hit 3kW peaks. I'm not sure anybody's had one that high without lots of "issues".

-- Gary

 

[geoff57]

hi Gary
the idea of adding an A123 cell to each paraellel block is the Idea that jozzer first proposed to me.
this is how it came about:-
I have 96 ping cells of the same type from a 48v 20ah pack and a 48v 10ah pack
I was going to join them with a schottky from justin which I bought and still have (40a size)
I then desided to change to a 72v setup and so bought a newbms from ping and rewired the 48v 10ah pack to make it a 24v 20ah pack.
the 2 packs are linked in series with diodes to protect the bmc boards.
the system worked fine until a connection in the wrong place fried the diode on the 24v pack when I next went to charge the 24v pack the charger fried as well, we live and learn.
at least the 48v pack was fine as are the cells so I am just using them.
So I thought it was time to do a compleat redesign of the battery pack, change to a pair of ping 36v BMS boards for charging and with jozzers help use a seperate LVC system that can handle a full 24s pack, jozzer has been able to help with that, at that time the amp cutoff was discused and we then started to talk about the idea of this hybrid battery pack, charged as 2 12s packs and discharged as a 24s pack.

now all I need Is the A123 cells to start the project going.
geoff

 

[Jozzer]

Just to say, I had 5p A123 cells paralleled (ie, each cell couple with the others in parallel, not just strings connected at the ends) with the 3p Lifebatt cells in the Aprilia for around 150 cycles and it worked very nicely. I noticed no difference in balance times when disconnected them and replaced them with more LifeBatt. It seems a reasonable aproach. If a cell fails, it will more likely be a ping cell, and if it takes just 1 of Geoff's a123's with it it will cost little to repair.
I believe if the A123's are individually paralleled with the ping cells the diodes become unnessacary, since the voltages of both strings will remain indentical and no current can flow anywhere it shouldn't..someone correct me if I wrong!

Hope this helps,
Steve

 

[GGoodrum]

I think you've got it right, Steve. I don't think diodes are necessary either. I think the only reason for having them when using the Ping BMS boards is if one board dies during discharge. It sounds like Geoff is only using the two 36V Ping BMS boards for charging, if you've set him up with a separate LVC board for discharging. Best of both worlds, I like it.

-- Gary