Homemade Battery Packs

[miniebiker]

Great info in this thread.

There is info about using these batteries to power electric bicycles and testing results in the book Electric Bicycles A Guide to Design and Use by William C. Morchin and Henry Oman starting on page 64.

This thread puts it all together !

 

[Samd]

Been following this excellent thread. I have been looking to home brew some cells not so much for high current supply but to supplement my two turnigy 8s 5.8Ah.

Yesterday I made up as 10Ah of 8s from old nokia mobile batteries I have had for 3 years. Photos below. Most of them still had well over 3.9 volts after sitting in a box in the shed for 4 years. I sent out for recycling any swollen ones.
Soldering onto these actually isnt so hard, I didnt lose any. I did have a lot of sand and concrete handy.

Seems to cope ok on their own with a burst of 30 amps into my aprilia electric. No temp change. I have a small digital thermometer from ebay for 2 bucks I will mount in the pack. More testing needed on long term current for these little lipos....

 

[DrkAngel]

Great info in this thread. ...
... This thread puts it all together !

Keep posting - your successes ...
and failures ... we'd rather learn from your mistakes ... than our own!

 

[Samd]

Not sure I have a failure or not with the above soldering job - tried charging fully but anything over 0.3 amps from my turnigy 8s charger and the pack goes open circuit. Suspect my bullet connectors between the cells may have been the problem. Will do some checking on the weekend and advise.....

 

[miniebiker]

Great info in this thread. ...
... This thread puts it all together !

Keep posting - your successes ...
and failures ... we'd rather learn from your mistakes ... than our own!

I bought a lot of batteries on eBay that would fit my laptop so they would be easy to test. Everyone charged up ok on the laptop.

Took them apart, no cuts but did get a blister.

I numbered 120 cells and am doing charge discharge tests with an Imax 6 on each cell and writing down the results.

Doing this while I work on bicycles. When the alarm goes off I write down the results and do another battery.

I want to end up with 10 3.7v packs. Each 3.7v pack with the same ah capacity.

 

[miniebiker]

I tried charging the packs in my HP nc6230 laptop before I took them apart and they all charged but some got hot.

Took them apart and got 50 good ones out of 120. All the Samsung pink ones were good.

All the Sony green ones were bad.

The ones that tested bad would be warm when charging and would do poorly on the discharge test.

Interesting that the computer did not see them as bad and still tried charging them.

 

[DrkAngel]

"Craftman" 12v cordless drill, Li-ion battery pack upgrade, looks good.
10 years ago this was my favorite drill.
6 cells, 3S2P fits easily, and should supply enough output.

14.8v Firestorm pack, is doing nicely!
I plan on converting, at least, 2 more packs.

 

[Floont]

...All the Sony green ones were bad...

Big surprise.

 

[miniebiker]

...All the Sony green ones were bad...

Big surprise.

I guess the suprise is that my laptops still tried to charge them. The whole bottom of the laptops got hot !

The bms on these has no way to tell the temp of the batteries ?

And how bad do these batteries have to get before the computer stops trying to charge them ?

 

[Spicerack]

The battery packs have temp sensors on them. I've also seen thermal fuses with 90 celcius ratings on the packs.... Seems pretty darn HOT to me!

 

[miniebiker]

The battery packs have temp sensors on them. I've also seen thermal fuses with 90 celcius ratings on the packs.... Seems pretty darn HOT to me!

Me too !

The Imax B6 came set at 80 degrees C and I put it on 40 degrees C cutoff for the tests.

None got that hot so far charging at 2.5 amps, even the Sonys that failed the discharge tests.

I am using a Fluke 62 IR meter also to check temps.

This is fun !

 

[18wheeler]

Finally I am testing all my battery's capacity.

they spread from ~80min - ~150min when discharging with 5 ohm resistor.

how comparable should it be in term of 'minutes'? within +/- 1' ? or 5' or 10' ?

Eagerly awaiting updates!

Any tips/hints on how to setup a capacity measurement?

Reliable method - does not require constant monitoring.
Just swap cells, mark time expired, and reset timer, when alarm sounds.

1st ... charge a batch of cells to 4.2v.
2nd ... let set 24hours, check voltage and remove any with significant drop.
3rd ... find 4v compatible device with consistent draw. (6v headlight, or 2?)
Finally ... attach "Voltage tester-Low voltage alarm"($6.29), to cell, set alarm to 3.7v, maybe 3.8v.
Set timer-stopwatch etc., engage power drain, when alarm goes off, mark elapsed time on cell.

Repeat.

Should give an accurate, "comparative" capacity.
Packs should be built from cells, with same "comparative" capacity.

 

[miniebiker]

18650 datasheets

http://www.batteryonestop.com/baotongusa/products/datasheets/li-ion/SANYO-UR18650F-26A.pdf

http://www.batteryonestop.com/baotongusa/products/datasheets/li-ion/Samsung-SDI-ICR18650-26A.pdf

http://www.batteryonestop.com/baotongusa/products/datasheets/li-ion/LR1865BE-1400.pdf

tests

http://lygte-info.dk/info/Batteries18650-2011%20UK.html

stuff I found online

 

[DrkAngel]

Great info!
I was especially impressed by the "tests".
Graphs gave some impressive stats:

2/3 of cells were tested at full discharge at a 2C rate, supplying 85-90% of rated.

Confirmed 3.5v, per cell, as optimal depth of discharge voltage. However, there seems to be a "class" of cells that works optimally to the 3.2v point. (5 cells demonstrated a near identical discharge curve.)

Tests

 

[miniebiker]

I want to do tests like that.

This will work and it will be ok to post battery test results using this ?

http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=6609

 

[liveforphysics]

Confirmed 3.5v, per cell, as optimal depth of discharge voltage. However, there seems to be a "class" of cells that works optimally to the 3.2v point. (5 cells demonstrated a near identical discharge curve.)

It's not a class of cell, those are all the same cell. I've visited the factories for some of these in my China/Korea/HongKong battery factory tours. The same factory will make the exact same product and shrink wrap it with all sorts of different brands and labels. It is extremely common in the 18650 market.

You see it with larger cells as well. Anyone can have their own brand of ThunderSag battery, you pay them and pick a color of plastic pellet you want to extrude your cases from, and what the sticker says you want to put on the cell vent.

You could have your own DrkAngel 18650 with never having to get up from your desk, a few emails to the right people and a batch shrink-wrapped with your ES name and avatar, and any capacity claim you like could be made and shipped out to you.

It's amazing what an impact only ~1.5c (5A) discharge makes on these weenie cells. The drop in voltage makes a bigger loss in energy (watt-hours) than the piddly loss in capacity, and they didn't make a graph to show it. You only get to use about 1/2-2/3's of the energy your pack is capable of delivering if you were to discharge it over a 10-15hr period (like the 0.2A tests).

 

[miniebiker]

Confirmed 3.5v, per cell, as optimal depth of discharge voltage. However, there seems to be a "class" of cells that works optimally to the 3.2v point. (5 cells demonstrated a near identical discharge curve.)

It's not a class of cell, those are all the same cell. I've visited the factories for some of these in my China/Korea/HongKong battery factory tours. The same factory will make the exact same product and shrink wrap it with all sorts of different brands and labels. It is extremely common in the 18650 market.

You see it with larger cells as well. Anyone can have their own brand of ThunderSag battery, you pay them and pick a color of plastic pellet you want to extrude your cases from, and what the sticker says you want to put on the cell vent.

You could have your own DrkAngel 18650 with never having to get up from your desk, a few emails to the right people and a batch shrink-wrapped with your ES name and avatar, and any capacity claim you like could be made and shipped out to you.

It's amazing what an impact only ~1.5c (5A) discharge makes on these weenie cells. The drop in voltage makes a bigger loss in energy (watt-hours) than the piddly loss in capacity, and they didn't make a graph to show it. You only get to use about 1/2-2/3's of the energy your pack is capable of delivering if you were to discharge it over a 10-15hr period (like the 0.2A tests).

I like the ones that are from the same factory that makes those pink Samsung cells !
They don't test the same as any others I have found !
Who makes those ?
And would the charger I showed be ok for tests posted on this forum ?

I want a charger that will give good results.

 

[miniebiker]

liveforphysics,

Why would anyone want to use cheap made in China batteries to build a homemade battery pack ?

( some have no name or numbers at all on them )

That's what people use that are selling battery packs.

I want it to be homemade because I want it to be the best.

 

[Ypedal]

liveforphysics,

Why would anyone want to use cheap made in China batteries to build a homemade battery pack ?

( some have no name or numbers at all on them )
That's what people use that are selling battery packs.

I want it to be homemade because I want it to be the best.

I can only assume those are factory rejects that are never meant to be sold to joe public, but make it out the back door... someone slips on generic no-name shrink tube and throws them on ebay. :wink:

 

[miniebiker]

liveforphysics,

Why would anyone want to use cheap made in China batteries to build a homemade battery pack ?

( some have no name or numbers at all on them )
That's what people use that are selling battery packs.

I want it to be homemade because I want it to be the best.

I can only assume those are factory rejects that are never meant to be sold to joe public, but make it out the back door... someone slips on generic no-name shrink tube and throws them on ebay. :wink:

I know the batteries that are not rejects.

The ones I pulled out of laptop packs.

And they tend to match each other real well in the same pack.

See the results of the fake Sony batteries ? All over the place. How ya gonna make a pack out of those ?

 

[DrkAngel]

The "2900's" seem to have a totally different discharge profile.
This would indicate a different chemistry-formulation.
They demonstrate a faster voltage drop, but the drop, extends further ... to a lower voltage.

18650 discharge capacities & profiles

 

[liveforphysics]

This is the chart that matters, and this is only 1.5C discharge.

 

[miniebiker]

This is the chart that matters, and this is only 1.5C discharge.

Are you using the numbers on the batteries or the numbers on the test to come up with 1.5C discharge is 5 amps ?

You don't want to use these cells at 5 amps discharge.

That is why you need to parallel them.

 

[liveforphysics]

5amps would be roughly 1.5-1.8C discharge on the cells in this capacity range.


I agree, you wouldn't want them to be operated at 5amps, you end up throwing away ~30% of your pack's energy to heating/losses.



You also may think that because you spend 4hours of riding around to drain your pack, that your C-rate is 0.25C (which would be your average C-rate).

However, average C-rate has NO effect on capacity and losses from your pack is determined only by the current you pull from the pack while under throttle. So for an example, if when you're drawing current you're drawing 30amps (or whatever) from a 15Ah pack, then it doesn't matter if you spend 2 weeks draining the pack down by hitting the throttle in little pulses once in a while, if every time you're on throttle it's pulling 30amps, then your pack will have the loss of 2C discharge even if your average C-rate is 0.000001C from spreading out the 2C discharge over weeks or whatever.

That's really important to realize for the folks who think, hmm... I spend 2hours riding to drain my pack, so I must be 0.5C discharge. The pack has no idea and no cares about what average C-rate may happen to be, it feels the losses instantly from voltage sag etc as you pull the current.

 

[DrkAngel]

This is the chart that matters most ... to me!

My typical usage! With my 25.9v - 31.2ah Li-ion battery pack, only 9 lbs.