Many people have built battery packs out of both Nickel and Lithium based batteries for their own personal use. Maybe as a battery bank for a Solar system or for a smaller use case. I have been contemplating to build my own version of a standardised(ish) DIY battery pack. One that I could use for multiple applications.
I will be using 18650 Lithium-Ion batteries mostly harvested from disused laptop battery packs. If you have access to laptop batteries in chances are that you would be able to salvage at least some of their cells for this use. The challenge for the future in this respect will be the ever thinner laptops are using more custom battery cells than 18650 cells in order to become as thin as possible. Another source could be surplus battery packs for discontinued laptops and other battery operated appliances. These would then be available for purchase at sell-out prices.
Finding suitable Lithium-Ion cells
Going into this project I thought that I had possibly 10 cells already harvested that I will have to check the capacity on. And then a few laptop battery packs I could harvest more cells from.
Evaluating these cells is critical in order to get a good battery pack. The three critical tests to evaluate the cells seems to be:
- Cell polarity
- Ability to hold its charge when not in use
1. Cell Polarity
First entry bar is to check for reverse polarity. When individual cells have been set in series, it can end up with its polarity reversed. Any such cell is usually not worth using.
2. Ability to hold its charge when not in use
The main advantage that Lithium based batteris have over Nickel based batteries is to hold its charge when not being used. The simple test is to charge each cell to 4.2 V and store the cell for one week and then check back on it after a week. It should still hold close 4.2 V.
The Capacity test is probably the most involved. The test is to again charge the cell to 4.2 V and then discharge it while measuring the energy used. I am going to use this to 1) decide if the battery is worth using. 2) Arrange the cells so that each series has cells of similar capacity.
DIY Battery Pack Design
My main goals for this pack is to use it in applications, where 12V is needed. As far as I can tell, it is easier and more efficient to use a buck converter to take the voltage down than it is to use a boost converter to boost the voltage up, especially under heavier loads. So my plan is to build a battery pack that runs above 12V but is as close to 12V as possible.
Lithium-ion batteries usually operates with a nominal voltage of 3.7 V, with a fully charged cell running at 4.2 V and a discharged voltage of 3.4 (3.0) V. In order to get a minimum of 12V, I decided to design my DIY battery pack with 4 cells in series (4S) and a number of those in parallel in order to increase capacity. So the battery pack can be anything from 4S1P (with a total of 4 cells) to for example 4S10P (a total of 40 cells).
One design philosophy could be to include a buck converter in the unit so that the outgoing voltage would always be 12V but in order to keep things simple, I have decided not to go this route.
Battery Pack design v1
The charge board I am initially basing my design on is this 4S 14.8V/16.8V 40A Li-ion Lithium Battery 3.7V 18650 Charger BMS Protection Board:
It is advertised as being capable of delivering 40A at 14.8 V which would be close to 600W. With a 4S10P battery pack of 18650 having an average of 1500 mAh that would amount to just over 22 minutes at full discharge. That is a lot more than I have no plans to utilise. But I guess that is a nice head room to have.
My plan is currently to include a D1 Mini clone in the design of the Battery Pack. That way each unit can be monitered both with regard to current voltage and charge/discharge. It will also provide historical data about the battery pack. I am going to hook up a V/A sensor in order to collect this data. The D1 Mini’s will connect to a server or database via its WiFi. Perhaps I will write an Android App to connect to the battery via BlueTooth.