Transform Dead Laptop Batteries into a reliable E-bike Battery!

 

I decided to write this guide to describe how to recycle non working or “dead” laptop batteries and to use them to build a recycled and reliable e-bike battery pack.

I had the inspiration to build this guide because of this discussion I read on the Endless-sphere.com forum. I therefore would like to thank all the subscribers of the thread that contributed with their ideas and knowledge. In particular, a big thank you goes to “DrkAngel” for his continuous supply of useful information on the topic.

 

Edit (12/1172012), I added the video for even better comprehension of the whole process:

 

In this little guide you will find information, methods and ways to build, test and use safely and e-bike battery using exhausted laptop batteries. As it will be too long for me to explain in details the basic of battery chemistry,I invite you to contact me, visit the link I posted above, read some good book and surf the web where you can find all the information you need.

Disclaimer: I am not responsible for any damage to people and/or things due to improper use of equipment, materials, batteries and cells included in this guide.If you are unsure, do not attempt to build anything. Call an expert and/or a technician to help you.

This text is written for educational purposes only.

Also, this guide is available for free.

Considerations on the use of this battery pack when completed.

These batteries are not built to be discharged deeply. The typical discharge is about 1C with short busts of up to 2C for a few seconds. If you exceed this values, you risk overheating the cells and making them unusable. In worst cases you could also have explosions and fires. Be considerate of these factors when building your battery pack!

For example the pack I am building will be about 864Wh and my typical usage will be between 300-500w (less than 0.6C) with peaks of 1200w ( about 1.4C) for a few seconds (less than 5 sec. in my normal use).

A good definition of C-rate is available here.

These cells are 3,6volts  nominal each and their capacity is typically between 1700 and 2900 mAh. A single cell fully charged has a voltage of 4.1-4.2v (this depends on the charger used, mine recharges cells up to 4.1v). It’s ok to discharge these cells until 3.0v but on my experience I suggest not to go under 3,2v under load to avoid any damage and excessive heat.

Material & Tools needed:

  • Laptop batteries (you can get them for free at local pc repair shop or for as little as 1$ a piece)
  • Li-ion battery charger (I used an RC battery charger, about $30)
  • 1-8s li-ion/life/lipo Battery alarm (About 2-4$ each on ebay)
  • Multimeter/Tester (from $20+)
  • 12v 50w and/or 20w light bulb (2-5$ each)
  • Solder and rosin (30$+ but a spot welder would work best)
  • Wire and alligator clips (10$+)
  • Permanent marker
  • Notebook and pen or pc to record result of the tests
  • Stopwatch (I used an app on my phone)
  • Other materials may be needed depending on how you want to use/store the battery.

Opening the battery

Before opening the batteries record on an piece of paper the details that you find on the sticker of the battery pack:

As you can see the voltage is 10.8 volts and the capacity is 4.4 Ah.
When I opened this battery I found 6 cells connected in a 3s2p configuration (3 series and 2 parallel) as you can see in the following picture.
It’s important  that you mark the capacity of the battery if you are dealing with batteries of different Ahs.

Use a flat head screwdriver to open the batteries. Be careful not to get hurt! Most of the batteries are glued to the plastic container. Normally if you peel the label you can see to the cells, but you still have to get rid of the plastic before you can actually access to them. Be careful not to create any shorts between cells and not to physically damage any cell. Even if you think they are “dead”, some cells may contain a lot of energy!

The next step is extract the battery pack from the plastic case and get rid of the electronic board by cutting the wires one at the time to avoid any shorts.

What I normally do at this point is to mark the cells with a letter and write the capacity on it with a permanent marker. Keep the cells in the same configuration and proceed to charging.

First charge:

I used a RC programmable charger for this project. They are not hard to find and you can get them over the internet for 25$ and up.
Do a first charge under the Li-ion program. In the menu you can choose different settings. I did the first recharge at 1A. If the voltage is too low and the charge won’t start, I use a little trick to boost up to voltage to get your first charge started.
Set the charger to “NiMh charge”. Start the charge at 1 A until the voltage reaches at least 3 volts per cell. For example if you have 3 cells, you can stop this quick charge at 9.5volt. Monitor this operation closely as usually it will take less than 2 minutes for the batteries to get to the minimum voltage. Stop the charge and go back to li-ion charging mode.

It’s important to charge the batteries in balance mode, so that every cell gets charged to the same voltage.To do that you will need balance leads and follow the instruction on your charger’s manual.

In this photo you can see the voltage of the single cells.

After the first let the battery pack rest for at least 24 hours. Then check the voltage of each single cell with the multimeter. If the voltage has dropped more than 0.1volt from the final charge voltage discard the cell.

At this stage,as I am testing hundred of cells, just to be organized I decided to make three bins for cells to be tested, good cells, and bad cells:

Testing.

Unless you have professional equipment it’s not easy to determine the exact capacity of each cell. I use a simple method to get an idea if the cells I am testing are good or not.

Once we have divided the good batteries from the bad ones we use a 12v 50w light bulb to check how good our cells are.

You will need:

12v 50w light bulb
3s2p packs (they can also be 3s3p or 3s4p)
2-8s battery alarm
Wire and alligator clips
Stopwatch (I use an app on my phone)
Glass Jar (optional)
A good dose of common sense.
Patience.

Connect the 1-8s battery alarm to the wires (you can do this in many ways) and alligator clips, I used an old floppy disk power lead, you probably can do better than me, the result is this:

Notice the tape in the “speakers”: the alarm is really loud.
A suggestion for a better result (I didn’t have time) using an hdd data cable:

Now prepare the lightbulb. It’s better if you have a proper holder, I didn’t so I used a bit of creativity.

Remember to use wires of adequate gauge: the more the Watts, the thicker the wire.

Connect the battery alarm to the cells as written behind it.
Remember to connect the minus clip last. If you connect it first it will not recognise your full battery pack configuration.

Now you can connect the light bulb to the battery and start the stopwatch as soon as possible.

Discharging 3s2p @ 12v 50w

Remeber to use common sense performing the test. Don’t leave the light bulb and the battery unattended!
Take notice of excess of heat coming from the battery. As we are discharging at around 1C it’s normal for the cells to get warm, but they shouldn’t get too hot. If this happens stop the test immediately.

For the first test set the alarm to 3.60 Volts. If the voltage drops quickly (1 minute or less) try lowering the threshold. I would suggest 3.20 Volts as a lower limit. Do not go under 3.00 volts or you risk permanent damage to the battery.
The purpose of this test is to see if the cells can still hold the charge and to spot any bad cell that has passed the first screening.
Once you found the optimal voltage alarm you can start recording your results. In particular The voltage of all cells when the alarm goes off and the cell that triggers the alarm. A good pack will have the voltages of the cells very close together typically between +/- 0,02 v difference.
Every brand of cell has his own specifications (especially the discharging curve) and it would be optimal to have the same cell brand,type and capacity to build our e-bike pack.

Once you got some records you can pair the cells together. I firstly recharged the cells to the same voltage (4.1v) and then I made some 3,6v “blocks”, in my case soldering 12 cells in parallel:

1s12p

Then when you have all the 3,6v “blocks” you need you can re-test them using the “light bulb method”. In my case I connected 3 x 3,6v 24Ah in series (total 10.8v 24Ah) and discharged them with 3x 12v 50w light bulbs in series (total 12v 150W). Remember to use wires of adequate size, for the test above I recommend 12AWG  or better (10 AWG).

If the tests are successful you can then proceed to build the pack:

Be careful no to make any short during the pack assembly.

The configuration below is not good, it will discharge the cell unevenly and eventually will cause some cells to fail:

Here is the battery pack completed (10s12p) with fuse and “Dean T connectors”, notice the 1-8s battery alarm connected to the pack to monitor in real-time every single 3.6v “block”:

Remember to insulate properly the battery to avoid any short!

I decided to make the pack “splittable” into 2x 5s12p so I can either recharge it in different configurations (5s or 10s).

Performance.

I tested the pack on my electric bike. The bike is equipped with a 36v 500w geared hub motor. So far the voltage does not go under 3.5v per cell under load (300-500w typical with 1200w peaks) and I estimate a range on my bike between 55 and 75km, so far my longest trip was 27km, and the battery was well over half full. Not bad for some batteries considered “dead”!!

For any questions feel free to comment or contact me privately.

Advertisements

13 thoughts on “Transform Dead Laptop Batteries into a reliable E-bike Battery!

  1. So you wire your pack in parallel and after you put them in series?
    I see most people put them in series first … What’s the difference, I am starting a pack and I am trying to find the best way to wired them …

  2. Hi ccr,
    Yes,first you put the cells in parallel. When you do this the cell should have the same voltage (maximum difference 0.2V between each other), this will avoid problems (e.g.cells that are charged will discharge into the ones that are not charged) and the pack will be easier to balance at the end. Moreover it’s more practical (at least to me) to parallel first: you make your 3.6v “blocks” and then you can decide how many in series you want to connect.
    Hope this help 🙂

  3. I used an Imax B6. Pretty much any RC charger will be good. If you have $$ you can choose more accurate & powerful ones like the Hyperions.

  4. hi great work, would like to ask if you could you make a wire diagram of where to place the balance leads and how you connect it to the balance charger cheers.

  5. Hi, thank you for putting this all together on one page. I read the 32 pages on the EndlessSphere site and it made my head spin.
    I am starting a 10S 10P battery but it is still not clear to me how to charge the final pack. Do you use your Imax B6 charger to charge 6 sets of cells then reconnect to charge the other 4 sets? Can two B6 chargers be used in tandem or will that create a ground loop problem. How long does it take you to recharge your 12p 10s and how many amps do you put into it?

    • I am using this http://www.bmsbattery.com/alloy-shell/23-alloy-shell-180w-lifepo4li-ionlead-acid-battery-ebike-charger.html for the whole pack. Fully discharged it takes about 4-5 hours to top it up, but I have discharged the pack fully only once. Normally in 2-3 hours max the pack is charged. It’s strongly suggested that you use a bms in this case. I used to balance manually but the cells but it’s a pain.

      Alternatively,You can use the imax b6 or any other rc charger simply splitting the pack in two 5s packs. You do a balanced charge on each 5s pack and then you connect the two packs in series on the ebike. In this case it’s worth having 2 imax b6 (they’re cheap).

      If you have the $$ you can use more powerful chargers, you can put as many amps as the Ah nominal of your pack, but I recommend to stay a bit lower, just for safety. For example my pack is 24Ah, and I could charge it at 24A but for me 10-12A would be quick and safe enough. Don’t forget that those cells are used, the less “stress” (meaning less amps, both ways) they get/give, the longer will last!

      • Thanks,
        I went on line planning to buy 2 Imax b-6 chargers and ended up ordering a Quattro 6×4. It wasn’t that much more than buying 2 b-6’s and I figured it should speed up the process of sorting and matching all those cells.

  6. I understand the reason for building the 3.7V parallel blocks from closely matched cells. How much difference in capacity would you consider acceptable between “blocks” being assembled in series? Did you record the individual cells capacity or the combined capacity of 2 or 3 paralleled cells with your light bulb test?
    I bought one lot of, what turned out to be, new unused laptop batteries and ended up with 72 cells that are at full rated capacity (2500 mah). the second batch of cells are not so good and so far the best of them are only about 1800 mah. How far can they deviate from the mean and still be useable?

    • Hi Art Brasher,

      sorry for the late reply, I have been quite busy lately.
      I measured the capacity of the cells as I found them in the laptop batteries, normally they are 3s2p (6 cells total, 3 series and 2 parallel).
      Don’t mix the 2500mAh cells with the 1800 mAh because when you parallel 2 or more cells with different capacity, all your cells will assume the lower capacity even if they have more mAh, in your case 1800mAh. The cells will still be usable but you loose a lot of capacity…
      I suggest stay within 100 mAh as max difference when paralleling cells. And remember they need to be at the same voltage (100 mV difference max ) before paralleling to avoid damage (fast charge/discharge on the cells you are paralleling).
      Happy holidays!

      • Thank you for the reply.
        I understand the necessity for using only closely matched cells for groups in parallel.
        When assembling these groups into series strings what would be the effect of differences in capacity between groups? My assumption is that all of the groups in series would come to match the capacity of the lowest one. The lower capacity cells have no problem reaching and holding 4.15~4.2 volts when full but have a shorter discharge curve.
        Happy New Year

      • Hi.
        Let’s see…in case of group of cells with different capacity connected in series,the cells that have more capacity will tend to maintain a higher voltage during discharge, so in the end you will need a bms to monitor closely the cells with lower capacity as the total pack voltage may seem fine while it’s not. If you’re using a bms, as soon as the lower cell reaches the low voltage cutoff, the bms will stop the discharge. So in the end you will end up not using your full pack capacity anyway.If you don’t use a bms you can be in big trouble, risking to damage the cells with lower capacity every time you discharge your pack. Mixing cells with different capacities is never a good idea, it’s still doable within some limits, but in my opinion the disadvantages are too big compared to the benefits.
        Happy New Year!!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s