How to build a individual 18650 battery pack, Example 12S8P

[solar.g]

Danger with soldering, as stated elsewhere by more experience ones than I, is that heat from soldering iron soaks into the cell and overheats the electrolyte (and other chemistry) just under the end you're soldering.
Very difficult to manage/limit the heat-soak from the soldering iron through the metal into the chemistry.
Spot-welding is much shorter-duration and more manageable from a heat soak-through point of view.
Some post-mortems on cells done both ways would be nice (thinking of vertical sections using a buzz-saw, with photos, to see visual effect of heat-soak with both methods.

 

[t3sla]

So we have seen this being stated numerous times, and I don't doubt that the cells take damage from the heat.
But in reality, how much of a problem is it? What negative effects can one expect to a soldered cell compared to a spotwelded one? (With this I mean soldered with good tools and minimal heat exposure)
I am evaluating if I am going to solder my 200cells or build a spotwelder. But I want the cards on the table if anyone have an idea.

Totally depends on equipment, certain applications call for solder and one o the most extravagant methods is a heated block that presses down momentarily. Cells are acting like a heat-sink and the cool down period isn't exactly small, the key would be to use a temp control iron that you can regulate and can maintain it's temperature (which decreases when in contact with the cell)

This is why welding rocks, a quick short burst of concentrated energy.

For example your typical run of the mill direct wall-plug soldering iron heat gets soaked like a sponge, the truth is you don't actually need a 200W iron but rather one that can maintain constant tempreture when soldering. I presume like all soldering surface prep help, so a light scuff light scuff should reduce bond time.

Saying that I'd still rather have a ghetto soldered 18650 pack with good cells over HK lipo myster cells.
Anyway that's just my opinion.....man

 

[madin88]

It's not inadequate from an electrical connection stand point, it's just ensuring every cell with a soldered joint to them IS damaged internally. The solvents decompose and introduce gas at best when holding the case at >150C to get the solder to wet the can, so you know you've damaged them just building it.

You say the cell IS damanged but you don't say in which way. Capacity, resistance, calendar life? You should explain!
Personally i don't think it's as bad as you claim. With good soldering tin and good iron you need less than 2sec to wet the can (if you know how to proceed).
In former NiMh / NiCD times cells also had been soldered - mostly for special high power packs - without problems if it was done the right way. One big reason is soldered packs have lower resistance than nickel band spot welded ones. This causes less heat at high discharge rates and less heat will extend lifespan.

You guys are aware that the can's are all the negative terminals right? And those in series are touching can's with other potentials and just relying on a very thing bit of shrink rubbing on shrink where the edges of cells contact each-other to prevent fire right? This is why even on the very shitty ebay 18650 packs they put hot-glue and a layer of fish paper or even card-stock paper or something between the rows of series groups.

Shrink does NOT chafe on shrink forever before it wears through, I can assure you of this. Likewise, unfused balance tap leads of that heavy gauge of wire cause some excitement.

In most battery packs there is not only shrink on shrink. Every cell is glued to each other so there is nearly zero rubbing in the pack, and if there is some rubbing it must also chafe through the glue.

 

[dnmun]

It's not inadequate from an electrical connection stand point, it's just ensuring every cell with a soldered joint to them IS damaged internally. The solvents decompose and introduce gas at best when holding the case at >150C to get the solder to wet the can, so you know you've damaged them just building it.

You say the cell IS damanged but you don't say in which way. Capacity, resistance, calendar life? You should explain!
Personally i don't think it's as bad as you claim. With good soldering tin and good iron you need less than 2sec to wet the can (if you know how to proceed).
In former NiMh / NiCD times cells also had been soldered - mostly for special high power packs - without problems if it was done the right way. One big reason is soldered packs have lower resistance than nickel band spot welded ones. This causes less heat at high discharge rates and less heat will extend lifespan.

You guys are aware that the can's are all the negative terminals right? And those in series are touching can's with other potentials and just relying on a very thing bit of shrink rubbing on shrink where the edges of cells contact each-other to prevent fire right? This is why even on the very shitty ebay 18650 packs they put hot-glue and a layer of fish paper or even card-stock paper or something between the rows of series groups.

Shrink does NOT chafe on shrink forever before it wears through, I can assure you of this. Likewise, unfused balance tap leads of that heavy gauge of wire cause some excitement.

In most battery packs there is not only shrink on shrink. Every cell is glued to each other so there is nearly zero rubbing in the pack, and if there is some rubbing it must also chafe through the glue.

if the soldering does create a lower internal resistance in the pack you should be able to demonstrate it. this battery seems to have 48mR internal resistance which means each cylinder has about 32mR resistance to get to the 48mR total pack resistance. it would help to know what the original cylinder had for internal resistance before it was soldered. knowing that would help your argument a lot.

there is another guy who has a soldered laptop pack and he has 178mR internal resistance for a 16S pack that i think is 12P wide. that implies 133mR for each cylinder after soldering. that is higher than 10 year old ping pouches.

but this pack is built with .2C laptop cells so it is built to only deliver 260mA from each cylinder, 2A total, so i don't know how much that would determine the internal resistance on discharge. it would help if someone would honestly untangle these fundamental elements of the battery behavior.

 

[madin88]

Ok, lets calculate a pack 10s6p 18650 cell:

lets presume 1mm² joints. I batt = 60A (so 10A per strand).
connector between cells about 20mm (so 9 connectors 0,18m).

P loss = R x I²

rho Copper = 0,017 Ohm x mm²/m x 0,18m = 3 mOhm x 10² x 6 = 1,8W heat
rho Nickel = 0,062 Ohm x mm²/m x 0,18m = 11,1 mOhm x 10² x 6 = 6,66w heat -> plus some more heat due to the tiny spot welded points (i don't know how to calculate this but i know the heat is there)

iron band which is also used for spot welding has even more poor 0,098 Ohm x mm²/m

 

[dnmun]

i did understand your argument about the heat generated in the spot weld because of the current density in the contact. i think that might be part of why the manufacturer recommends the .2C because of the internal ultrasonic welding of the electrode foils to the current collector inside the cylinder. or they may recommend .2C because the dendrite growth on the anode under higher discharge rates leads to separator damage because of the separator material they are using.

what would be useful instead of theoretical arguments is for someone to actually establish what the internal resistance is of the cylinder as delivered by samsung. samsung should be able to provide a real number for the internal resistance even though they do not post it on the data sheet.

then after the pack is built up it could be discharged into a load. at the 1C rate it would deliver 10.4A and at 3C it would deliver about 31A.

you should be able to create a dummy load that will draw the 31A and then measure the voltage sag on each cell at 1C, 2C, and 3C to determine the internal resistance of each cell in the pack after it has been built.

to me, this is the most important number to determine. you may have other opinions.

if we can find someone who has built a laptop cell pack using a spot welder then we could try to get the same numbers and even better try to examine the spot welds with a thermal imaging camera to see if we can see how much hotter the weld spot is than the surrounding nickel conductor and cylinder contact.

i think that would help to establish at least part of the uncertainty of soldering these large gauge wires to the pretinned spot on the cell. it should be relatively easy for you guys to establish the internal resistance since you already have the packs built so the only uncertainty is what the initial internal resistance of the can is coming from the formation charge at the factory. samsung should provide you with that number if you spent this much money on them.

 

[madin88]

then after the pack is built up it could be discharged into a load. at the 1C rate it would deliver 10.4A and at 3C it would deliver about 31A.

you should be able to create a dummy load that will draw the 31A and then measure the voltage sag on each cell at 1C, 2C, and 3C to determine the internal resistance of each cell in the pack after it has been built.

I know what you mean, but if you measure a built up pack you have nickel vs copper and solder vs spot weld joints and this takes some influence to the measurement.
Measuring resistance of single cells is not very easy and you need some very good calibrated equipment.
I don't have the time and money to do this. I even don't have a spot welder.

the best way woud be to measure every single cell before soldering and after with something like this:

both cells should have same temperature and same volts.
It also could be better to put on every cell 5 similar cycles before this test, because as far as i know new cells have some chemical stuff added to the electrolyte to extend calendar life (store life). After some cycles this electrolyt stuff disappears and the cell gives the full capacity and power. I know this from LiPo, but i'm not sure if this is so with other types of Li cells.

 

[Hillhater]

..... i think that might be part of why the manufacturer recommends the .2C because of the internal ultrasonic welding of the electrode foils to the current collector inside the cylinder. or they may recommend .2C because the dendrite growth on the anode under higher discharge rates leads to separator damage because of the separator material they are using..

That "0.2C" rating seems to be quoted by several battery makers in their official spec sheets.
I have a suspicion that it is some kind of "industry standard" for use in establishing the cell capacity when testing,... rather than an actual recommended normal duty discharge rating.
You can even see that 0.2C quoted for "high drain" cells which we know can handle 2C - 5C continuous discharge.
As an aside, i also have my suspicions that spot welded foil tabs may not be as desirable as good "wetted" solder braid connections, for high current situations on these newer "high drain" cells.

 

[madin88]

One good thing to prevent damage to the cell (we do not know what kind of damage for so far) is to spot weld little nickel tabs on the cans (with lots of spots for good resistance) and than wet them with solder. Over the spot welded points there will go much less heat inside the cell.
The positive side of the battery is anyway no real problem..

 

[ridethelightning]

i need some advice regarding balence taps and 1420i

if charging say, an 11s pack, of subpacks of 5s and 6s(connected in series), there are 2 balence plugs to hyperion. 1 with 6wires, 1 with 7wires respectively(if im right) including ground for each sub pack

as the sub packs are connected in series while charging,

will i need the ground wire between s5 and s6?

help much appreciated, as im all new to balence chargers

 

[ecross]

i need some advice regarding balence taps and 1420i

if charging say, an 11s pack, of subpacks of 5s and 6s(connected in series), there are 2 balence plugs to hyperion. 1 with 6wires, 1 with 7wires respectively(if im right) including ground for each sub pack

as the sub packs are connected in series while charging,

will i need the ground wire between s5 and s6?

help much appreciated, as im all new to balence chargers

Hi ridethelightning,

you have only 1 x ground, it is the minus from pack "1".

-for my split packs I have solder a "V" cabel to connect 2 split packs to my Hyperion 1420.

-please send more infos to your packs and I can send more input

BR
ecross

 

[ridethelightning]

thanks ecross!

my pack is actually 22s10p samsung 20r li ion.

it is made up of 2x 6s and 2x5s sub packs.

i have wired the 5 and 6spacks together, so it can be charged as 2 x11s packs.

so from what you said my balence plugs to hyperion shoud be- (plug 1) ground,1s,2s,3s,4s,5s. and (plug2)6s,7s,8s,9s,10s,11s .




i have 2 hyperion 1420is, however no psu yet... not sure what to go with. i was thinking id start by just charging 1 11s pack at a time till i get the hang of it and get

enough psu.s
i could use my adaptto max e controller to bulk charge, but i guesse balence is best first and i dont have the adaptto bms system yet.

other pieces of the puzzle i still need to find out are:

1.what recomended max voltage to charge these cells to,in bulk and ballence charge.
2.max charge rate
3.lvc voltage.



please extrapolate about your v cable, is it for paralelle ballence charging 2 split packs?

 

[madin88]

other pieces of the puzzle i still need to find out are:

1.what recomended max voltage to charge these cells to,in bulk and ballence charge.
2.max charge rate
3.lvc voltage.

i hope its ok for ecross if i comment:

first: read the datasheet for samsung 20r cells.

to 1: for longer lifespan i would charge to 4,1V per cell. This will cause in about 10% less capacity, but will more than double the useable watthours from the battery over whole lifespan.
to 2: with 10p you can set your 1420 charger to max output (this will be about 1C charge with your samsung 20r cells)
to 3: i would not discharge below 3,3V (leave 10-20% capacity in there for longer lifespan)

do you have some pics of the packs?

 

[ridethelightning]

thnx madin
i realise the data sheet would be a good start, however it may not tell you stuff about increasing lifespan etc, exactly like what you just wrote. thanks heaps.


as for pics, id love to post some, however iv had trouble getting the E.s to accept the file type i have even after iv downsized the files.
i definately want to get my build up there at some point though. im not all that IT savvty to be honest. iv kinda chucked i in the too-hard basket for the moment, but praps its time for another shot at it. might be time to talk to my bro who works as a programmer

my pack came from supower in tabwelded rows of 10p with cell spacers. they did a great job.

 

[ecross]

other pieces of the puzzle i still need to find out are:

1.what recomended max voltage to charge these cells to,in bulk and ballence charge.
2.max charge rate
3.lvc voltage.

i hope its ok for ecross if i comment:

first: read the datasheet for samsung 20r cells.

to 1: for longer lifespan i would charge to 4,1V per cell. This will cause in about 10% less capacity, but will more than double the useable watthours from the battery over whole lifespan.
to 2: with 10p you can set your 1420 charger to max output (this will be about 1C charge with your samsung 20r cells)
to 3: i would not discharge below 3,3V (leave 10-20% capacity in there for longer lifespan)

do you have some pics of the packs?

Hi madin88,

no prob, everwhere can comment everything here, but I told you into the past, please, please dont`t comment here all with your little general knowledge.
Please use your theread for your pro general knowledge after 1 E-Bike building.

But certainly you can help with the special V-connector for ridethelightning, please send some infos how do this connector and send also a pic and more, this is very helpfull.

 

[Spacey]

Ecross, that was a little bit rude dude. madin88's advice is not completely wrong at all...and he was very polite.

 

[madin88]

Hi madin88,

no prob, everwhere can comment everything here, but I told you into the past, please, please dont`t comment here all with your little general knowledge.
Please use your theread for your pro general knowledge after 1 E-Bike building.

But certainly you can help with the special V-connector for ridethelightning, please send some infos how do this connector and send also a pic and more, this is very helpfull.

yes i only build one e-bike, but why do you think that this e-bike is the only experience i have regarding to batteries, chargers etc.?
You make no friends here if you continue to comment like this..

i need some advice regarding balence taps and 1420i

if charging say, an 11s pack, of subpacks of 5s and 6s(connected in series), there are 2 balence plugs to hyperion. 1 with 6wires, 1 with 7wires respectively(if im right) including ground for each sub pack

as the sub packs are connected in series while charging,

will i need the ground wire between s5 and s6?

help much appreciated, as im all new to balence chargers

no balance ground wire needed. just make it like this:

first: battery 1 (6s) to charger negative (-) and to balance port A (A= LOW on display)
second: battery 2 (5s) to charger positive (+) and to balance port B (B= HIGH on display)
third: connect battery 1 positive to battery 2 negative in series.
start charging

always connect the balance plug of battery, which is connect to charger negative (-), to port A. never to port B!

there is no need to solder a v-cable for balance wires. Just for the main connectors if they are built as a pair.
OK?
It is also explained in the manual "charging split packs" :wink:

 

[ridethelightning]

Ecross, that was a little bit rude dude. madin88's advice is not completely wrong at all...and he was very polite.

at least not as rude as your profile picture, spacey, but i love it all the same...bit of spice to colour up the forum....

im most greatful for all the advice i get,very valuable to me, thanx again all!

 

[dnmun]

i think spacey has a nice ass and don't think it rude for her to show it off. but there is no place to click on it to enlarge it.

no need to be cross with ecross. he spent 30 years learning how to make batteries so he has the seniority on all of us in that.

 

[ecross]

Hi madin88,

no prob, everwhere can comment everything here, but I told you into the past, please, please dont`t comment here all with your little general knowledge.
Please use your theread for your pro general knowledge after 1 E-Bike building.

But certainly you can help with the special V-connector for ridethelightning, please send some infos how do this connector and send also a pic and more, this is very helpfull.

yes i only build one e-bike, but why do you think that this e-bike is the only experience i have regarding to batteries, chargers etc.?
You make no friends here if you continue to comment like this..

i need some advice regarding balence taps and 1420i

if charging say, an 11s pack, of subpacks of 5s and 6s(connected in series), there are 2 balence plugs to hyperion. 1 with 6wires, 1 with 7wires respectively(if im right) including ground for each sub pack

as the sub packs are connected in series while charging,

will i need the ground wire between s5 and s6?

help much appreciated, as im all new to balence chargers

no balance ground wire needed. just make it like this:

first: battery 1 (6s) to charger negative (-) and to balance port A (A= LOW on display)
second: battery 2 (5s) to charger positive (+) and to balance port B (B= HIGH on display)
third: connect battery 1 positive to battery 2 negative in series.
start charging

always connect the balance plug of battery, which is connect to charger negative (-), to port A. never to port B!

there is no need to solder a v-cable for balance wires. Just for the main connectors if they are built as a pair.
OK?
It is also explained in the manual "charging split packs" :wink:

Hi,

sorry my buddy but from time to time I must say the truth to ES-Members like you!

Why....because your Manual for ridethelightning is wrong. absolutely wrong, you have no Hyperion 1420, right, or! and you writing false statements here. And yes, I need no friend like you, because you are a little troll.

Please do not enter in future my threads. Open your own threads with your friends. This forum is big enough for all.

I have no problem to help ES-User but I hate the type of the self-made pros like you...after 1 build and so little experience like you.

 

[ecross]

Ecross, that was a little bit rude dude. madin88's advice is not completely wrong at all...and he was very polite.

at least not as rude as your profile picture, spacey, but i love it all the same...bit of spice to colour up the forum....

im most greatful for all the advice i get,very valuable to me, thanx again all!

Thanks ridethelightning for the friendly words, please do not follow the false manual from madin88, please send some more infos from your balancer cables setup so I can send you the right infos. Also a pic is welcome, because a pic is better then many words. I have the great 1420 over 3 Years and I have min. 4 pcs into my lab. The 1420 charger is a good basic charger like the Graupner Ultramat 18.

 

[ecross]

i think spacey has a nice ass and don't think it rude for her to show it off. but there is no place to click on it to enlarge it.

no need to be cross with ecross. he spent 30 years learning how to make batteries so he has the seniority on all of us in that.

-thanks also for the friendly words, I'm always looking to give a helping hand specialty for the great and save 18650 cells from Sanyo, Sony, Panasonic, Samsung because I'm involved into a R&D department for this cells+corepacks like into Power-Tools and so on....

 

[madin88]

Hi,

sorry my buddy but from time to time I must say the truth to ES-Members like you!

Why....because your Manual for ridethelightning is wrong. absolutely wrong, you have no Hyperion 1420, right, or! and you writing false statements here. And yes, I need no friend like you, because you are a little troll.

Please do not enter in future my threads. Open your own threads with your friends. This forum is big enough for all.

I have no problem to help ES-User but I hate the type of the self-made pros like you...after 1 build and so little experience like you.

whats wrong with you guy?

A bit more respect pls, or i will contact a moderator. I don't bother if this is your thread or not. Your words are not OK.

About the question from ridethelightning: I have illustrated it correct, so don't say i am wrong!
There is no need to solder a balancer v-cable for his 5s and 6s packs.

http://www.hyperion-eu.com/downloads/eos/manuals/1420inet_connection_guide.pdf

what do you say now as an owner of this charger?

 

[Spacey]

Guys, I just love the fact that this forum is full of polite people...even if they do fantasize over my avatar's gorgeous behind

But just wanted to point out that there is no place for rudeness, I admit I had only read that page of the thread and agreed with statements like:

Not charging and discharging the cells fully will lengthen their life.

Heat soldering cells together will damage them.

Those statements I agree with totally and have experienced this myself (hence why I bought a battery welder ages ago). But let's keep it civil and have respect for each other.....can I get a group hug....I'll bring my avatar

 

[ecross]

Hi,

sorry my buddy but from time to time I must say the truth to ES-Members like you!

Why....because your Manual for ridethelightning is wrong. absolutely wrong, you have no Hyperion 1420, right, or! and you writing false statements here. And yes, I need no friend like you, because you are a little troll.

Please do not enter in future my threads. Open your own threads with your friends. This forum is big enough for all.

I have no problem to help ES-User but I hate the type of the self-made pros like you...after 1 build and so little experience like you.

whats wrong with you guy?

A bit more respect pls, or i will contact a moderator. I don't bother if this is your thread or not. Your words are not OK.

About the question from ridethelightning: I have illustrated it correct, so don't say i am wrong!
There is no need to solder a balancer v-cable for his 5s and 6s packs.

http://www.hyperion-eu.com/downloads/eos/manuals/1420inet_connection_guide.pdf

what do you say now as an owner of this charger?

-I'm lazy to discuss stupid thinks with you, please go out here and do not lurk me. You have not a Hyperion 1420 and a 5 and 6 cells battery, and this set-up will not work on a 1420. That's all to you my friend, please writing your half wise statements in your threads. I'm out for a while.