18650 series discharge protection

[alxptt]

Hi! So I've been playing a lot with 18650 batteries, and I'm trying to design circuits that allow them to be used essentially like rechargeable alkalines. You take them out and plug them into a charger, then put them back in your thingy. The problem is, the load I'm using would just keep draining them well past the safety point of 3v or so, and also might discharge too quickly.

I'm aware that there are series balancer circuits on eBay -- but are they the best solution if I intend to simply take the batteries out to charge? I'd like something as simple as possible, and they include charging circuitry that I don't need. Can anyone point me in the right direction for finding or making a simple discharge protection circuit?

My additional question is, would I need balancing if all I'm doing is discharging the cells? That is, is there a risk of one cell being discharged faster than the others, or would they self-balance during discharge? If the latter is the case, is it possible that I could get away with a simple circuit that cuts the battery off at, say, 12V for a 4S battery pack? I realize this wouldn't have current limiting, but maybe I could achieve it some other way.

My goal is to be as simple as possible. Any inspiration would be very appreciated!

[IanB]

Why are you trying to reinvent a wheel? The market is completely saturated with the product you are contemplating. All the parts and assemblies you are talking about are made and produced in China at cost of a few cents per unit. What's more, consumer ready battery products are available from multiple sources that do everything you describe.

I really don't understand your goals here.

[alxptt]

Well, for one thing, I want to be able to understand how the components work. Simplifying my design as much as possible makes that more feasible, since I'm a beginner. Just slapping a generic module on might work, but honestly I'd rather use something I actually understand rather than a black box.

[sleemanj]

The standard China solution is the DW01 Protection IC and 8205 Dual Mosfet, this combination provides over discharge (~2.5v) over charge (~4.35v) over **discharge** current and short circuit protection.

Note that it does not protect against over charge current, just discharge, and also note that the typical over charge voltage cut off is quite high (there are different grades of DW01 with different cut offs, but from China who knows what you get).

Both ICs are readily available for a few cents on the usual chinese sources, not a complicated circuit.

Modules with them already made are also available, including circular ones which get welded and heat-shrunk onto 18650 cells to make them into protected 18650 cells.  Ideally that is the best way, put protection circuits right on your 18650 cells.

I doubt you can make a cheaper or simpler solution even if you had a couple of opamps spare in a circuit.

[nathanpc]

You could just buy a simple DW01 module on eBay and be done with it, but apparently you really want to get a better understanding of the circuit which is a great thing.

First of all, about the discharge balancing process you've mentioned. The cells are going to discharge at different rates because of the internal impedance of them, it is literally the same process that happens when charging, but in reverse this time, so when discharging a 4S pack to 12V your cells might be at different voltages and this problem just becomes aggravated when you charge the pack back up, so you do still need to balance your pack.

Now to the discharge protection circuit, you can learn a lot from datasheets of ICs you might be trying to replicate. For example, if you take a look at the DW01A datasheet you'll notice that they have provided a block diagram:



Of course it is a bit complex since the chip is trying to do multiple things, but if you focus on the left-hand side where the they are describing the over-discharge detector you can see that it's basically just a classic comparator circuit with a voltage reference. For example in this case you could use low-voltage opamps as comparators like the TLV272IP, or even better in this case, the MCP6292 which can work down to 2.4V. The circuit would be almost identical, but you could use a 1.8V standard red LED as a voltage reference and then just add a voltage divider to the cell so that 3V would equate to 1.8V at the input of the opamp. Then just use the output of the opamp to cut off the cell using a MOSFET.

EDIT: As a simple example for you here's a very simple solution to cut off the load when the cell voltage drops below 3V, this is a very simple circuit of how to use an N-Channel MOSFET on the low-side and a nice base to start your experiments with other circuit topologies.

[BravoV]

Well, for one thing, I want to be able to understand how the components work. Simplifying my design as much as possible makes that more feasible, since I'm a beginner. Just slapping a generic module on might work, but honestly I'd rather use something I actually understand rather than a black box.

As many recommendations, these are not blackbox, all you need is to read their datasheet to understand how it works.

As you admitted your self you're a beginner, not to rain your parade, I doubt you can improve much the current design, price wise and size wise.

Most sold are rounded to be placed on top of battery, or separated board like this below.
Just look at the size compared to a TO-220 sized body, imo, its already quite small.

[alxptt]

@nathanpc

Wow, thank you for all that detail! That's fantastic and I really appreciate it! Totally addressed all the initial operation questions I had. I'll definitely be playing with this project as op amp practice.

One question though -- since there is a 25k combined resistance in the resistor divider... it will have 0.24 mA of quiescent current draw, right? By my math, if I had a 3Ah cell, it would drain it in about 500 days. Is this correct? I am wondering (not for this project, but in the long term) if there is any way of creating a discharge protection that doesn't have a quiescent current draw, e.g. for an emergency system that's charged once and then left in my emergency backpack, or somesuch. How difficult would that be? (I'm assuming here that the Li Ion cells themselves don't have any significant self-discharge rate, since I'm told by the 18650 people that they don't. let me know if this is wrong!)

[KL27x]

Easy way to do this is use something like 3 pin voltage detector which are specifically made for something like this. There is in/power pin, ground pin, and output pin (which you can get in open collector or push/pull logic), and they come in voltage set point in every increment between 2.5-3.3ish V and more, to cover the cutoff voltage you are interested in using, drawing a gnat fart of current.

Take the output and connect that to the gate of a logic level Nfet. Connect the source to the ground terminal of the battery, and seal that over. The drain is now the negative terminal of the protected battery.

You can put these cells in series and each cell is individually protected. The battery will cut out when the weakest cell hits the limit. But you have to deal with the hysteresis or bounce with your main circuit, somehow.

This is two tiny 3 pin components. You technically don't need anything else but maybe a pulldown resistor on the FET gate. 3 components. Add pullup on the output pin of the voltage detector if you use the open collector variant. 4 components, max.

[alxptt]

Easy way to do this is use something like 3 pin voltage detector which are specifically made for something like this.

What type of voltage detector chip families should I be looking at here?

But you have to deal with the hysteresis or bounce with your main circuit, somehow.

What do you mean, bounce?

gnat fart of current

That is my favoritest thing ever and I'm stealing it now.

[KL27x]

What type of voltage detector chip families should I be looking at here?

I don't care about brands, but the one I have used is Microchip TC54VC/N xxxx. The VC is logic output. The VN is the open drain. I didn't care too much about cost, so I didn't look at a lot of other options. I just happened to find it one day while looking under "battery management" IC's on the Microchip website. Read the datasheet and as far as I can tell you would need to spend a lot of time and space and components cost/count to make something to compete with this and still end up drawing more current. As you can imagine, you are not the first to identify this application. There are many companies who have developed products, such as this, specifically to tackle the problem you are describing.

That is my favoritest thing ever and I'm stealing it now.

This is a common phrase around here. You would not be stealing it from me.

What do you mean, bounce?

The cell will rebound and the voltage will rise when the FET switches off. So this protects the cell from damage. It doesn't protect your device from browning out and turning back on/off over and over.