Building battery from MAKITA Konion PACK

[Doctorbass]

I decided to show you what is a good way to dissassemble the makita pack ( in this exemple i'm using the 1.5Ah).. all 3.0Ah are sold

A picture worth thousand words so.. here they are:


Oh.. before some advices...

- take appart the low cells ( under 2.9 or 3.0V) and keep them away from your work bench.. this will avoid you to introduce them by mistake in your new pack! ( some will keep the cells above 2.5V and under 2.9 or 3.0V and will make a booster pack.. This is an excellent idea if you have some voltage margin availlable from your controler max voltage

- Return the bad cells to recycling!!! dont trash them!.. be eco friendly! ( you can put some tape on the tab to avoid short when they are all together touching others!

- make sure to charge the cells to the same level BEFORE TO PARALLEL THEM TOGETHER.. or ensure to have less than +/-150mV difference

-When assembling multiple parallel group, ensure to have an insulator between the cells from one group to the next to avoid short dur to fragile green skin of these 18650 format cells. This will prevent short during vibration of your new assembled pack!

-Before to use the pack flully charge ALL the cells to 4.1V or 4.2V.. (4.1 will prolong the life of these cells but will give you a bit less capacity)

-Dont OVERHEAT THE CELLS TAB! that will damage the cells ( the trick of soldering on the nickel sheet that is already spotwelded on them if to prevent this... and avoiding you to remove the white stuff

-For 90% of people that dont want this step or dont have the equipment to match the cells to get perfect groups of the same total capacity, i suggest to mix ALL the cells and than to randomly pick some to make your parallel cells groups,

-REMOVE ANY JEWELRY or watch... this will prevent KFC !! and plasma event in your shop !!


---------------- 22 Dec 2012 ADDED INFO ABOUT CROSS CURRENT MANAGEMENT CARE down to that page--------

[Doctorbass]

then

[Doctorbass]

then...

[Doctorbass]

finally...

[Doctorbass]

Real test.. Real results!

around 6.5Ah from 5 cells parallel... and.... 3.8miliohm!.. not bad for these auto balancing 18650VT cells!

That's 24Wh out of 5 these used VT cells

At 5$ per packs that give 4 to 5 cells each... let say 4 cells that's 6.25$ per 24Wh = 0.26$ per Wh

Just 5 of these cells in parallel can sustain 50A continuous and 100A burst... just like a 18$ headway cell


Now there is also a very important advice about HOW TO CONNECT THE CELLS TOGETHER!!

To get max performances and share the stress on all cells with uniformity, please follow these advices:

You want to avoid cross current problem!

Let me explain:

let say you have 10 parallel cells that can give each 10A max.. you would assume that you can draw 100A max.. but it's right only if every cell will give 10A,

BUT.. it is rare that the current will share equally from each cells. You may have one cell that give 25A while some some other cells only give 7A... and we must avoid that for many reasons: parallel cells could age differently causing faster battery degradation, also the cell that give too much current can overheat and cause serious damadges.

Every of these problems occur because that between every cells and the main terminal or main wire, the resistance ( even though it's in miliohm difference) is important and the path the current choose is depending on that resistance. By puting many smaller gauge wire between every paralleled cells tab, the low resistance become more iniform because the resistance of that small gauge wire is more important.

Doc

[texaspyro]

I would rotate the cells 90 degrees and not fold over the tabs. Attach the copper bus bar across the length of the string of cells, then fold the tabs over the top of the copper, and solder again. Much less chance of a solder joint failure.

[999zip999]

Doctorbass, Thank you for for all the good pics..

[Doctorbass]

I would rotate the cells 90 degrees and not fold over the tabs. Attach the copper bus bar across the length of the string of cells, then fold the tabs over the top of the copper, and solder again. Much less chance of a solder joint failure.

That's what i do normally, but with a 80 watts solder iron, you dont need to apply solder onboth side of the copper strip to help transfer heat to melt the solder between the cell and the coppper...

BAD IDEA to flip cell 90 degree... The nickel part can touch the next cell and puncture the fragile green plastic... that'S why i kept them 90 degree from the copper strip...

Over the 2200 cells i assembled.. i never had any problem of solder join failure with thsi method

Doc

[999zip999]

Looks good but my color must be off that looks like the color of brass in the pics.. I'm going to build a 55v 13ah pack and have alot of brass wire flat 12ga.x 18ga. and 9ga. round or 18 ga.sheet copper 1.5in.x 15in. or 20ga. 3ft.x3ft copper sheet. If I could use the big copper sheet is best or the brass flat wire is best for me. This is for the paralle of makita 1.5 cells 2p out of the 3.0ah packs. Yes there are in the 2p with nickel spot welded join togeter at the factory. So to make a 8p cell for 15s battery.

[John in CR]

My question is why cut the factory tab connections at all? In the example above with VT cells you've taken 1p5s and cut the factory tabs and made 10 new solder connections to make them 5p1s, but for what purpose? 1s certainly isn't useful, except maybe as a battery for a LED light, so those 5p1s blocks need to be connected with others in series to be a useful pack. Breaking up a perfectly matched 5s pack seems like a waste to me, and it multiplies the amount of work by about 10X.

eg Lets say you have 10 of those 1p5s packs and want a 5p10s pack (37V6ah), then after just 5 simple tab on tab solder connections to connect pairs of packs in series, you're ready to attach your leads or bus bars to the five 1p10s strings. The other way will require 100 solder connections for the same 5p10s pack. Call me lazy but I'd much rather make 5 easy tab on tab solder joints than 100 solder joints, one at each end of each battery. Plus I see our biggest risk during DIY battery construction as being the heat risk of making solder connections to the batteries, and disturbing the time proven factory tab weld construction is asking for trouble in addition to creating lots of work.

[Ypedal]

agreed, to a point.. these cells remain ballanced very well, but by grouping P groups first, then series those you can monitor the voltage of each group as a single display. instead of each string seperately.. if you care to monitor.. and you should..

[Doctorbass]

My question is why cut the factory tab connections at all? In the example above with VT cells you've taken 1p5s and cut the factory tabs and made 10 new solder connections to make them 5p1s, but for what purpose? 1s certainly isn't useful, except maybe as a battery for a LED light, so those 5p1s blocks need to be connected with others in series to be a useful pack. Breaking up a perfectly matched 5s pack seems like a waste to me, and it multiplies the amount of work by about 10X.

eg Lets say you have 10 of those 1p5s packs and want a 5p10s pack (37V6ah), then after just 5 simple tab on tab solder connections to connect pairs of packs in series, you're ready to attach your leads or bus bars to the five 1p10s strings. The other way will require 100 solder connections for the same 5p10s pack. Call me lazy but I'd much rather make 5 easy tab on tab solder joints than 100 solder joints, one at each end of each battery. Plus I see our biggest risk during DIY battery construction as being the heat risk of making solder connections to the batteries, and disturbing the time proven factory tab weld construction is asking for trouble in addition to creating lots of work.

John.. I know it might appear a strange decision, but let me explain:

The original nickel link are too wide compare to the diameter of the cell. and that will create a DANGEROUS short possibility to thecell next to it. in fact, the large nickel link will touch the green plastic shin of the next cell and fo rthe negative tab, there is no worrie, but fo rthe positive tab, WE MUST BE CAREFULL

Plus when you solder the nickel sheet link, it will melt the green plastic that is touching and will short to the negative shell of the cell that is VERY close to the positive tab!!

that'S why i preffer recommanding my method.

Byt, by usig my method oc cell assembly i never had any problem after 7000km if hard ebike ride and alot of vibration over 432 cells ASSEMBLED BY THIS WAY

yes... 432 cells so it's 861 connections....

Sumarry:
These really thin nickel link are very SHARP and i found that on the positive tab, when you put many cell together i parallel , these sharp edges can puncture really easy the green skin of the cell next to it.

By doing the methode above, you also add a kind of soft interface from teh cell to teh bus bar and it help absorbing vibrations. The dual plied side of the nickel link act like a little spring.

I already had some issue reported by some Makita pack buyers that did not pay attention to that ( eventhoug i always ensure that people that buy teh cells will HAVE the skill and knowledge to assemble their pack) and they had some short at this exact place.

You MUSt pay attention to that ! when assembling your pack!!

Doc

[Doctorbass]

agreed, to a point.. these cells remain ballanced very well, but by grouping P groups first, then series those you can monitor the voltage of each group as a single display. instead of each string seperately.. if you care to monitor.. and you should..

Yes.. i'm i the group of person who preffer PARALLEL FIRST and than serie...

This was debated alot and John, in the industrie you will NEVER find any "serie first and than parallel"

Too many seperate induvidual cell without any monitoring!!

Only LiMn and nicad MUST be done the way you suggest due to the particular chemistry.

Doc

[999zip999]

I asked earlier about what ga. can be used for the buss bar. The nickel is only 24ga. at best and does not (?) conduct as well as brass or copper. Can I use 20ga. copper sheet for the buss bar for 55v 13ah v cell pack. The v packs are in 2p for alot easier build. Or what is the rule for a proper buss bar ? Thanks for the help.

[Doctorbass]

I asked earlier about what ga. can be used for the buss bar. The nickel is only 24ga. at best and does not (?) conduct as well as brass or copper. Can I use 20ga. copper sheet for the buss bar for 55v 13ah v cell pack. The v packs are in 2p for alot easier build. Or what is the rule for a proper buss bar ? Thanks for the help.

Same rule that appl;y to the conductor you will use

The rule about that conductor is simple.. how many amp will you cary between the battery and the controller?

ex: 10 gauge is ok for 30-40A.. maybe 50 Burst battery current.. so 10 gauge wire equivalent is ok for the bus bar.. no need of special calculation.. just what make good sence!

you can find the help with this page showing the cros sectional or circular mils of diferent gauge wire:

( in teh WIRE section) by cliking on the right of the page
http://www.bcae1.com/

The Ah (Ampere hour) is not importnat for the size of the bus bar.. it's the A ( current) that is important.

Doc

[John in CR]

agreed, to a point.. these cells remain ballanced very well, but by grouping P groups first, then series those you can monitor the voltage of each group as a single display. instead of each string seperately.. if you care to monitor.. and you should..

Yes.. i'm i the group of person who preffer PARALLEL FIRST and than serie...
This was debated alot and John, in the industrie you will NEVER find any "serie first and than parallel"
Too many seperate induvidual cell without any monitoring!!
Only LiMn and nicad MUST be done the way you suggest due to the particular chemistry.

Doc

For those who want to parallel each level it's still less than half the work if you don't take them apart. Plus you only need to solder the parallel connection where the tab is at the negative cell end, so you don't risk melting or weakening the plastic protector at the positive end. I cringe when I see people break these packs into individual cells, because I know the work that they've created for themselves, along with the cell matching they've lost. You don't break up the 2p V packs, so why do it for the 1p VT packs?

As far as monitoring at the cell level, it's not done on hundreds of millions of lead acid car batteries, and these LiMN cells self balance through the same mechanism, so I don't see much point unless you're running your pack at stressful levels.

A required step not mentioned though is capacity matching.

FWIW, the 2000+ Konion cells that I've built into packs for myself and others haven't had a single failure. Some of those packs are over 1k cycles, including those on my daily rider that has seen 2 cycles on a typical day for over 2.5 years, so dismissing my easy build build method as incorrect is unwarranted based on real world results. If pack building was really as hard as some make it, then I wouldn't even build my own packs. These cells don't burst into flames like Lipo, so there's no real risk building them in a less than technically ideal manner, and the time I save is worth more than the pack itself.

[Doctorbass]

John,

I'm glad to see we offer two good solutions for the future pack builder

Doc

[beast775]

im sure this is my next step in a battery.thanks for all the posts to make it easier-john-drbass,ive been using 18650 in high drain hobby applications for years and have never had one go bad.there seconds from 2nd tier battery suppliers.so i can imagine the sony are alot better.how many packs will i need for 48v 12ah drbass?

[Doctorbass]

.how many packs will i need for 48v 12ah drbass?

HOW MANY CELLS DO I NEED FOR A BATTERY OF X ah and Y volt ?

It is simple to calculate:

12Ah ?.. ok... divide 12 by 1.5Ah if you choose the V cells or divide it by 1.3 if you choose the VT cells

ex for the VT: 12 / 1.3 = 9.2.. so we round that number to 10.. for 13Ah

now you want 48V.

Generally we talk about NOMINAL voltage.. that mean this is the voltage the cell have at 50% SOC. ex a makita cells have 3.7V at nominal(50%) and 4.2V full (100%) and 3.0V at 0%

This is not a direct proportional relationship with the voltage due to the discharge curve shape.

so for 48V you calculate 48 / 3.7V = 12.9.. we generally round that number to 12 because it is more common to have 12 cells and more compatible for some measuring device.

12 x 3.7 = 44.4V

but you can still choose 13 x 3.7 = 48.1V if you wish

now you know how many cell in parallel you need ( 10 cells parallel) and how many of these parallel block of these cells in serie you need (13)

so you need 10 x 13 = 130 VT cells

generally i recommand people to get 10% more cells to have some spare to replace damaged cells during the construction or other replacement needed.

Doc

[beast775]

thanks drbass, i kinda got what you said .so i think there is 10 cells per pack? i need to go study a bit on these packs and look at your youtube channel.seems like i might need a shitload of packs! merci mec J'apprécie votre aide que je vous dois un mec .

[999zip999]

Yes both ways work. But with Doctorbass I could put a balance wires on for easier monitoring on a cell level in parelle first. Gald to hear these cells are so versible.

[999zip999]

JohnCR, How do check cell balance and charge the strings before assembly ?

[marty]

x
-Before to use the pack flully charge ALL the cells to 4.1V or 4.2V.. (4.1 will prolong the life of these cells but will give you a bit less capacity)
x

JohnCR, How do check cell balance and charge the strings before assembly ?

I got a CBA III - Computerized Battery Analyzer to check cell capacity.
http://www.westmountainradio.com/cba.php

What charger should I buy for my used Makita Lithium-Ion battery building experiments? Should I charge each cell separately or in parallel groups?

[Doctorbass]

x
-Before to use the pack flully charge ALL the cells to 4.1V or 4.2V.. (4.1 will prolong the life of these cells but will give you a bit less capacity)
x

JohnCR, How do check cell balance and charge the strings before assembly ?

I got a CBA III - Computerized Battery Analyzer to check cell capacity.
http://www.westmountainradio.com/cba.php

What charger should I buy for my used Makita Lithium-Ion battery building experiments? Should I charge each cell separately or in parallel groups?

Any RC charger would do the job.. check on hobbyking. Or a lab powersupply you can buy on ebay that can ve adjusted at 4.20V.. so any power supply with adjustable current and voltage up to 5V would be ok.

Before to put all cells in parallel group, you should ensure that the voltage difference between all cells you want to parallel is not too high.. let say 200mV or less is acceptable.. otherwise the lower cell will charge too fast and be damaged. to charge them, I suggest to put all cells in group of voltage.. all group with less than 200mV between the higher and the lower cell.

( this is just to charge them).... when you'll make your parallel group, I suggest to randomly choose cells and to group them. that will average all your group of paralleled cells for your pack.

Doc

[marty]

To build a big battery pack from Makita cells, what do you think about using a whole bunch of these plastic holders. Eliminate all the labor of soldering.
14.8V 4 x 18650 Battery Holder Case Box with Leads
http://www.dealextreme.com/p/14-8v-4-x- ... ads-103855