Build a DIY Powerbank with NiMH Batteries: Guidelines & Tips

[PhysicoRaj]

Hi everyone,
I wanted to make a powerbank for myself, to be able to charge smartphones (and if possible laptops), and last for as long as possible.
The reason I'm attached to Ni-MH batteries is because of this eneloop IV gen, claiming such attractive values for charge retention (70% after 5 years), maximum cycles (2100) and is relatively affordable.
Because longlife is my main aim as I stated.

Load requirements:

• I wish to have two USB Type-C outputs, each has to deliver a maximum of 4A at 5V (I don't actually plan to plug superfast charging devices, I would use recommended chargers for them).
• One USB Type-A output, 1.5A (DCP)
• And if feasible, a 20V 4A DC output for laptops.

So the power required = (5*4)*2 + (5*1.5) + (20*4)
Pmax = 128W
and if this figure looks crazy, omit the laptop charge, Pmin = 48W.

The input for charging the bank would be a micro USB female, I actually am not concerned very much about charging time. I want to be able to deliver 48W from a pack of eneloop cells. Each cell is 2000mAh, 1.2V and discharge current? Um, not very clear, seems like they'll give 2C easily but I need some help in deciding discharge current.
I know that unlike Li-ion cells, charging Ni-MH is challenging, and I'm not finding good Ni-MH charging modules easily. Lastly, the series-parallel configuration for the battery pack, I'm scratching my head. Each cell is 1.2v, but connecting them in series to increase voltage would make charging them difficult (as I'm aware Ni-MH require individual cell monitoring if charging in a pack), so connecting every cell in parallel seems to be the option, then boost 1.2v to required voltage for those USB supply controller boards.
48W at 1.2V would mean 40 amps. If each cell can discharge at 2C (4A) then 10 such cells in parallel could make the pack.

Correct me anywhere if I'm wrong and your advice would be of great help!
Thanks.

 

[anorlunda]

Sorry you're not getting other answers. Does this article help?

https://en.m.wikipedia.org/wiki/Nickel–cadmium_battery#Charging

 

[CWatters]

You should read up on the USB charging protocol. As I understand it the device being charged has to tell the charger what current it can handle, otherwise the charger should limit the current to something like 200mA. This means your charger (I mean your powerbank) has to be inteligent.

 

[PhysicoRaj]

Does this article help?
https://en.m.wikipedia.org/wiki/Nickel–cadmium_battery#Charging

Thanks. I had read on charging Nickel based cells and understood that it is quite different from charging Li-ion cells, w.r.t to charging current, voltage and overcharge or trickle charge. So I want to use charging circuits specifically for Ni-MH. All Ni-MH chargers I've seen charge each cell independently, but I want to know if a pack of Ni-MH cells all in parallel, can be charged without balance charging schemes, like most Li-ion based packs.

You should read up on the USB charging protocol. As I understand it the device being charged has to tell the charger what current it can handle, otherwise the charger should limit the current to something like 200mA. This means your charger (I mean your powerbank) has to be intelligent.

Yes, I had read http://composter.com.ua/documents/BC1.2_FINAL.pdf, so I'm planning to use USB power delivery circuit boards, that are pre-configured to work as Dedicated Charging Port controllers. Like inside a Type - C charger, I should find a controller that processes device requests and performs hardware handshakes. That'll make the ports intelligent, but to make the whole powerbank intelligent, there should be some understanding between these port controllers, or a coordinator that decides when and which port to shut down if the battery pack is unable to power all ports at the same time.

That's how I have been thinking.

Thanks.

 

[CWatters]

I don't quite understand why you are so keen on NiMH cells? I've found lithium cells fine in power tools. They retain charge for long periods meaning my drill is ready for that infrequent DIY job.

 

[anorlunda]

but I want to know if a pack of Ni-MH cells all in parallel, can be charged without balance charging schemes, like most Li-ion based packs.

Answer 1: When you put batteries in parallel, consider that one of the batteries is different than the others a little bit. It draws more (or less) current. That makes its temperature go up (down), and that changes the battery properties. If the change is in the direction to balance that battery with the others (negative feedback) your good. If the change is the other way (positive feedback) you get a runaway situation and the battery will be destroyed. Calculating net feedback effects is very complicated.

Answer 2: If all the NICAD chargers you see do batteries independently, there is a reason. Don't you think that if they could do is more simply, they would?

 

[donpacino]

Do you plan on buying a "smart charger" or building your own. If you are buying one, the answer to your question depends on the charger itself.

The easiest way to do it is have one charger, and use bleed resistors to balance the charge across the cells, you'll need to independently monitor the voltage across each cell to do this.

 

[donpacino]

That'll make the ports intelligent, but to make the whole powerbank intelligent, there should be some understanding between these port controllers, or a coordinator that decides when and which port to shut down if the battery pack is unable to power all ports at the same time.

That sounds like you need to implement custom logic. You will need to determine what the action is, and where the priorities are.

 

[PhysicoRaj]

I don't quite understand why you are so keen on NiMH cells? I've found lithium cells fine in power tools. They retain charge for long periods meaning my drill is ready for that infrequent DIY job.

The eneloop Ni-MH cells have higher maximum number of cycles (like 4 times more I read?) and are cheaper, so I thought they'll last longer for my kind of work. But if this turns out infeasible I might go with lithiums.

Answer 1: When you put batteries in parallel, consider that one of the batteries is different than the others a little bit. It draws more (or less) current. That makes its temperature go up (down), and that changes the battery properties. If the change is in the direction to balance that battery with the others (negative feedback) your good. If the change is the other way (positive feedback) you get a runaway situation and the battery will be destroyed. Calculating net feedback effects is very complicated.

Answer 2: If all the NICAD chargers you see do batteries independently, there is a reason. Don't you think that if they could do it more simply, they would?

I understand. So maybe I can rig up a circuit that connects cells in parallel while discharging and then switches to a configuration to charge cells independently when plugged in. But this will make simultaneous charging and discharging difficult (I think), unlike powerbanks nowadays.
The other option is to use lithium cells, and if I am right they can be charged as a whole pack of parallel cells.

Do you plan on buying a "smart charger" or building your own. If you are buying one, the answer to your question depends on the charger itself.

The easiest way to do it is have one charger, and use bleed resistors to balance the charge across the cells, you'll need to independently monitor the voltage across each cell to do this.

I am not planning to implement the USB protocols myself, I'll use some readily available DCP boards. Voltage monitoring was not in my plan since I thought it'd be complex. I think I'll reasearch more on this. Thanks for the input.

That sounds like you need to implement custom logic. You will need to determine what the action is, and where the priorities are.

Yes. That's what I thought at first, but I feel this can be achieved more easily? I'm not very sure but if I can set a current limit at the battery discharge, all the ports can work independently and still not fry the pack, only the devices will be charged slowly rather than being deprived of it.

Corrections and suggestions are welcome.
Thanks.

 

[Lord Crc]

48W at 1.2V would mean 40 amps. If each cell can discharge at 2C (4A) then 10 such cells in parallel could make the pack.

10 cells mighr be able to handle the max current, but you're forgetting the energy. A typical smartphone battery is around 3300mAh at 4.2V, so say 13Wh. Your pack with 10 cells is tops 24 Wh, so would barely be able to charge two almost-drained smart phones.

As for charging NiMh, have you looked into RC car battery chargers? They use sticks of 6 batteries in series for 7.2v, and I know they got NiNh sticks.

 

[PhysicoRaj]

10 cells mighr be able to handle the max current, but you're forgetting the energy. A typical smartphone battery is around 3300mAh at 4.2V, so say 13Wh. Your pack with 10 cells is tops 24 Wh, so would barely be able to charge two almost-drained smart phones.

As for charging NiMh, have you looked into RC car battery chargers? They use sticks of 6 batteries in series for 7.2v, and I know they got NiNh sticks.

Sorry for returning late. Thanks for these inputs, I'll look into them.

 

[sophiecentaur]

For someone who is about to launch out into DIY with Batteries (fairly high capacity, too), I have to bring in the obligatory warning about safety, appropriate fusing and heat dissipation requirements. Sorry if I'm telling you how to suck eggs but these things have to be said. (Unless you want to do your charging down at the bottom of the garden)

 

[Lord Crc]

Good point. I know from personal experience the 7.2v RC packs are quite capable of starting a fire...