Homemade Battery Packs

You will find it's more difficult than that to measure capacity. The current through the resistor decreases as the SOC drops, as well as the Ri increasing, so quite a few shifting variables that the resistor method misses.

An RC charger with a "discharge" function is the cheapest method to get accurate capacity readings.
 
liveforphysics said:
You will find it's more difficult than that to measure capacity. The current through the resistor decreases as the SOC drops, as well as the Ri increasing, so quite a few shifting variables that the resistor method misses.

An RC charger with a "discharge" function is the cheapest method to get accurate capacity readings.

OK, so it's not a linear drop. Should be a pretty good exponential like Beer's Law, so take the logV vs. time and you'll get a straight line through the LVC, then integrate under the original curve to get total mamp. The advantage of this is that you don't have to go anywhere near the LVC as long as you can plot a good curve. Come on, now, Ri is on the order of milliohms, so that will be negligible.

555's < $0.50/ea and LM324's < $0.17/op amp (4/IC) [all US sellers, fixed price, del w/in 3 days] . You can get a lot of them for 1 RC charger and get a pretty good estimate of the capacities of a lot of batteries at a time.

Cameron
 
Red_Liner740 said:
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.
 
in the process of testing cells right now...i'm charging them to 4.1v, what i'll be charging them when in use.

all cells are paired in two. i have four 11w 12v bulbs.

i take 3 of these 2 packs and parallel them up to balance the cells inbetween themselves, a few minutes of this and they are all the same. attach them in series and attach to one 11w bulb. 4.1v x 3 = 12.3v. 11w/12.3 = 0.89A draw. If my calculations are correct. Someone please correct me if i'm wrong.

soo...these packs are supposed 2600mAh so 2.6amp, so technically it should take them almost 3 hours to discharge...hmm...i might have to wire two bulbs in series to bring that up close to 1C discharge rate.
 
Red_Liner740 said:
...
all cells are paired in two. i have four 11w 12v bulbs.

...
soo...these packs are supposed 2600mAh so 2.6amp, so technically it should take them almost 3 hours to discharge...hmm...i might have to wire two bulbs in series to bring that up close to 1C discharge rate.

2600mAh: Is that per cell or per pair of cells? Most Lithium Cobalt cell are high energy density, so 2600mAh per cell is more likely. If that's the case, your pack capacity is 5.2Ah (5200mAh).

You don't want to wire two bulbs in series, that would cut the voltage through each bulb by half. You want to run 2 bulbs in parallel so each load (bulb) gets the full voltage.

Eventhough your cells are charged to 4.1v, you still want to use the nominal 3.7v voltage in your calculation.
 
Correct on all accounts sir. i forgot that they are parallel which would give me the 5200 A/h.

So far i've culled a few that run out of steam very quickly, about 30% of what the rest of the cells are holding...

i just wish i had a way to bulk charge these SOBs, i got the little 18650 chargers of ebay but they are 0.6amp charge, so thats 0.3amps per cell....even with 5 of them going at once, it takes a while to charge, rest for 24hrs, discharge and monitor....lots of spare time needed.

to the point where i'm thinking HK Lipos might just be a better way to go. Well, i mean, we all know that!! :mrgreen:
 
Red_Liner740 said:
i just wish i had a way to bulk charge these SOBs...

Here's how I do it:
I have a 12Volt, 5Amp lead-acid charger. This is the cheap, common charger that most car owners have lying in their garage.
I built a bunch of 1s6p packs. Each of my cell capacity is 2.5Ah, so each pack is 15Ah. I strung 3 of them in series to form a 3s6p pack (11.1V, 15Ah). Since my charger is only 5A, I'm charging the pack at 0.333c. Because of the low charging current, the packs tend to "cell balance". But not always. I monitored the voltage of each pack separately and bleed off when necessary (very rarely). When the voltage hit 12.6 (4.2V per pack) I stop.
 
thats risky business. i would and have done that when im around to watch the packs. But i would only use an SLA charger on random cells for random projects...i dont think i have the balls to hook a SLA charger to my 20cell lithium pack...too much work to destroy, when i could prevent the very thought by buying a 40 dollar charger or splitting your pack and charging at a lower voltage with a lithium charger you have kicking around.
 
I agree. That's why I monitored the voltage.

Anyway, this is only temporary until I make up my mind on either LiPo charger or full BMS.
 
^^ Agreed. This is only temporary until my meanwell charger comes in, but i'm impatient. And i dont have a 12v car charger, i only have a 1amp battery trickle charger for my motorcycles and from what i understood, trickle chargers are a no-no.

I also have 5 more of those little 18650 chargers coming in for a total of 10, ought to make the going faster in the mean time.
 
SamTexas said:
DrkAngel said:
iMax B6 charger - $23.78 - shipped


Will charge 1-6cells at up to 5 amps. Balance function available with balance connector - plug.
Looks good. Which balance connector-plug do I need? Where can I buy it? Thanks

For 3 cells - Sets RC Lipo Battery Balance Charger Plug 11.1V 3S1P

For 6 cells - RC Lipo battery Balance charger plug 6s1p 22.2v 6cell

Assortment 2s 3s 4s 5s 6s w/22ga silicone wire, (much better quality) - 2S-6S 7.4V 11.1V 14.8V 22.2V lipo balance plug set Cool
 
Homemade Li-ion 3000 mile 25.9v - 31.2ah pack retired to backup-auxiliary status.

3000mile new balance.JPG

Before "retirement", pack was separated, and tested.
1 dbl cell replaced and 2 cracked solder connections repaired.
"Modern" balance plug added.
Then, range tested - 30+ miles at 16mph!
 
Finally together! 25.9v 31.2ah, in an EZip pack.

I started with 6 cell Gateway notebook batteries, (2600mah cells).

attachment.php

6cell oem.JPG
I really lucked out, these packs had cells connected, in such a manner, that 2/3rds the soldering was already done.

attachment.php

6pack.JPG

I ganged, similar voltaged cells, together, then charged "en mass", (6cell balance charger at 5 amps). Balance charger "balances" at a very low rate, so I individually "top charged" lower banks to attain faster equalization, at 4.2v. Cells were separated into original 6 packs, after equalization, to allow individual "bleed down" for 24 hours. Cells that maintained 4.19v +or- .01v, were selected for build.

attachment.php

6cell gang balance.JPG
Pack fit nicely into the OEM EZip pack, I did have to cut out the center screw mounts. 14ga wiring was replaced with 12ga and the 22ga charging wire was replaced with 16ga. 8wire, 7 cell, balance plug was also added and made retractable, through a small side hole.

attachment.php

EZip 31.2ah Liion (600 x 450).jpg
Cells were securely wrapped with clear boxing tape and sheets of closed cell foam added, to firmly secure cells.
1/4" "Masonite" board was added, as solid base, for cells, and as secure barrier between cells and bottom contacts.

Next?
Road trial!
Just have to wait for some decent weather.
I estimate range at 40 - 50 miles, at the EZip Oem 16mph speed.
20 mph cruising speed should attain 30 - 40 miles.

The OEM, 10ah SLA batteries, when expended in 1 hour, output, only , 6.14ah!
 
The pack that currie never wanted to to sell you.
 
37v 20.8ah, Li-ion pack, will also fit nicely into the same EZip pack! Requires trimming of the, now useless, side-center "pillars". It even leaves room, to integrate, a PCB, above the cells, next to the charging port.

Also possible to integrate PCB into the 25.9v - 31.2ah pack. Requires side pillar removal and cells moved to one side.
 
Handy little item I made up. A 3s balance - charger adapter.


Works great for 11.1v notebook pack cells.
Requires Lipo - Li-ion balance charger.

3Sbalance.JPG
Wires are separated, in order of connection, to simplify proper connection order.

I started by soldering balance adapter plug, to each set of cells.
Decided it was too much work.
 
Hillhater said:
wow , 32 Ahr ! .. very nice pack.
what does it weigh ?
What C rating are those cells ?

Total weight -10lb 6.4oz

C rating ?
Tested a 11.1v 5.2ah "6pack" w/"Toro 250" 12v blower
Surged past 20amps, settled to 11amps,
decreased to 10 amps after 1 minute,
decreased to 9 amps after 5 minutes.

Maximum* C rating:
= 4C surge (maybe higher? - limitation of test device )
= 2C continuous (as tested for 1 minute)
= 1.8C continuous (as tested for 5 minutes)
*NOTE: Exceeds what I consider "acceptable", or usable, discharge rates!
But demonstrates brief, or emergency, capability.
Note: cells became "comfortably warm" after 5min test. (Held against face - maybe 120F)

Recommended C Rating
= 2C brief surge
= 1C maximum use
= < .5C Continuous
 
DrkAngel said:

Wires are separated, in order of connection, to simplify proper connection order.

I started by soldering balance adapter plug, to each set of cells.
Decided it was too much work.

Looks like a short waiting to happen.

As for your calculating of the C rate, your method of calculation is not valid.
All battery chemistries can put out very high amounts of amps..
Voltage sag is what you need to look at.

A car's 50ah lead acid battery is rated to put out something like 500 cold cranking amps.
That doesn't make it a 10C battery.
 
neptronix said:
As for your calculating of the C rate, your method of calculation is not valid.
All battery chemistries can put out very high amounts of amps..
Voltage sag is what you need to look at.

A car's 50ah lead acid battery is rated to put out something like 500 cold cranking amps.
That doesn't make it a 10C battery.

Obviously you have never looked up the definition of "C rate"!
"Voltage sag" is not a factor in calculating "C rate"

By definition, the "C" "discharge rate" of a battery is 1hr/expended to fully discharged voltage, (after load removed), time (expended in 6min - 60min/6min=10C)
True "maximum discharge rate" would be determined by a dead short, till, when short removed, minimum voltage reached.
Due to massive inefficiencies, and danger, this seems worthless in determining usable amps, or ah's.

My "Surge C" and "Continuous C" are rated by the actual usable amps available.
Note: Even during "surge", voltage never dropped, (sagged), to the "rated" 11.1v.

Since voltage never dipped to the rated voltage, actual "C" rating , based on "voltage sag", should be higher! ... ???
I just need a more powerful test motor!

Anyhow, ... it is enough for me, that I demonstrated that my 31.2ah pack is capable of 125amp surge and 60amp continuous.

My actual usage is a mere 30amp surge and 10-15 amp continuous. = 2-3 hours biking time!
From a ten pound battery!

1C continuous usage gives you 1 hour biking time.
10C continuous usage gives you 6 minutes biking time!

What is your definition of the "C rate"?
 
DarkAngel- Just for fun, where would you place the maximum usable C-rate for this LG 18650?
It has 3.05watt-hours usable energy at 3C (and can only deliver that much energy if you're hitting it continuously so the cell stays hot from the initial voltage drop heating it).

Or, 7.7 watt-hours usable energy at 1C.

Or, 9.4 watt-hours usable energy at 0.2C

18650_2400mAh.jpg




Also, would you continue to call it maximum C-rate if the cell sagged below 1/2 resting voltage at that C-rate (normally these little feces cells just saturate the current collector and fall on their face before this happens), so that as current output increases, the amount of power you're getting from the cell is decreasing?
 
liveforphysics said:
DarkAngel- Just for fun, where would you place the maximum usable C-rate for this LG 18650?

You forgot to list the C rate usage, for each different colored line!

Which is the entire purpose of the graph!
 
Back
Top