How to upgrade to lithium?

[IsmAvatar]

I've got a few gadgets that run on AA/AAA batteries at around 5 volts (like 3-4 or batteries in series or 6 batteries with a parallel in the middle) which I use often enough to eat through batteries and I feel they'd be more convenient as rechargeable lithium. Lithium has the downside of being less than 4 volts, so I pair it with a TP4056/7 board (very small, very handy) and the the gadget is happy and I'm happy.

Now 18650s are very handy, except for their limited AH capacity, and this is where my question comes in. From what I understand, it's common to take multiple matched 18650s, pair them with a BMS of some sort, and power a project that way. I'd be down for that but I don't see anything that fits my use case - generally a BMS is a fair bit larger, and puts the cells in series (2S, 3S, etc) thus boosting the voltage well above 6v.

So how can I do 2P or 3P with ~5V discharge? I'm not having much luck finding a "2P BMS" (I'm thinking lithium batteries maybe don't play well in parallel - so would we use a 2S BMS + a voltage regulator? If so, any suggestions on where to find a small board with that?)
Or do I have to give up on 18650s and stick to pouch batteries?

Much appreciated, and forgive my ignorance. I'm still learning to navigate the electronics nomenclatures, but I'm having a blast with tinkering.

[wasedadoc]

Do not tinker with Lithium cells.

[Infraviolet]

Buy some Nickel Metal Hydride (NiMH) AA and AAA cells from a reputable seller (Farnell, Mouser, RS, digikey...). They can, for most devices, be readily swapped in for alkaline AA cells. NiMH won't catch fire the way Lithium Ion can. NiMH cells will take a long time to recharge (you can buy a wall wart unit to pop them in to when charging is needed), and they aren't great for devices that sit around rarely used for a long time, NiMH wil self-discharge over a period of months meaning you need to recharge them before using a device which has been unused for a while. But other than that, they're an easy replacement for alkaline cells in standard form factors.

[indeterminate]

Eneloop's solve the self discharge problem
https://www.panasonic.com/global/energy/products/eneloop/en/lineup/eneloop.html

[Whales]

Lithium has the downside of being less than 4 volts, so I pair it with a TP4056/7 board (very small, very handy) and the the gadget is happy and I'm happy.

A TP4056/7 will not fix that particular problem.  Those chips only charge a cell, they don't convert its voltage to other voltages.  Perhaps you're using a DC-DC converter?

I'm not having much luck finding a "2P BMS" (I'm thinking lithium batteries maybe don't play well in parallel

First of all it's possible to not have a BMS for both series and parallel battery Li-ion arrangements.  Possible, but typically a very bad idea (fires & explosion).

Most of the problems come from charging.  During charging you cannot:

* Take any one cell above its max rated voltage (typically 4.2V for most li-ion, but it varies)
* Charge a cell above its max safe charging current (typically expressed in units of 'C', like 1C)
* Continue charging if abnormalities are detected (eg one bank of batteries at too low of voltage, perhaps failed or disconnected)
* Charge outside certain cell temperature ranges.

Batteries in series will have slightly different capacities, so they will charge at different rates if you put the same current through all of them.  A BMS in this scenario makes sure that no one cell goes above 4.2V.  Ideally the BMS will also balance them, but cheaper ones don't bother and instead stop charging.

Batteries in parallel do not need balancing (they will all be at the same voltage) but you need to be careful about your charging currents.  If one or more cells disconnect (keep in mind many 18650s have hidden internal fuses) then you are now charging less cells with the full charging current.  This might be above the safe rating.  In this scenario some BMS can monitor cells individually, but most often the charging currents are low enough not to bother (unless you're dealing with a fast charger, but that's a whole other can of worms).

In practice: most small products using 18650s in parallel simply pretend that they are one big battery and charge at a current level that 1 cell could handle (just in case all the others fail).  A single TP4056/7 is going to be fine for this most of the time.

BUT

There are other hazards worth mentioning.

Connecting two 18650's in parallel is actually quite complicated:

* You MUST match both cells to be within a few mV of the same voltage before connecting them together.  Otherwise the higher charged cell will dump enormous amounts of current into the lower charged cell, possibly violating their ratings.  Typically you match cells by slowly discharing them individually with resistors to the same voltage.
* It is generally unsafe to install 18650's in parallel into spring sockets (or similar), because accidentally installing one cell backwards will also lead to insane and dangerous current dumping.  The cells may start fires within seconds.
* It is generally frowned upon to connect 18650's by soldering directly to their terminals.  In theory it's possible to damage the cells or cause them to enter thermal runaway (catch fire and explode) but I have not seen this yet.  They also have a lot of thermal mass so it's very hard to get solder to stick.
* The "correct" way is to spotweld metal strips to the batteries.  But that requires special equipment.


I might perhaps suggest a simpler solution.  Don't integrate a charger into your device, instead take the 18650s out and individually charge them (thus you won't need a BMS).  Rather than putting cells in parallel, put them in series and use an off the shelf buck converter to reduce the voltage (eliminating the challenges of assembling cells into parallel).  This also makes adding fuses simpler and easier to understand.

Also please, please use "protected" 18650s.  These are slightly longer because they have a little circuit board at one end that provides some extra protection.  It's not perfect, but it helps.

This is not an exhaustive safety guide, if you follow all of the above then it's still perfectly possible for people and property to be harmed or killed. Don't trust me, I'm a random guy on the internet, you need to consult multiple sources and make informed decisions.  NEVER try and charge a lithium primary cell (like a CR123) https://www.candlepowerforums.com/forums/smoke-and-fire-hot-cells-and-close-calls.107/

[Whales]

Oh and I'm also assuming that only you will be operating these devices.  If there is any chance then someone else (especially a small child) can gain access to the batteries in your devices then I recommend reconsidering.  AA batteries might be small but that gives them a LOT of advantages in terms of safety when compared with 18650s.  The other suggestions for rechargeable NiMH's made by other people here are very much worth considering.

What devices are you retrofitting?

[IsmAvatar]

That's really great info, thanks. I'll certainly remember those cautions.

The devices I'm retrofitting are a couple of portable lights and a label maker (yes, I'm the only person that uses any device I modify, but the warnings are worth restating, so I appreciate that). There have been a few instances where I went to use them in the field and dealing with yet another 6 dead batteries (sometimes with a lid with a screw...) in that situation was a nuisance, and I thought "if only I could just recharge this with a wall wart or a power cell".

[Whales]

Another thing to consider is whether or not you can modify these items to have a USB-C port and run directly off it (with no internal batteries).  You can buy little DC-DC boards with a USB-C input and an adjustable voltage output for only a few dollars.  Then carry off-the-shelf USB power banks as your batteries.  If you ever run out of power banks then a laptop or phone with USB-C can be used instead.

[J-R]

Not sure what you mean by 18650 has a limited AH capacity.  18650 is simply a form factor.  There are cells available in different capacities and lithium ion formulations and a company makes products in various cell shapes/sizes which primarily dictates it's capacity.

There are generational advancements or improvements for certain applications, but for Li-Ion pretty much the higher the capacity the lower the maximum discharge/charge rate, so selecting the right cell for your application is important.  For example, Molicel has two new 18650 cells.  One is rated for 35A discharge and 8.4A charge and has a capacity of 2,800mAh, while the other is rated for 10A/1.7A and has a capacity of 3,500mAh.


Assuming the device can support a slightly lower voltage, the top solution for replacing alkalines is the Eneloop AA/AAA cells and I recommend specifically the ones labelled with 70% capacity after 10 years.  There are different generations out there, but I think those are the best investment.  Locally, I get mine from Home Depot, but I've also found some deals on eBay.


If your device doesn't like NiMH's voltage curve, then an easy but slightly questionable solution is to go with the drop-in AA/AAA USB rechargeable lithium ion cells, such as from EBL.  I say questionable because of two main reasons: first, EBL isn't top tier like Eneloop, and second, the circuit necessary to drop the voltage can mean a poor shelf life and possibly noise injected into the device.  Minor issue is you can't charge them in the device.


Now we are down to the DIY conversion option, and I agree with the sentiment to use holders along with protected cells and charge outside the device.  The protection circuit will kick in to prevent over discharge especially in series configurations and it also should save you in the case of short circuits.

With regard to parallel connections, matching to within a few mV is not necessary, more like a hundred mV.  You can experiment with a typical lithium ion cell to see why this is the case.  Yes, lithium ion can supply a lot of current, but this requires a decently large voltage differential.  So if I try to charge an 18650 cell that is sitting at 3.5V with 1A, the voltage at the cell might be closer to 3.6V.  Similarly, if I pull 10A from a fully charged cell at 4.2V, the voltage will drop below 4V immediately.  So multiple cells in parallel will equalize fine if they are reasonably close.  But if you made the mistake of installing a fully charged 18650 in parallel with a fully discharged 18650, there could be trouble.  Although some abuse can be tolerated.  If both cells are at room temperature, then by the time either cell gets hot, the current flow will be lower and you won't have to call the fire department.

Here's a review of the aforementioned Molicel 2,800mAh cell where you can see the discharge curves: https://lygte-info.dk/review/batteries2012/Molicel%20INR18650-P28A%202800mAh%20%28Gray%29%20UK.html
At 30A it drops to 3.6V essentially immediately, then flattens out at just below 3.5V for a bit before beginning a further drop.


As a side comment, there is no perfect battery chemistry, and that is why NiMH, Li-Ion, lead acid, etc. are all very popular.  It depends greatly on the application.

[kjr18]

Well, for portable lights I would suggest Ni-Mh rechargeable cells, there is no need for lithium, probably space in battery compartment would be a problem, as for label maker, you could also try some, it depends on how low your power source could get before mentioned laber maker stops working and reports empty battery. It's because Ni-Mh has lower nominal voltage, than normal alkaline cell.

Another option Whales mentioned is to use DC-DC converter and typical power bank with sufficient power rating, but this is probably applicable only to label maker, as only this device would benefit from this.

[paulca]

Get on AliExpress.  You will find many a gizmo you need.

The most basic is to keep what you have, learn and parallel some 18650s and use a boost converter to get your 5V.

In fact if you search for a "Power bank module" it will provide you with a 5V charge circuit and a 5V discharge circuit from however many you want in parallel.

18650s and lithium cells come in different chemistries and different charge regimes.  Do NOT mix them.  The charge/discharge circuit must be paired to the correct chemistry.

The safety advice above is worth heeding.  A lot of it is ... debatable... however as you sound like you new to lithium I will refrain from telling you the "But.."

LiFePO4 cells are typically considered safer and longer lasting.  However they DO NOT match standard lithium cobolt and other "normal" chemistries.

If you don't want to parallel cells, AliExpress sell lithium cells up to 300Ah! 

[BeBuLamar]

Eneloop's solve the self discharge problem
https://www.panasonic.com/global/energy/products/eneloop/en/lineup/eneloop.html

Yes the Eneloop and Ladda from Ikea do have low self discharge. For application where the lower voltage of the NiMH is a problem you can try using AA Li-Ion rechargeables. These are 3.7V Li-Ion with built in regulator and BMS to deliver constant 1.5V for almost all of the discharge cycle.

[ledtester]

You might be interested in looking into Sodium-Ion battery technology.

[Refrigerator]

Buy some Nickel Metal Hydride (NiMH) AA and AAA cells from a reputable seller (Farnell, Mouser, RS, digikey...). They can, for most devices, be readily swapped in for alkaline AA cells. NiMH won't catch fire the way Lithium Ion can. NiMH cells will take a long time to recharge (you can buy a wall wart unit to pop them in to when charging is needed), and they aren't great for devices that sit around rarely used for a long time, NiMH wil self-discharge over a period of months meaning you need to recharge them before using a device which has been unused for a while. But other than that, they're an easy replacement for alkaline cells in standard form factors.

To be honest it's getting increasingly harder to make lithium cells catch fire. We have a little makerspace here, where we shoot lithium cells with a stapler sometimes and good name-brand cells are very hard to make runaway. Most of the time absolutely nothing happens even if the cell is charged to 4.5V.
Some 18650 cells have an over-pressure diaphragm, which will disconnect the cell internally if the pressure rises too high. Others have just a simple burst vent like an electrolytic capacitor.
But recently i got my hands on some no-name Chinese cells and they runaway in a spectacular fashion. When you shoot some staples in them nothing happens at first but as soon as they start hissing, they immediately runaway and explode, shooting their guts all over the place. They also smell absolutely rancid when they blow up.

But, other than shooting with a stapler, i haven't ever had a lithium cell fail over the past 10 years that i've been playing with them. So i consider them to be quite safe, assuming they're made by a reputable manufacturer (LG, Sony, Samsung and the like).

Ni-Mh cells are a great alternative to alkalines, and the self-discharge has been mostly solved over the years. Nimh cells also take a lot of abuse and they don't need any balancing. They're also highly available and very affordable.

[Phil1977]

To be honest it's getting increasingly harder to make lithium cells catch fire. We have a little makerspace here, where we shoot lithium cells with a stapler sometimes and good name-brand cells are very hard to make runaway. Most of the time absolutely nothing happens even if the cell is charged to 4.5V.
Some 18650 cells have an over-pressure diaphragm, which will disconnect the cell internally if the pressure rises too high. Others have just a simple burst vent like an electrolytic capacitor.
But recently i got my hands on some no-name Chinese cells and they runaway in a spectacular fashion. When you shoot some staples in them nothing happens at first but as soon as they start hissing, they immediately runaway and explode, shooting their guts all over the place. They also smell absolutely rancid when they blow up.

But, other than shooting with a stapler, i haven't ever had a lithium cell fail over the past 10 years that i've been playing with them. So i consider them to be quite safe, assuming they're made by a reputable manufacturer (LG, Sony, Samsung and the like).

Thanks! Probably most of us do not dare to write this, because better be safe than sorry etc. But the vast amount of Li-Ion-Pouches everywhere shows it´s good to be not too afraid of them.

Anyhow, be cautious. Not only because of cell explosions but also just because of high currents that can come from any low-ESR-battery. It´s easy to set something on fire if something little goes wrong, so please handle unprotected cells with lots of respect and add protection as soon as possible. Use protection only for protection and not for regular operation like ending the charge after 4.2V are reached.

The only time I´ve seen a Li-Ion-pouch go havoc was when a student tried to open a tablet computer and he drilled into the pouch. It vented quite a lot of things you don't want to breath and luckily an open window was near where the smoking tablet could be ejected. Since that event I try to have a sand-filled metal box next to me when working with these cells.

[Refrigerator]

To be honest it's getting increasingly harder to make lithium cells catch fire. We have a little makerspace here, where we shoot lithium cells with a stapler sometimes and good name-brand cells are very hard to make runaway. Most of the time absolutely nothing happens even if the cell is charged to 4.5V.
Some 18650 cells have an over-pressure diaphragm, which will disconnect the cell internally if the pressure rises too high. Others have just a simple burst vent like an electrolytic capacitor.
But recently i got my hands on some no-name Chinese cells and they runaway in a spectacular fashion. When you shoot some staples in them nothing happens at first but as soon as they start hissing, they immediately runaway and explode, shooting their guts all over the place. They also smell absolutely rancid when they blow up.

But, other than shooting with a stapler, i haven't ever had a lithium cell fail over the past 10 years that i've been playing with them. So i consider them to be quite safe, assuming they're made by a reputable manufacturer (LG, Sony, Samsung and the like).

Thanks! Probably most of us do not dare to write this, because better be safe than sorry etc. But the vast amount of Li-Ion-Pouches everywhere shows it´s good to be not too afraid of them.

Anyhow, be cautious. Not only because of cell explosions but also just because of high currents that can come from any low-ESR-battery. It´s easy to set something on fire if something little goes wrong, so please handle unprotected cells with lots of respect and add protection as soon as possible. Use protection only for protection and not for regular operation like ending the charge after 4.2V are reached.

The only time I´ve seen a Li-Ion-pouch go havoc was when a student tried to open a tablet computer and he drilled into the pouch. It vented quite a lot of things you don't want to breath and luckily an open window was near where the smoking tablet could be ejected. Since that event I try to have a sand-filled metal box next to me when working with these cells.

Shorting lithium cells can be very damaging to them. One time i accidentally shorted a lithium cell for just a split second and that was enough to cause internal damage that made the cell break down internally at 4V, which made it not possible to charge the cell fully. Below 4V it was still perfectly usable though, yet the damage was already done.

[Phil1977]

Shorting lithium cells can be very damaging to them. One time i accidentally shorted a lithium cell for just a split second and that was enough to cause internal damage that made the cell break down internally at 4V, which made it not possible to charge the cell fully. Below 4V it was still perfectly usable though, yet the damage was already done.

Sounds like quite a zombie-cell with severe damage to the separator. I would discard it ASAP.

[Paul T]

There is a new board in town which combines TP4056 + MT3608 adjustable dc boost circuit on one board.  Do a search for TP4056 + MT3608 on AliExpress.  I've bought a couple but have not implemented them yet.  Not sure how low drain they are when not running your device.

I see no problem with the path of using a 2S BMS so long as you use a CV-CC board as well to control your charging.  The output could be fed to a 5V LDO voltage regulator with low quiescent current draw (appx 5-10 uA) and a couple 1uF ceramic capacitors per the chip's datasheet application circuit.

Another option is to get one of the "tube" style single-18650 powerbanks which do not shut off at low current draw.  You can use either use it as-is, or pull out the circuitry and adapt to your application. 

I personally have these all over the place in my standalone 5V projects.  I carefully desolder the 5V USB-A output, and wire the output directly to my project and the cell connections to a LiPo of appropriate size.  Then I mount the board at the edge of the case so I can plug USB micro charger into it. 

I found a local store selling these tube powerbanks for $1.69 each.  I bought everything they had left!  I also collected a few as promotional items from businesses giving them away with their logon on them.

An example of such a powerbank and it's circuit was reviewed by Big Clive: https://youtu.be/TYBWNLyJmTM?si=_Y61kOnuVGIit-rr

And it doesn't have to be a tube type powerbank.  There are plenty of skinny 2500 - 5000 mAh powerbanks which also do not shut off at low current draw.  I've come across a few but it's never in the product specs.  You basically need to evaluate them yourself to see if they are suitable.

Attached is pictures of a "Rolodex" milliohmmeter with two 5V power sources (one for the LED display, one to supply constant current).  I zoomed in on one of the modules which is actually a tube powerbank chip converting 3.7V to 5V.  A DPDT switch connects both power sources simultaneously.

[Paul T]

To just drop down 1S to around 3V, it's possible to use the TP4056 to charge and a LDO voltage regulator.

I recently posted a few conversions for inexpensive multimeters.

If the cell is too big for the battery bay, and open space is available inside...  ANENG 613 Lithium Ion Conversion (Reversible)

If the cell and charger can fit inside the battery bay, everything done without opening the unit... Another 2AAA Lithium Ion Upgrade (UNI-T UT116C)

[David Hess]

As far as I know, there is nothing wrong with operating lithium cells in parallel, and in that case, the only BMS required is for protection against things like over-voltage or under-voltage, which is the same BMS required for a single cell.