Homemade Battery Packs

It seems that, with no doubt, lowering charge-float voltage will prolong capacity and lifespan of Li-ion cells.
file.php

Multiple sources estimate a doubling of usable cycles from a .10V reduction in peak charge.
In some cases this is a mere 10% capacity reduction in preliminary capacity.
But soon, the capacity in the lower charge voltage cells exceeds the capacity of the higher voltage cells.

More info - Determining Optimal Charge Voltage

LiFePO4 types are much more difficult to quantify due to their charge-discharge process.
 
DrkAngel said:
It seems that, with no doubt, lowering charge-float voltage will prolong capacity and lifespan of Li-ion cells.
................
LiFePO4 types are much more difficult to quantify due to their charge-discharge process.

I charge my LiFePO4's to 3.4 volts. The extra amphours I would get by going to 3.65 is not worth the decreased cycle life.
 
I was thinking about going makita packs..since I can get a 18s8p delivered for 260.00. However have to sort through everything and ect. However they are also 5C cells!

Could the same pack be put together with laptop cells for cheaper? What model laptops should the batterys be coming from? Any higher discharge cells that come in lots from recent laptops? 2012+
 
Great price for Makita cells, dependent on condition ... if you need the C rate.

Laptop cells, in the 2600mAh range are cheaper and more plentiful.
I have seen few "actual" higher capacity cells, forgeries, mislabeled, abound, especially in "replacement" packs.
 
ok great, I only need about 30 amp max discharge, as Im going to go for a little higher voltage..around 72v.

What model packs should I look for however? I want to match the cells vs getting 10 different types of cells.


Will be built in 3 6s packs and balanced charged with bc charger.
 
1KW said:
ok great, I only need about 30 amp max discharge, as Im going to go for a little higher voltage..around 72v.

What model packs should I look for however? I want to match the cells vs getting 10 different types of cells.


Will be built in 3 6s packs and balanced charged with bc charger.
At 30A max, you want a minimum of 30Ah, (12p = 31.2Ah - 2600mAh cells)), more is better!

I highly recommend cells of same capacity and brand!

"Lot laptop batteries -new" search on ebay might be your best bet.
Local recycling centers?
Large companies often replace quantities of laptop packs annually, "ask da boss"?
Craigs list?

After finding cells, refer to - Cell Testing and Selection
 
In regard to charging cells, there is a difference between "charge" and "float" voltage.

"Float" voltage involves applying the desired resultant voltage. This provides an initial high amperage charge rate that reduces till desired pack voltage is attained. Amperage reduces to an ultimate .0 Amps. The disadvantage is a prolonged, minute "topping" charge. This method is used in charging using a variable power supply, such as the venerable MeanWell, adjusting the resultant voltage then waiting till pack nears set voltage.

"Charge voltage", by contrast, typically refers to the speedier method of charging cells by supplying a higher voltage to cells and disconnecting voltage when amperage reduces to a predetermined amount. Charge voltage is higher than the desired resultant voltage. This method requires a "smart charger" that stops charging at an estimated higher voltage, as amperage drops to a certain level. The purpose is to attain a final voltage by estimating the voltage and amperage necessary to acquire the desired resultant voltage after charge stopped. The advantage is a much quicker "topping" voltage.

Hypothetically, a battery could be fully charged at a 1C amperage rate in 1 hour. However, this would involve hazardous voltages, a 4.2V rated cell might require a 4.5V charge voltage. This could be extremely damaging and dangerous. Typically, "charge voltage" is limited to 4.25-4.3V and regulated by multiple cutoffs.

"Smart" chargers often "cycle" applying charge voltage, then stopping, waiting, and measuring the retained voltage, then reapplying charge etc. There are

Personally, after determining optimal charge voltages, I might charge at 4.2V per cell but manually stop charge when amperage reduces to <1amp. Reduced amp estimated as point to attain a 4.1V, or other desired voltage, final "float" voltage.
 
fwiw, termination current seems to matter too, from here.

the author said:
Battery charge voltage and current

In my charger reviews I often comment on a high termination current, and says that the battery will not be fully charged, but I do not specify more precise what that means.
In this test here I have taken a battery and charged it with many different charge settings and then measured the capacity and energy in the battery. All the measurements was done on automated equipment, i.e. voltage, current and timing was exactly the same in each test, except the charger termination setting. For the test I have used a fairly new AW18650-26 cell.

The charger started at 1.3 ampere and followed a perfect CC/CV curve until the specified limits.
After charging the battery rested for one hour.
The discharge was done at 1 ampere and stopped when the battery was down to 2.8 volt.

In the table the green line is what a typical hobby charger will do at the 1A LiPo setting.
CDTest-AW18650-26.png



Plotting the voltage after 60 minute versus capacity shows a very good correlation:
CDCurve-AW18650-26.png

As can be seen from the table, both the termination current and termination voltage is very significant to get the most energy from the cell. The downside to filling the cell completely is lifetime, a cell with a high resting voltage will wear down a lot faster and this will happen even if it is not used.
 
But what if a BMS is used?
The charger could (should?) be a "dumb" charger, with no CC or CV in that situation, right?
Because the BMS takes control of everything?

I can get "or" able power supplies roughly 56V/5A for $20 (for 13s LiPo pack).

Would be perfect for me as I can add more power supplies at some point if 5A charging feels too slow.

But they are just simple power supplies, no CC/CV.
Would those work for a BMS?

Thanks!
 
powersupply said:
But what if a BMS is used?
The charger could (should?) be a "dumb" charger, with no CC or CV in that situation, right?
Because the BMS takes control of everything?

I can get "or" able power supplies roughly 56V/5A for $20 (for 13s LiPo pack).

Would be perfect for me as I can add more power supplies at some point if 5A charging feels too slow.

But they are just simple power supplies, no CC/CV.
Would those work for a BMS?

Thanks!
Very much depends on the BMS!
Some seem worse than useless, while others control every segment of the charge-discharge process.

I use a bulk (dumb) charger because I want the capability of adjusting the final voltage.
Effectively, I can "program" the Ah capacity vs longevity of the battery pack.

Manufactures tend to recommend a maximum of a .8C charge rate.
I try to not exceed .5C charge, or continuous discharge, for my recycled laptop cells.

Non-adjustable chargers can have their voltage trimmed by running through one or more diodes.
Voltage drop can be between .3V - .7V per diode.
Smart charger might require a return voltage which the diodes block.
A small wattage resister bypassing the diodes might resolve this malfunction.
I extended my charger cord with an additional cord, connected with a Deans (T) plug.
Double T plug, inserts (modules) can be built and placed in line for various "adjustments".
 
It is for the SmartBMS unit.
The charger would be mounted onto my scooter and the charging rate would be 0.25C over 4 hours.

It is a simple power supply:


Thanks!!
 
Yes, fan of course, plus some casing against splatter etc.

The 275W is the unit I was aiming at.
The 280W cannot be adjusted unfortunately, love its isolated 12V though.

In the comments (of the 280W) somebody writes that it is not easy to modify the voltage, but no details on how.

So yould that 275W power supply work in my cenario?

Thanks!
 
My 29.6V 31.2A (8s12p) Sanyo 18650 cells seem to have suffered an early demise.
After barely 1000 miles, pack is deteriorating quickly.

Cells began as an 6s16p (22.2V 41.6Ah) stuffed into an eZip RMB shell.
See -22.2V Battery Woes

file.php

While initial tests provide 60 miles @16+mph and ...
"41.6Ah pack (4.20 - 3.71V per cell discharge) accepted 32.86Ah recharge, at an estimated 80-85% DOD, for a ~38-41Ah capacity!
So cells appear to be of excellent capacity condition."
... I felt performance "... inadequate".
Initially I considered converting pack to 12s8p (44.4V 20.8Ah) but decided on a more utilitarian 8s12p (29.6 31.2Ah).
After initial excellent speed and range, recharge Ah began reducing with nearly every cycle.
I blame my own ignorance!
Initially I discharged to near 3.6V and charged to 4.20V, using my mAh/100thV Discharge Capacity Map as my guide.
file.php

Unfortunetly, of fortunetly, I developed a method of capacity mapping using charge mAh/200thV.

This method precisely quantifies the amount of electricity to raise pack voltage between steps of voltage.
Results were enlightening!
file.php

It seems discharging below 3.72V is needlessly damaging!
Further, all experts agree, lowering peak charge voltage greatly increases capacity and cycle life!
A 1/10V reduction in charge voltage will double usable life!

As indicated by the graph, the last capacity bulge ends at 4.05V.
Keeping voltages between 3.72V and 4.05V should increase usable life to 300%!
This from an initial loss of usable Ah of a measured 18%.

Well, this is merely a prelude to my next project using my best Sanyo 18650 cells.
9s12p (33.3 31.2Ah) will squeeze into an eZip RMB pack.
I have 108+ saved (best bleed down) cells.
Bleed down test revised to maintained above 4.05V after 6 months - my last bulk charge of candidate cells.
But I intend on retesting all my Sanyo cells with my Bulk Capacity Test, 60 cells at a time.
All cells will be discharged rated from 4.10V to 3.7V.

33.3V pack will work on my 36V controller Winter eZip and my universal controller eZip LS.
I intend on keeping a spare oem geared wheel-motor combo, ready for quick swap for high torque 20mph legality from 33.3V.

I have to keep my upgeared eZips for my 25.9V Lipo packs.
7s12p is the best I can squeeze into the RMB pack.
Last build gave adequate performance for 5000+ miles. (Still use it for quick trips around town - 6500+ miles)
Charge Capacity mapping indicates a 50% increase in usable life by reducing charge to 4.15V, with a miniscule loss of capacity!

Test was run twice to confirm odd capacity bulge between 4.10-4.15V.
file.php

Alternatively, reducing charge voltage below 4.05V increases usable life +300% ... but with a 25%+ loss of capacity.
25.9V 26Ah Lipo pack reduced to to ~18Ah usable ... might be worth it for 6 years of daily use - vs 2 years.
Best of all ... I can use a 2nd battery in the rack for longer trips ...
 
Keep in mind that running a 1C pack at 1C isn't going to net you very long lifetime, regardless of float voltage.
 
Teh Stork said:
Keep in mind that running a 1C pack at 1C isn't going to net you very long lifetime, regardless of float voltage.
Of course ... with laptop cells, I recommend a maximum surge of 1C, with a continuous drain of .5C or less.
For the eZips with 30-35A controllers-450w motor, I've always recommended a 30Ah+ pack.

The exception is my laptop recycled Lipo builds.
Unfortunately, the eZip oem battery pack can fit a mere 25.92Ah at 25.9V.
But even with full throttle starts I've never blown the 30A fuse I installed ...

Deep discharges ...
High float voltage ...
High discharge rates ...
High charge rates ...
Anything that produces heat, is damaging!

Also, interesting to consider ...
As capacity decreases, charge and discharge rates should be likewise reduced.
In other words, the C rate should be based on actual capacity rather than original rated capacity.
This is why batteries begin to deteriorate at an accelerated rate.
If retasked to low drain applications, while cells are still of reasonably capacity, they might enjoy a fruitful ... retirement ...

Power backup-emergency power - with inverter
Lights, many power tools etc., run well on 120V DC
A camp florescent or LED lantern
Part of a laarrge pack for mobility scooter or trike
 
Building 2 packs for Winter
For the first time I will be building 9s 33.3V packs.
I have 2 eZips with compatible controllers.
1st 2008 Moutain Trailz converted to 36V controller and rigged for winter, studded tires etc.
2nd is my 2009 Trailz LS which is 22.2V - 37V+ capable.
Packs will be as large as possible to offset cold effects.

Lipo - recycled laptop cells
I plan on using a minimum of 21 6-packs of 2160mAh cells.
9s14p for 11.1V 30.24Ah = 1.007kWh
I'm hoping to stuff into a 6 pack insulated cooler.
Ideally, pack will be mounted between suspension fork and handlebars and run wire back to controller.
Capacity mapping indicates optimal voltages above 3.75V and about 4.15V peak charge.
(My previous build using these brand cells, at 4.20V charge, provided 2 years - 5000+ miles of acceptable use.)
However, I intend on limiting charge to 4.05V, although this neglects a large capacity bulge, it will, reputably, more than double the usable life. (Will take several years to confirm!)

file.php


Li-ion - 18650 recycled laptop cells
Careful planning will allow me to squeeze 9s12p into an oem eZip RMB battery shell.
9s12p = 108 2600mAh cells = 33.3V 31.2Ah = 1.038kWh.

About 8 months ago I bulk charged my entire supply of cells and used the B rated cells for my 22.2V and 29.6V builds.
Longevity was disappointing, so peak charge voltage will be lowered to prolong useful lifespan.
Optimal charge voltage, for these Sanyo cells looks to be 4.05V.
I just checked the residual voltage on the remaining cells and selected 150+ that, after setting for 8 months, still retain 4.05V or higher.
1. I will equalize all cells to 4.10V.
2. I will use my bulk capacity test method to rate all cells.
3. Best 108 cells will be selected and arranged into 9 banks of equal mAh capacity.
Capacity mapping recommends limiting charge to 4.05V and discharge to above 3.7V.
Hopefully limiting these voltages will greatly prolong usable life.
I will be able to compare directly with this years previous builds, (same batch Sanyo cells).

file.php


Standardizing at 4.05V per cell will allow 1 charger for the Winter.
1 MeanWell S-350-48 modified to bulk charge at 36.35V 9.5A. (< .3C charging rate recommended ... by me.)
(I never found a bargain price for 36V S series MeanWell)
MeanWell S-350-48 mods will be documented on the ES Wiki.

S-150-XX mods are well documented or being updated in stages.
 
Recycled Lipo charged to 4.20V per cell, were previously documented at 2 year 6000 mile usability.
Multiple reputable sources insist that lowering charge voltage from 4.20V to 4.10V will double the number of usable cycles ... at a mere 10% capacity sacrifice.
Lowering voltage to 4.05V should increase usable cycles to 250-300%.
Usable capacity from the 4.05V might exceed the usable capacity of the 4.20V pack before the 1st year of use ends.
Previous 2 year usable life might extend into 5-6 years and previous 6000 mile total might exceed 12-15 thousand miles.

Since my previous charges were typically partial, rather than full depth, my use shouldn't suffer at all.
Instead of the previous ~500 cycles, ~1500 seems viable.

These cells vary in age, but even some dated 2004 still measured at better than 90% of oem rated capacity!
 
Bump
 
A question on your experience with recycled 18650 batteries...

I recently bought 49 dell laptop battery packs in 2 lots. One lot was 27 packs and the other 22 packs. The total price was just under $70 USA for all the packs.

I have yet to fully count the total number of batteries but it's about 325 used 18650's. Mostly Samsung 2.600 amp hour cells but also Sony's LG's and others I have yet to identify.

Right now, I have them separated into 3 groups.
A: Those that are above 3.25 volts,
B: Those that are 3.24 to 2 volts
C: and those that are 1.99 to zero volts.....I have about 113 of these.

At this point my question is what should I do with the 113or so that are 1.99 to 0 volts. They may be the dangerous ones and I am wondering if they are worth trying to revive with a slow charge to see if any are salvageable.

From your experience is it worth trying to salvage low voltage cells?

Thanks!

:D
 
e-beach said:
A question on your experience with recycled 18650 batteries...


At this point my question is what should I do with the 113or so that are 1.99 to 0 volts. They may be the dangerous ones and I am wondering if they are worth trying to revive with a slow charge to see if any are salvageable.

From your experience is it worth trying to salvage low voltage cells?

Thanks!

:D
Want to play it safe?
Apply a 3.5V charge (or lower, 3.25V?) to each cell.
Cell voltage will climb to that point very quickly ... or it won't.

Check for heat during charge.
Check for voltage loss or heat a few minutes after charge removed (self-discharge)
Eliminate any that fail.

If cells retain fair voltage after 1 hour, save for full charge and test.

Method should eliminate "unsafe" cells.
 
DrkAngel said:
some dated 2004 still measured at better than 90% of oem rated capacity!

define 'some'!

some people live past a hundred!
what does that tell you about your chances?!!

DrkAngel said:
Previous 2 year usable life might extend into 5-6 years and previous 6000 mile total might exceed 12-15 thousand miles.

increase cycle life, yes!
calendar life, no!


Battery performance deteriorates over time whether the battery is used or not. This is known as "calendar fade". Performance allso deteriorates with usage and this is known as "cycle fade"
http://www.mpoweruk.com/life.htm


After one year on the shelf, a LiFePO4 cell typically has approximately the same energy density as a LiCoO2 Li-ion cell, because of LFP's slower decline of energy density. Thereafter, LiFePO4 likely has a higher density.
http://en.wikipedia.org/wiki/Lithium_iron_phosphate_battery
 

Attachments

  • MSA-calendar.pdf
    219.4 KB · Views: 188
Back
Top