Homemade Battery Packs

On the battery resistance/impedance testers, I have a 12v Elk tester that I got years ago and use to track the state of 12v SLA/AGM/Gel cells.

As for the "SLA replacement" LiFePO4 alternatives, anyone have experience with the BatterySpace.com offerings, especially this one:
https://www.batteryspace.com/LiFePO4-Battery-12V-9Ah-108Wh-18A-rate-with-LED-Indicator.aspx

or this one:
https://www.batteryspace.com/LiFePO4-Battery-12V-10Ah-120Wh-40A-rate-----Replace-SLA-12V-10Ah.aspx

Specs do look better than the "Bamboo" CR1210 from Clean Republic, with the latter 40A peak LiFePO4 more than adequate for use with an eZip 450w motor.

I am still confused on the old eZip Trailz motor ratings, so if they all are 450w as several on this site have confirmed, then these Powerizer branded batteries from http://batteryspace.com seem to have adequate current rating for the eZip.

Even better for my old Currie Mountain ebike from 2001 which has dual 12v12ah in the frame mount pack, the folks from batteryspace.com have this "12v SLA replacement LiFePO4" with a built in BMS, not just a battery protection circuit:
https://www.batteryspace.com/Powerizer-LiFePO4-Battery-12V-12Ah-144Wh-30A-rate-with-Balancing-PCM.aspx

It's not really the wattage of the motor itself you'd worry about with the battery pack capability (c-rate), but rather the controller's current limit. As long as the battery can output whatever the controller is capable of pulling from it, without damage to the battery (or cutting off from overcurrent protection, if any), then the battery is suitable for use with it.

7tronics said:

I am still confused on the old eZip Trailz motor ratings, so if they all are 450w as I am told, then these Powerizer branded batteries from batteryspace.com seem to have adequate current rating for the eZip.

Click to expand...

Ezips with 450w motor use a 24V 35A controller.
24V x 35A = 840 watt input! - battery drain.
450w is the motors peak output watts.
Peak HP at near 40% of top no load RPMs, is approximately 50% efficient.
So a battery output of near 900w is required to meet 450w motor output.

This is what will work for sure.

http://www.batteryspace.com/lifepo4-prismatic-box-battery-12.8v-20ah-256wh-10c-rate-without-pcm.aspx

Won't fit in the stock ezip battery box. You have to make up your own box and mounting.

Mine stay high and dry in a milk crate on the back rack.

They sell them with balancers too.

http://www.batteryspace.com/lifepo4-prismatic-battery-12.8v-20ah-256wh-10c-rate-with-led-balancing.aspx

I bought the ones without balancers and put balance leads on them so they can be balanced charged with the Hyperion 1420.

Lipo alarms for LVC.

DrkAngel said:

An eZip with 29.6V 31.2Ah pack, (8s12p), without any regearing, will push to (legal - USA) 20mph capability!
With a substantial torque increase ...
A 24% torque increase, as opposed to the 20% (initial) torque decrease associated with 16T or similar with the 13T mod.
Problem is, finding a controller that is 29.6V capable.
Except for the pre-mid 2008 eZips, universal voltage controllers seem to be the only available.

Representative 24v to 29.6V comparison

Power (red lines), beyond the road load-wind resistance line (black), looks like ...
+50% - 10 mph
+100% - 12.5 mph
+200% - 15 mph

Hill climbing capability seems impressively increased!

Click to expand...

Put a 96 cell 8s12p 18650 build into an eZip oem pack.
29.6V 31.2Ah is the maximum 923.52Wh, rather than a "new" build I rebuilt my 96 cell 6s16p 22.2V 41.6Ah pack.
See 22.2V Battery Woes

Newer eZips have a controller that limits pack voltage to below 30V.
Fortunately, I have a 2008 eZip Mountain Trailz and a 2009 eZip Trailz LS that accept better than 36V (42V fully charged).
My 2009 LS has a 25.9V recycled battery, the 16T mod, 2.125" comfort tires and cruised at ~19mph.
29.6V pack pushed the cruising speed to 22-23mph a ~4mph increase.

I had intended to use the 29.6V on an unmodified eZip as a better way to attain the "legal 20mph" without the torque loss of regearing.

Difficult to find, but I did add an 8s balance cable and will charge-balance with an iMax B8.
After several monitored cycles I will begin charging with a modified MeanWell S-350-36, modded for 32.8V 10A.
Yes, I will post up instructions and values for current mod and probably voltage mods.

Unfortunately:
1. Both my 30V+ compatible bikes have been re-geared for 25.9V optimal 20mph.
2. So far, only the 2007- early 2008 and some 2009s have 29.6V compatibility.
3. No aftermarket controllers are compatible with this voltage, (multi-voltage controllers will work but will not fit in the eZip rack, as a same position replacement.)


Not as pretty as I would like ... 29.6V build cannibalized-patched from 22.2V build.

Fortunately:
1. I can restrain throttle use "in public" and not attract attention.
2. I can carry my 25.9V 26Ah (recycled Lipo) pack in one rack and 29.6V 31.2Ah (recycled 18650 cells) in the other and change with the flick of a switch.
That extra 4mph and added torque helps me blow by those spandexters. Cruising up hills at 20mph, while they huff and puff at 10mph ... well ... you know ...

One of the problems of building large packs with recycled cells is, the time consuming step of capacity testing every cell.

For one of my most recent builds I experimented with matching a particularly good, possibly never used, batch of cells solely using a bleed down method-elimination.
This involves:
1. charging the entire quantity of cells to an identical voltage.
2. monitoring voltage loss over several days and eliminating any with voltage "leakage".
Results were nearly perfect. Discharge tests showed the best cells to be at nearly 100% of rated capacity.

Unfortunately ... one bank reaches optimal DOD a bit sooner ... which leaves the remaining banks with unused capacity. Since this discharges the weaker bank more deeply and at a faster rate the capacity differential will get worse! Finding the weaker cell-cells could involve a disassembly and thorough capacity testing, or ...

Finding That Weak Cell - easy method!

Pack = 8s12p 96 cells
Problem = 1(?) weak cell
Diagnosis:
1. charge-balance all cells to optimal full charge voltage
2. remove all balance wires, leaving the negative and positive "rails".
3. discharge normally to near optimal DOD
Result = weak cells in any "s"(series) will display a noticeably lower voltage
Cure = replace any cells displaying lower voltage
4. re-attach balance wires
5. charge-balance pack, use as normal, recheck voltage balance at optimal DOD = fixed!(?)

Also see - Bulk Capacity Testing

8s 29.6V 31.2Ah eZip pack build - run through 2 cycles.
35 miles @ ~19mph dropped cell voltage to 3.72V.
2/100th variance between cells.
Max optimal DOD rated at 3.62V.
22Ah recharge.
27Ah+ total pack capacity? (using optimal charge-discharge voltages)
I intend on trying my "finding weak cells" method to equalize cell voltages at optimal DOD!
Will replace bad cells.
If no "bad" cells found, will reshuffle, trade weak cells from weak banks with strong cells from stronger banks.

7s 25.9V 25.92Ah (6000 mile Lipo recycle build) degraded to 12-13Ah capacity.
Cells suffering imbalance at near optimal DOD 3.7V.
Will rebuild with newer, accurately balanced cells.

Will re-recycle usable cells into some bulk pack :
Giant pack for eTrike or mobility scooter;
BIG! 11.1V pack for 110V AC inverter for emergency power;

Bulk packs are fun.
Instead of thorough capacity testing, just add more cells to any weak bank.

25.9V 26Ah recycled Lipo.
6000+ miles. (EOL) Will rebuild.
Capacity-range reduced to 12Ah 20 gentle miles (without pedaling)

Initially my 10A charger charged this pack (~24Ah) with no noticeable heat production.
Now, at ~12Ah capacity, the 10Ah charger creates a noticeable warmness to the pack.

Apparently ... the C charge rate should be regulated based on present rather than original capacity?
.25C - .33C <.5C recommended charge rate.

Best lipo config for backpack, 12S 10ah






p.s. I dont recommend zippy flightmax 5ah, after 2C discharge inner temperature >60C

Hello.

I have been reading a lot about 18650 laptop battery packs, and I would like to ask about a few questions, that I couldnt found anywhere...

1. How many cells can be "SAFELY" get in parallel? I was thinking in more than 50 to get something like 12-14S and maybe 50-60P to get 3 KW-h in a stacionary enviroment. Im thinking about an isolated house with no power connection and photovoltaic solar panels.

2. Any recomendation of BMS? What about balancing so big packages? I suppose standar 12-16S BMSs are all 3-5P oriented..., not 50P

3. I allways see welding cells. Could be possible build a pack with some clipping mechanism (no welding) to conect cells. I have reading something about "Neodymium magnets" instead weld... and it could be better for build and mantain the pack (removing and replacing bad cells).

4. Any recomendation to get even more KW-h. I.e. reach 15 KW-h with 5 of this packs in parallel?

5. Any recomendation for the sun power charge profile (It can take big fluctuactions with clouds...etc.) and the need to mantain the supply power to the house, thats means some diferent as tipical LI-ION batterys charge-discharge mobile phones, ebykes. etc.

segalion said:

Hello.

I have been reading a lot about 18650 laptop battery packs, and I would like to ask about a few questions, that I couldn't find anywhere...

1. How many cells can be "SAFELY" get in parallel? I was thinking in more than 50 to get something like 12-14S and maybe 50-60P to get 3 KW-h in a stationary environment. I'm thinking about an isolated house with no power connection and photovoltaic solar panels.

Click to expand...

1st, decide on what working voltage you want to run.
12V, 24V, 48V through an 120V AC inverter?
120V direct DC?
But, in regard to how many can safely be run in parallel ...
I don't foresee any "danger" in an unlimited quantity run in parallel, the more the better.

segalion said:

Hello.2. Any recommendation of BMS? What about balancing so big packages? I suppose standard 12-16S BMSs are all 3-5P oriented..., not 50P

Click to expand...

"Balancing" is not as important as an initial build to banks of equal capacity.
Li-ions will "naturally" discharge, then recharge to the same "balanced" voltages per bank.
If care is taken in the original construction, optimal charge and discharge "balance" will maintain automatically.
If equalized at full charge, but one bank discharges deeper at max discharge voltage, than add one, or more cells to the weak bank .
You do not need an equal number of cells in parallel banks. - You do want equal capacity in each parallel banks.
It is, however important to limit charge voltage and limit DOD (Depth of Discharge)to optimal cell voltages.
AC inverters of fair size can usually be set to shut off if a lower voltage is reached.
But any BMS is not affected by any limit on quantity of cells in parallel.

segalion said:

Hello.3. I always see welding cells. Could be possible build a pack with some clipping mechanism (no welding) to connect cells. I have reading something about "Neodymium magnets" instead weld... and it could be better for build and maintain the pack (removing and replacing bad cells).

Click to expand...

A good connection to each cell is important!
Spot welding is best, soldering is good, soldering spot welded tabs seems the best for most applications.
"Clips" and probably magnets are subject to tarnish, corrosion, arcing etc ... not durable or reliable.

segalion said:

Hello.4. Any recommendation to get even more KW-h. I.e. reach 15 KW-h with 5 of this packs in parallel?

Click to expand...

"Packs" in parallel is a great idea.
You can even run multiple packs through the same BMS, you just need a modular plug with a sufficient number of connecting wires, I have used 20 and 24 pin computer power supply connectors on my early builds.
Modular packs are a good idea, you can detach and maintenance-repair one module while retaining functionality at a slightly reduced capacity.

segalion said:

Hello.5. Any recommendation for the sun power charge profile (It can take big fluctuations with clouds...etc.) and the need to maintain the supply power to the house, that's means some different as typical LI-ION batteries charge-discharge mobile phones, ebikes. etc.

Click to expand...

With laptop type 18650 cells ...
Limit charge and discharge voltage to below .5C, to limit wasted heat production.
Li-ion are extremely efficient at storing and delivering electricity, at modest rates.
Charge and discharge efficiencies are both in the 98% efficient area.
Compare this to SLA which can run at 75%, or worse, each way!
! You might need 1/2 the number of solar panels to produce-store-provide the same amount of electricity? !
You certainly will need a smaller "rated capacity" of battery to store the same amount of usable electricity.

Limiting charge and discharge voltages will greatly increase the longevity of a Li-ion battery pack.
Of course this will require a larger pack ... but might increase daily cycles into the 1000's???

Many "Greeners" add wind power to their home power systems.
Winter solar production is reduced while natural defoliation increases wind production,.
Perhaps more importantly, nighttime production is maintained!

Thanks a lot DrkAngel...

Thats really sound so easy....

Please, let me make some appoints...

1.
I was thinking in 48V. Solar voltage pannels, MPPTs and chargers run tipically at this... Probably its a 13S config.
Then there are an AC inverter to 220V for supply home electricity...

Glad to see more parallel cells better... But please, could you explain what happens with a bad cell inside a big bank? The cell became a short circuit (damage or explote everything) or an open circuit (not affect and run as with 1 less cell...)?

2.

I understand importance of pre-balancing, cells homogeneity, security Volts and speed Cs... DOD, temp etc. But I dont undesrtand very well the BMS function.... I read a lot of stuff about balance, somenthing about current limit between active balancing... and so this is why I ask for a BMC circuit that can manage so big Amp. I can estimate 0.5C of discharge but with i.e. this speed (0.5C) means (in a 50P 18650 2.000mAh/cell ) 50A. Is not a problem for BMS circuits that currents between serial banks while balancing reach 50A?

Other great thing is add new cells to the weak bank!!!

3.
Not to say ... Seems to be better welding/solder...


4.
You say 1 modular battery with 13S & [ 50P - 50P - 50P - 50P - 50P], with only one BMS and 24 pin conectors between each bank...!!!!
Its really sound fantastic Is like a 250P... but what about size of the cable of 25 pins conector... Is it sure a little pins and cable for so power big Amps?

I´m really worry about the security of system (basically burn my house), as it has to run automatically without supervision for weeks and weeks...

5.
What different point of view between solar experts and battery experts...
- Solar experts say to add lots of pannels, and are concerned about cost and realibility of batterys. Up today they dont see Li-Ion as alternative because cost (and probably safety), and you only see SLA stuff around this...
- Battery experts (you) say about the efficiency of batterys to "produce-store-provide"... Solar pannels are more and more cheapper every year, and problem is not the quantity to produce even the availabylity when you need. Usually solar people has lots and lots of KWatts that they cant use (cost of produced KW-h can be very very low), and then they dont appreciate battery efficiency. They even could preffer use 100KWh to charge 25KWh if battery was 1/10 cost and 10x realiability (clicles). SLAs has only 500 clicles, and this is very important on the total cost of the stored KW-h.


Sincerely . Thanks a lot for your great help....

"Bad" cells, typically, just lose function.
But, occasionally, can produce heat!
During original cell testing I have found cells that overheated during charging but, with many hundreds of cells over many years I have never had a cell turn "thermal".
Monitoring should include a thermal monitor alarm above pack in a fireproof unventilated battery "box".
Unlike SLA batteries Li-ion cells produce negligible heat, under moderate charge-discharge rates.

"Poor" cells can slowly bleed down voltage.
A "maintenance schedule" of semi-annual cell retesting should be employed.
Pick up a IR (infrared) meter <$20, and scan over cells to pinpoint any thermal rise. Replace any cells showing a higher temperature than its neighbors, "hot" cell will conduct heat into it's immediate neighbors. Heat production is typically a progressive defect so replace "bad" cells ... before they become a problem!

BMS typically monitors each bank for HVC (High Voltage Cutoff)and LVC (Low Voltage Cutoff) protecting against over charging or over discharging at the bank-cell level.
Most units will "balance-equalize" the peak charged voltage of all cells. Usually this involves bleeding down any bank-cell that attains a higher voltage charge than the others.

~20 pin connector to attach 13s pack to 13s BMS (14 wires needed.)
3 wires to positive, 3 wires to negative, 12 wires to extremely low drain BMS function.
Usually separate HD connectors for Pos Neg connections tho.

Ultimate safety for battery location?
How about an external power station - concrete pillbox.

Planning electrical use is important! Minimize battery use.
Time larger hot water tank for daylight heating only
Laundry, cooking, EV charging during daylight hours etc.

Thanks a lot again...

What a great ideas...

- Pick up a IR (infrared) meter <$20,

- How about an external power station - concrete pillbox.

- How about an external power station - concrete pillbox.
Better safe than sorry.
I had a battery fire of cells just sitting in my garage. Not being charged just sitting for 6 mos. and thremal runaway and fire out the vents. They where used cells from makita packs. Of 530 cells just gave all to the recycler.

Love this thread. Filled with links and pics for a future reference. I have one question. What are the best choices for a "spray-on" or a "brush-on" insulation. for situations like in this pic? I'm assuming that a flexible sealing insulator is better than a rigid one that would develop cracks, but I'd like a list of several verified choices by those who use applied insulation...

spinningmagnets said:

Love this thread. Filled with links and pics for a future reference. I have one question. What are the best choices for a "spray-on" or a "brush-on" insulation.

Click to expand...

Plasti Dip

Plasti Dip electrical tape spray

Tho ... Personally, I would just cover it with a layer of non-conductive closed cell foam and a single strip of duct tape ... till confirmed of impeccable reliability.

See the big list -eBike Toolbox - Bargains

Building recycled packs effectively-safely requires capacity testing large numbers of cells.
This is a time consuming tedious process ... so I decided to develop a method for testing large numbers in a reasonably simultaneous manner.
The development of a process is still in progress but the below link leads to a thread that documents my step by step development.
First test processed 60 cells with a fairly accurate comparative capacity in a few hours.
Second test processed 48 cells, with accurate mAh rating, for each paired cells, in about 4-5 hours.
Process refining anticipates accurate individual mAh rating of paired cells in batches of 60 cells in 2-3hours.

See - Bulk Capacity Testing

knighty said:

I went a little crazy buying cells, and have about 6,000 used laptop cells here to charge/test :-o

got a psu and loads of cell holders, so I can charge 100 at a time

Click to expand...

Are you sane? :lol:

Wow, if you survive w/o burning down first, then you have already achieved a lot. Seriously though, I wish you well. Are these all of the same type/brand so you have matched cells/Ah or ?

It will be interesting to see how you settle on some sorting/screening method to match these up.

deVries said:

knighty said:

I went a little crazy buying cells, and have about 6,000 used laptop cells here to charge/test :-o

got a psu and loads of cell holders, so I can charge 100 at a time

Click to expand...

Are you sane? :lol:

Wow, if you survive w/o burning down first, then you have already achieved a lot. Seriously though, I wish you well. Are these all of the same type/brand so you have matched cells/Ah or ?

It will be interesting to see how you settle on some sorting/screening method to match these up.

Click to expand...

Developing a simple-fast-accurate method ... right now!
See - Bulk Capacity Testing
Updating-perfecting method continuously.
Stay tuned - Step by step instructions of method ... available soon.

TIME STUDY...

Could you guys that are building packs do a time study too that will give us an indication of the time commitment to sort, test, and screen cells for assembly. ONLY that phase of the building your battery is sooooo time consuming, so it would be great to be able to know approximately how fast one method compares to another AND the overall time commitment involved with matching cells.

Thank you very much!

fastest pack is soldering china pouches like this

this pack created 2 years ago, still working good

DrkAngel,

When I was building packs using recycled cells, for matching them I tested large numbers of them at once to determine relative capacity. They were blocks of 2p cells with each 2p nominally 3ah. I paralleled all of the same size blocks to charge all cells to an identical voltage. Then I connected them all in series and using a series string of incandescent light bulbs I discharged 2ah. Then I separated them and measured their ending voltage and assigned a grade to each block of cells. Those with the highest voltage had the highest capacity and got an A grade for example.

Just Measured up a plan for a 9s12p eZip pack rebuild.
33.3V 31.2Ah = 1039mAh = 1kW pack!!!
vs OEM SLA 24V x 10Ah = 240mAh = (180mAh usable))

Importantly, should work well with 36V SLA controller!
TNCscooters.com has eZip scooter controllers and matched 36V throttles that match the eZip bike connectors.
About $55, but will lose PAS function.
450w / 24V x 33.3V = ~625W motor output!

Perfect for my pedal eBike - 2008 Trailz with 22mph, motor only capability, from 25.9V, some assist to 25mph.
25mph / 25.9v x 33.3V = (some assist till) 32mph!
33.3V, low rolling resistance kevlar belted tires, aero bars and 90rpm through 44-11T sprockets ... 30mph ... begins to look feasible ... with heavy assist.

Which brand/model and capacity cells print and encode on the cell itself the date of manufacturer, so one can easily understand or decode and decipher to determine when that happened???

Or, does the battery shell-casing for the cells have a date of manufacture or part number to locate that info?

Other ways to narrow this date down too?

Thanks.