Some Questions About 18650 Cells and Designing A Battery

[NinjaKitten]

Hey all, new to the forum and had some quetions.

I'm planning on making a 12V battery from used cells so I can power some LED light strips. Getting ready for hurricane season. But during my research for everything I had some questions and confusion that popped up and I haven't been able to find a reliable answer for some stuff. So I figured I'd ask about it all here.

Balancing Salvaged Cells Before Building The Pack, and Selecting Cells To Use

• Balancing Cells Correctly – Do I just hook them up to a charger, make sure they don’t get too hot, and match the ones that all charge up to 4.2 volts?
• Checking Capacity – Do I just use a cell charger to charge, discharge, then charge them again to get the reading of the mah capacity? How CLOSE would I match usable cells together? Would 2000mah and 2050mah (50 mah difference) be acceptable?
• Cycle Count – How would you know that you can use a Cell without knowing its cycle count? Is that where the rated vs actual capacity comes into play? Would you recommend only using cells that are at least 85% of their rated capacity only? OR is there a better general idea I should go by.
• Mixing Different Sourced Batteries – Generally I know you should only match batteries with the same brand/model in a battery pack. But what if the specs of the 2 different batteries are same, just different brands? Would you recommend using used laptop batteries mixed in with old power tool batteries? Or keep it strictly the same brand/model no matter what?
• Self Discharging – After the cells are charged, how long do you let them sit before checking voltage again to make sure they aren’t self discharging? And is this necessary for low drain applications (battery for 12v light strip – lower amps) instead of high drain applications (ebike – higher amps)?
• Internal Resistance – How would you check this accurately? I know there are tools you can by to do so, but I’ve read that Internal resistance can change depending on the state of charge, full, half way, empty. Also can be different from the make, type of cell, etc. Is it generally speaking good enough to charge the battery to full 4.2v then check it? And when checking it, what’s a good general number for used cells? Anything under 50? Of course a good number to look for would depend on the application I’m assuming. Higher drain cells would be lower from my understanding, and low drain cells would be generally a bit higher.

• To summarize, so I can have a general idea of what to go by, does this seem like the necessary matching criteria to pair cells with?
  • Voltage must be at 4.2v when full, should not get too hot when checking capacity
  • Voltage should stay the same with very little movement if any after sitting for a few days to a week.
  • Capacity mah rating needs to be within a 50mah difference between cells
  • Internal resistance of each cell, (depending on application) should be no more than ~50 +/- mOhm
  • Batteries should be match with the same brand/model to avoid major differences
  • Amp Specs need to match (Max Amp Discharge/Recharge)

Thickness Need For Nickel Strip Series & Parallel Connections

From what I researched, a 0.15mm strip can handle about 7a. So if my math is right….

Say I have a 3S5P 18650 battery. Each battery is rated for 15Amp Discharge, 5Amp Charging, and 2500mah. Each series connection (5) x 15 Amps, means this battery can supply a max of 75 Amps Continuous Load. Or 75a x 12v = 900w maximum load.

However the actual battery capacity will be 2.5ah x 5 = 12.5ah or 150wh (12v x12.5ah)

Anyway, I would need to use 2 of those 0.15mm strips on each series & parallel connection in order to handle 15A from each battery, so the full 75 Amps is distributed evenly across the entire pack. (14a is close enough to 15a for this example )

Of course, I wouldn’t plan on pulling that many amps from such a small battery pack. But I just want to make sure I’m doing the math right more than anything.

For the charging, I would be able to use a 12.6v MAX 5Amp (since each cell is rated for a max of 5A) charger, I think.



Using the correct AWG wire for the main Positive/Negative Leads & Connectors, & Selecting a BMS

After doing the math..

The main leads from the battery for positive and negative would need to be at least a 4AWG wire (which can handle 85Amps) and I would need a BSM rated for more than 75 Amps.

Now my next question about this part. Would attaching the main leads ONLY to the ends of the positive and negative part of the battery be enough for that full 75amp load? I’ve seen people have the wire itself soldered all the way down the entire length of the main positive and negative ends on top of the batteries in parallel.

I’m thinking that since that’s where the full load will leave the battery, the nickel strips on those two parts will need that extra wire to handle the current and load, since the connections themselves can only handle 15amps max continuous? So running the leads down that path will help with the flow of current? That and bigger batteries with higher loads should be setup that way, otherwise more of the cells towards the top will be under most of that current load….

Correct me if I’m wrong though! But this is what I mean by having the wires go down the path –

https://imgur.com/6OBkobp

But in essence, I should be able to use a BMS and wires for whatever AMPS that match (or ideally is rated for a bit more) the power needed for the demand, so long as it doesn’t exceed 75 amps in this example.

I wouldn’t necessarily NEED to have a 80 amp BMS or 4 awg wire unless I need to pull the max power I can from this battery.



Securing & Wrapping The Battery Pack

My plan for my 12v battery in terms of wrapping it and finishing it up:

Use plastic 18650 spacers for each cell

Use the ring terminal insulators on the positive ends

After spot welding nickel strips on and soldering my main leads for the positive and negative ends… wrap the whole thing in some Kapton tape

Wrap it in some heat resistant foam on the outside as well for some added protection and then heatsink it all together.

If anyone has opinions or suggestions to properly insulate and wrap a battery though, I’m all ears!



After watching so many YouTube videos, and reading posts on reddit. I think I finally did enough research to get started. I don’t plan on making an ebike battery anytime soon. And if I do, I’ll use new cells instead of salvaged ones. I only plan on making a decent sized 12v battery to power some lights for hours since we got the hurricane storm coming up. But I couldn’t help but ask these questions since I couldn’t find any in depth answers.

Preciate any help or opinions!

 

[calab]

There is a lot there, its to late to organize my thoughts in a response to your lengthy, informative, easy to read (thank you) thread. The summarize part was especially appreciated.

Building the pack - Voltages of each 18650 needs to be all the same within a very small margin before you start tab welding it all together.
If the voltages are way off then you damage them. Some people will buy chargers to stick individual 18650's in and charge to whatever voltage you want, you can buy 18650 holders off fasttech or dx.com to have an easy way to charge bare 18650's.

Cycle Count - This is why you buy from a good company, then your not guessing.

Buying used stuff you have no clue how the battery or cells were handled, stored or used. I'd try to keep the 2 different cells all together within a single pack build and not mix them up, however I read that a random mix up can be a solid option.

Used cells - Group the cells on the discharge test. Have a healthy pile, a questionable pile and a throw out pile. Like a used 18650 at 1.00v, but some people slowly bring those back to life and use them. They'd have to charge those at a very low rate, 0.1a charge rate, then do a capacity test.

Sit/Rest cells - You just need the resting voltage after some time has past, say 5 minutes while you work on other stuff. The bms is what can drain the battery. How long would it take for a new 18650 or healthy used 18650 to drain on its own, a very long time. If the Internal Resistance is high, they would stick out from the others after they sit for months.

 

[NinjaKitten]

There is a lot there, its to late to organize my thoughts in a response to your lengthy, informative, easy to read (thank you) thread. The summarize part was especially appreciated.

Building the pack - Voltages of each 18650 needs to be all the same within a very small margin before you start tab welding it all together.
If the voltages are way off then you damage them. Some people will buy chargers to stick individual 18650's in and charge to whatever voltage you want, you can buy 18650 holders off fasttech or dx.com to have an easy way to charge bare 18650's.

Cycle Count - This is why you buy from a good company, then your not guessing.

Buying used stuff you have no clue how the battery or cells were handled, stored or used. I'd try to keep the 2 different cells all together within a single pack build and not mix them up, however I read that a random mix up can be a solid option.

Used cells - Group the cells on the discharge test. Have a healthy pile, a questionable pile and a throw out pile. Like a used 18650 at 1.00v, but some people slowly bring those back to life and use them. They'd have to charge those at a very low rate, 0.1a charge rate, then do a capacity test.

Sit/Rest cells - You just need the resting voltage after some time has past, say 5 minutes while you work on other stuff. The bms is what can drain the battery. How long would it take for a new 18650 or healthy used 18650 to drain on its own, a very long time. If the Internal Resistance is high, they would stick out from the others after they sit for months.

Thank you for the reply. I appreciate that! Was doing more research and asking the same questions on a few more websites and everything is matching up information/advice wise.

Just still confused on the current distribution between parallel groups, connections and series connections still. But I made another (shorter) post about that.

 

[rafalg]

Hi, regarding the current distribution - i think your way of thinking is correct here, thick main leads at both ends of batter are very reasonable.
And some additional thoughts about construction of bigger packs from salvaged cells:
- have some way of testing the cells. Volt meter is not enough, would be good if you can estimate the capacity and self discharge. One faulty cell can waste your work. I'm buying used cells in already tested batches to avoid the tedious procedure, but it's not perfect, sometimes an 'almost good' one will slip thru and will cause problems later.
- mixing different types of cells is possible, but keep every section the same capacity. So for example use 3x3000 mAh + 3x2500 mAh in every 6P section
- equalizing the voltage before connecting cells in parallel. I'm bit lazy with that and try to get them all within 0.1V difference (but sometimes 0.15V goes as well). Did not see any damage so far but who knows.
- think about some way of diagnosing and servicing the battery. Used cells aren't as predictable as new ones and there's a bigger chance some of them will degrade fast. How to find such cell and replace it? My idea was to split the battery into separate sections, for example recently built 10S6P battery was split into 5 separate sections of 2S6P, connected with wago connectors. Each section is built with plastic holders, and all battery is held together in a plastic box, so it's all reasonably fixed and secured against rattling and moving.
The idea is that i can take out a faulty section and either fix it or replace it with a new one - easier than taking apart the whole battery. But for 12V battery it might be an overkill
- balancing the voltages. Some BMSes have the balancing function, but it's usually working with a very low current. For bigger capacity packs the balancing current needs to be increased. I'm using some cheap balancer boards with extra resistors added. Then it works reasonably (but it creates some noticeable amount of heat when balancer is working). But if it's just 3S battery maybe it can be balanced with a proper charger instead.

 

[Diggs]

Balancing Cells Correctly – Do I just hook them up to a charger, make sure they don’t get too hot, and match the ones that all charge up to 4.2 volts?

You need to pick up an 18650 cell tester that will tell you the capacity and approx resistance. They are cheap and readily available. Then you can sort cells and build your packs with each parallel group being equal. Mixing same spec batteries is not a big issue. The problem is when a 3,000 mah cell (measuring 2100 mah tested) is used with a 2200 mah cell that is measuring 2100 mh. Although they test the same. They shouldn't be used together.

 

[Bullfrog]

Thank you for the reply. I appreciate that! Was doing more research and asking the same questions on a few more websites and everything is matching up information/advice wise.

Just still confused on the current distribution between parallel groups, connections and series connections still. But I made another (shorter) post about that.

In a series circuit amperage is constant.

In parallel circuits amperage is additive.

Voltage is the opposite of amperage i.e. additive in a series circuit and constant for parallel circuits.