Homemade Battery Packs

Hi guys,

very interesting thread.

I have a couple of questions:

to what voltage is safe to discharge those laptop batteries?
Is 3.0 volt too low discharging at 1C or less?

I am building my own recycled pack :)
 
spuzzete said:
to what voltage is safe to discharge those laptop batteries?
Is 3.0 volt too low discharging at 1C or less?

I am building my own recycled pack :)
Click to expand...
Not as easy to answer as you might think.

1st, every cell type, manufacturer, capacity cell, has different optimal voltages.
In my testing, I have rated optimal depth of discharges at 3.4V, 3.5V, 3.6V, 3.7V, (static voltage), dependent on the cell type.

I would recommend never dropping to 3V, even under maximum throttle.

The red 18650 cells appear to be 95% discharged at 3.6V, any further discharge will drop voltage precipitously, resulting in unnecessary cell damage.

So ... find your optimal depth of discharge (DOD), then apply a 1C load to determine optimal voltage under load.
 
Here is a datasheet i found on sanyo 2600mah 18650 cells

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

they state that you can discharge to 3.0v on there cells at 1c discharge rate.

If you have a very well balanced pack that you built using a cell matching then going that low would probably be ok however if you don't then going a bit higher should protect your cells from one of the cell banks going too low a voltage before the lvc kicks in.

when i test 18650 cells and discharge to 3.0v at 5 amps they bounce back to 3.5ish afterwards which shows they haven't been overly discharged.
 
chilledoutuk said:
Here is a datasheet i found on sanyo 2600mah 18650 cells

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

they state that you can discharge to 3.0v on there cells at 1c discharge rate.

If you have a very well balanced pack that you built using a cell matching then going that low would probably be ok however if you don't then going a bit higher should protect your cells from one of the cell banks going too low a voltage before the lvc kicks in.

when i test 18650 cells and discharge to 3.0v at 5 amps they bounce back to 3.5ish afterwards which shows they haven't been overly discharged.
Click to expand...

My assertion is that ...
The amount of energy available between 3.6V and 3V is miniscule!.
Especially compared to the resulting damage caused to the cells!
Damage in the form of a permanent loss of capacity and the reduction of available cycles.
 
DrkAngel said:
1st, every cell type, manufacturer, capacity cell, has different optimal voltages.
In my testing, I have rated optimal depth of discharges at 3.4V, 3.5V, 3.6V 3.7V dependent on the cell type. (static voltage)

I would never recommend dropping to 3V, even under maximum throttle.

The red 18650 cells appear to be 95% discharged at 3.6V, any further discharge will drop voltage precipitously, resulting in unnecessary cell damage.

So ... find your optimal DOD (depth of discharge), then apply a 1C load to determine optimal voltage, under load.
Click to expand...


Thank you DrkAngel,

Now I have a clearer picture of how I will perform my tests :)
 
My method of determining optimal voltages - Lithiums - mAh/100th V - Discharge Tests

A low tech alternative.
Full charge - check voltage.
Run lap, (around track? around block? etc.) the shorter the more accurate.
(Maintain consistent laps, same speed or same throttle usage ... etc.)
Measure voltage at stop after each "lap". (Digital meter, highly recommended.)
Chart it.
Voltage will drop sharply till optimal charged voltage is reached.

When voltage begins to drop sharply, again, you have exceeded "optimal" low voltage.
 
Ypedal said:
this entire thread is a trainwreck........ it reeks of " safe " ..
Click to expand...

I don't think anyone else on this thread owns a metalized fire suit. Oh well.... :lol: :lol: :lol:
 
DrkAngel said:
My method of determining optimal voltages - Lithiums - mAh/100th V - Discharge Tests

A low tech alternative.
Full charge - check voltage.
Run lap, (around track? around block? etc.) the shorter the more accurate.
(Maintain consistent laps, same speed or same throttle usage ... etc.)
Measure voltage at stop after each "lap". (Digital meter, highly recommended.)
Chart it.
Voltage will drop sharply till optimal charged voltage is reached.

When voltage begins to drop sharply, again, you have exceeded "optimal" low voltage.
Click to expand...

Thanks for the info!

I am testing the 3s2p (10.8v charged at 12.3v - 4Ah) packs using a 12v 50W light bulb. I use the 1-8s voltage alarm (the one that was posted also in this thread) set to 3.6 to each cell and I record the time until it beeps. So far the best pack timed 19.5 minutes to reach 3.6volt, all the cells were discharging with the voltages being very close together.
 
spuzzete said:
Thanks for the info!

I am testing the 3s2p (10.8v charged at 12.3v - 4Ah) packs using a 12v 50W light bulb. I use the 1-8s voltage alarm (the one that was posted also in this thread) set to 3.6 to each cell and I record the time until it beeps. So far the best pack timed 19.5 minutes to reach 3.6volt, all the cells were discharging with the voltages being very close together.
Click to expand...
Good methodology!
"Equalize" pack by building banks of equal "minutes".

Example - 36V 20Ah pack
3 = 10.8V ....+3 = 21.6V .....+ 3 = 32.4V ....+ 1 = 36V
17 min ......... 16 .............. 15 .............. 17
16 .............. 15 .............. 16.5 ............ 17
18 .............. 17 .............. 14.5 ............ 17
16.5 ............ 17 .............. 18 .............. 16
15.5 ............ 18 .............. 19 .............. 16
-----------------------------------------------------------------------
83 min ......... 83 .............. 83 .............. 83

Actual bank "balances" might vary, under your typical use.
So keep your numbers available for "re-shuffling".

If you want to get precise ... note the voltage variance of the cells in each 6-pack ...

Notes:
With recycled notebook cells, I like to keep continuous usage under .5C and surge under 1C.
With my 24V 450W - 36V 675W motor, that requires a 30Ah pack. (10 - 15lb pack)

Good Luck
Keep us posted!
 
SamTexas said:
DrkAngel said:
Lithium Cobalt charging, at moderate rate, is 98-99% efficient, a good measure of capacity.
...
, but lithiums, particularly Lithium Cobalt, at moderate charge rates, attain near 100% efficiency!
Click to expand...
Those percentages sound impressive. But have you verified their accuracy?
Click to expand...
Battery University.com
"Manufacturers recommend charging the 18650 cell at 0.8C or less. Charge efficiency is 97 to 99 percent and the cell remains cool during charge."

A very informative article - Charging Lithium-ion
Required reading, for anyone interested in using lithium ion batteries!
 
The difference in remaining capacity at standing of 3.6 to 3.0v may be minuscule but at 2c its most of the cells capacity.

If your running your cells at even 1c discharge a lvc of 3.6 will still be cutting you off way too early.

Discharging cells at 5A they don't reach the cliff until they are about 3.2 to 3.3 volts.

What I am saying is if your running your 18650 cells at a highish discharge rate 3.6 lvc would be cutting you off way to early.

if you have a huge pack and only run at say 0.5c peak discharge current then maybe 3.6v might work ok.

But however in the same situation that pack would be fine down to say 3.1v if the pack was built properly will cell matching used.

For example when building one of my packs I tested the capacity of every cell and then was able to get all cell rows within 2mah capacity of each other using a cell matching program kindly provided by another ES member.
 
Sorry ... thought I was being clear.

I recommended 3.6V as the minimum resting voltage. ("Static Voltage")
Not the voltage under load.
To determine optimal LVC ... (Low Voltage Cutoff)
Deplete pack to 3.6V idle voltage.
Then apply your typical throttle.
Take note of voltage under this sustained load.
This should give you your optimal LVC voltage-point.

Regardless ... I still recommend not dropping below 3V, under throttle.
 
Thanks guys for your feedback.

I would like to make a tutorial (pdf) and a video of what I've learned here and my way to build the battery pack once I've finished testing & assembling the batteries.

I think it would be useful for many people, and quicker to read/get to the point. This thread is excellent but is 25 pages long :shock:
 
spuzzete said:
Thanks guys for your feedback.

I would like to make a tutorial (pdf) and a video of what I've learned here and my way to build the battery pack once I've finished testing & assembling the batteries.

I think it would be useful for many people, and quicker to read/get to the point. This thread is excellent but is 25 pages long :shock:
Click to expand...


that would be nice, I'm pretty new to all this stuff, and read all 25 pages and still feel very lost on several things, I still am very unsure of how to go about building my battery pack, not to mention all the testing and charging after building the pack... I just can't afford to run out and drop $400+ on a battery pack, and can't stand how bad the SLA's are.

one thing I'd like some clarity on is what is the xC (like .5C - 4C) that's being talked about so much, and what is the meaning of the things like 14S8P, I'm guessing the configuration of the batteries, but have not found confirmation... would that be 14 series, 8 parallel? so a 51.8 volt pack, then running 8 of these packs in parallel to boost Ah ?

I have an ezip and the battery died in less then a year, bikes been siting 2 years now due to new baby and no time to do anything, so now I have some time and would like to get the bike up and running, but better then it was.
 
I also highly recommend adding your information to the ES Wiki, in the appropriate pages if they exist, or creating new ones if they do not:

http://endless-sphere.com/w

RIght now the wiki is a mere skeleton, and needs fleshing out by anyone knowledgeable on specific subjects. :)
 
swampjeep said:
one thing I'd like some clarity on is what is the xC (like .5C - 4C) that's being talked about so much, and what is the meaning of the things like 14S8P, I'm guessing the configuration of the batteries, but have not found confirmation... would that be 14 series, 8 parallel? so a 51.8 volt pack, then running 8 of these packs in parallel to boost Ah ?

I have an ezip and the battery died in less then a year, bikes been siting 2 years now due to new baby and no time to do anything, so now I have some time and would like to get the bike up and running, but better then it was.
Click to expand...


You are correct, 14s8p is 14 series 8 parallel. You parallel the batteries to have more Ah available. It's important that you parallel only batteries of the same chemistry,same producer, same Ah, same voltage. On the packs we are building with laptop batteries it's important to have a lot of Ah available to keep the C rate low, that's why it's important to put a lot of batteries in parallel. For a better understanding of what "C" rate means here is a link with a good explanation: http://www.cameronsoftware.com/ev/EV_BatteryPhysics.html

Testing the batteries takes a lot of time, I choose to build a battery pack to learn something more :)

Keep us posted with your project :)
 
chilledoutuk said:
The difference in remaining capacity at standing of 3.6 to 3.0v may be minuscule but at 2c its most of the cells capacity.

If your running your cells at even 1c discharge a lvc of 3.6 will still be cutting you off way too early.

Discharging cells at 5A they don't reach the cliff until they are about 3.2 to 3.3 volts.

What I am saying is if your running your 18650 cells at a highish discharge rate 3.6 lvc would be cutting you off way to early.

if you have a huge pack and only run at say 0.5c peak discharge current then maybe 3.6v might work ok.

But however in the same situation that pack would be fine down to say 3.1v if the pack was built properly will cell matching used.

For example when building one of my packs I tested the capacity of every cell and then was able to get all cell rows within 2mah capacity of each other using a cell matching program kindly provided by another ES member.
Click to expand...

I agree with you.

I run some tests today.

Using a 12v 50w lightbulb on a 4s2p pack (green colour cells 14.4v nominal 16.4v charged-4Ah) fully charged it was only 2 minutes until it got to 3.6 volt. Same with other cells/pack of the same brand/colour.

So I run another test: I tried to recharge those packs and they were pratically still fully charged (they would take max 200mAh).

Then I discharged the same pack of the first test (green cells) with the iMax B6 @ 0.4 A connecting the 1-8s battery alarm to monitor the cell voltage during discharge. At the moment the test is still running, about 2,3 Ah have been drained from the batteries and the voltage is about 3,45v for each cell. So I think I should lower the 3,6v threshold for these cells.

With the red cell it was good the 3,6v limit.

I am preparing some videos, I will post them as soon as I have some more time.

:)
 
swampjeep said:
one thing I'd like some clarity on is what is the xC (like .5C - 4C) that's being talked about so much, and what is the meaning of the things like 14S8P, I'm guessing the configuration of the batteries,
Click to expand...
"C" is Capacity
The "C rate" is the "rate of discharge".

Amps drain / C-Amphours battery capacity
6(Amp drain) /30(Ah battery capacity) = .2C
60(Amp drain) /30(Ah battery capacity) = 2C

The simplest representation is that if you discharge at:
30 amps from a 30Ah battery = 1C or C (discharge rate) - 1 hour discharge time
15 amps from a 30Ah battery = .5C (C/2 alternate method) - 2 hour discharge time
6 amps from a 30Ah battery = .2C (C/5 alternate method) - 5 hour discharge time

oem "rated capacity" is, typically, obtainable only at the slowest discharge rates.
(SLA (Sealed Lead Acid) batteries are typically rated at .05C! (20 hour discharge time))

At higher discharge rates, the cell voltage "lags",or "sags" noticeably, AND usable capacity is reduced!
Accurate cell voltage is only measurable after removing drain.

file.php


"S" & "P"
3S2P = 3Series 2Parallel
With 3 x 3.7V - 2 x 2600mah cells this would be a 11.1V - 5200mah "pack"

Note: Laptop batteries are designed and built to be discharged at 2-5 hours!(.5C - .2C)
Not recommended for discharge rates above 1C, they will "sag" deeply under heavy drain.
 
This graph is invaluable and should be understood and memorized by anyone using or wanting to use laptop 18650 cells. Continuous discharge above 1C not only reduces the cell capacity, it also raises the cell temperature to unacceptable and potentially dangerous level. That said, it's perfectly okay to reach 2C for a short period of time (60 secs) every 5 minutes or so.
 
Hi guys,

I didn't have time to run other tests, but I have another question.

Is there a charger/discharger I can buy to log the discharging curve and obtain discharge a graph like the one posted? Any suggestion?

Thanks!
 
I put together some data from a test I did last week.

Here is a graph:

Fullscreen capture 512012 93513 PM.bmp.jpg

I discharged those 3 batteries in series (3s12p 10.8v 24ah) with 3 12v 50w lights bulb in parallel.

Not bad for batteries worth $0 !!

8)
 
spuzzete said:
Not bad for batteries worth $0 !!
Click to expand...
Not bad at all! Judging from your graph, you could easily go for another 10 to 15 minutes. Under 0.6C load, you can safely go down to 3.00V. Once the load is removed, the voltage would jump back up to around 3.5V.
 
SamTexas said:
Not bad at all! Judging from your graph, you could easily go for another 10 to 15 minutes. Under 0.6C load, you can safely go down to 3.00V. Once the load is removed, the voltage would jump back up to around 3.5V.
Click to expand...

I just wanted to see how many minutes they would last under load, and I am pretty happy with the results :). I did not want to push those batteries too much. I think once they will be on the bike it will be different, as the load will not be constant, and maybe that's a good thing especially when no power is used it will give the battery pack time to "breathe" a little.
Also I want to improve the welding configuration, i think It will improve a bit the energy flow.

I am so happy I found this thread :), thanks guys!
 
spuzzete said:
I put together some data from a test I did last week.

Here is a graph:



I discharged those 3 batteries in series (3s12p 10.8v 24ah) with 3 12v 50w lights bulb in parallel.

Not bad for batteries worth $0 !!

8)
Click to expand...
Voltage sag seems excessive?
If your leads, or jumper wires, were less than 14ga, that might explain it.
12ga+ recommended, for 24Ah battery, on bike.
 
You must log in or register to reply here.

Similar threads

digitalbilly
Garmata EEB battery box says it can hold 400pcs 18650 cells. I think I can fit more.
Replies
7
Views
257
digitalbilly
digitalbilly
M
Help with first pack build
Replies
10
Views
468
mynameisjoe
M
hyperknot
First DIY pack for heated gloves
Replies
4
Views
652
amberwolf
amberwolf
G
New or reclaimed cells for sale samsung 50S, LG M50LT, Samsung 40T
Replies
2
Views
620
glebor
G
A
[$29 plus shipping USA BASED] 36v 12.8aH 460wH (lg18650mh1 3200mAh cells) ~3.5 year Jump Bike old battery
Replies
1
Views
731
pens100
pens100
Back
Top