Homemade Battery Packs

DrkAngel:

I'm going to take delivery of 1000lbs of laptop batteries on a pallet this week, and intend to tear down and reclaim at least some of that quantity. I'm not really set-up to do massive testing of 18650 and prismatic cells at the moment, but would like to be able to do it slightly faster than 1 at a time. I have read through your guide about reclaiming cells once, skimming a bit, and will read it again undoubtedly, but I am not sure how you test to many in a shorter time span.

All I have now is a Accucell8 and a few alligator clips. Not ideal I know, and I'm trying to keep the cost of this project down.

If there is any basic equipment I should get, that isn't going to rape my wallet, that would be good for this process please let me know!

I've seen some people using multiple hobby king/ ichargers at once, I've sort of seen DoctorBass's A123 reclaiming setup... although I haven't a clue how he controls it, and I have a few ideas how I would do this, but really I'd like to get a few tips from someone who's done this first. I've also watched Rinoa Super Genius' youtube channel... and he has a few good tips.
 
Thanks, I had not seen that part of the thread. I also briefly looked at your bleed-down testing post. That's good information to have!

I'm thinking I'll make up a testing rig with spring-connectors out of wood or some-such for easy cell removal/replacement.... perhaps for 20P or so.
 
dequinox said:
Thanks, I had not seen that part of the thread. I also briefly looked at your bleed-down testing post. That's good information to have!

I'm thinking I'll make up a testing rig with spring-connectors out of wood or some-such for easy cell removal/replacement.... perhaps for 20P or so.
I'm planning on Neodymium magnets and paper clips ...
 
DrkAngel said:
dequinox said:
Thanks, I had not seen that part of the thread. I also briefly looked at your bleed-down testing post. That's good information to have!

I'm thinking I'll make up a testing rig with spring-connectors out of wood or some-such for easy cell removal/replacement.... perhaps for 20P or so.
I'm planning on Neodymium magnets and paper clips ...

The old tricks are the best tricks, eh? :lol:
 
1000 lbs? That's a boat-load. I might be interested in some of those depending on the price and types of packs there.
Do you have any idea what you'll be getting? :lol:
 
Check list for bulk testing
Multi-Meter
MeanWell S-150-5(w/Mega-Mod - optional)
V-A-W-Ah meter w/external power (or dual 60w light fixture)
Jumper wires w/alligator clips
Neodymium magnets (100?) and a box of steel paper clips make a great tool.
Bleed down device - to lower bank voltage (12V headlight etc.?)

Handy
MeanWell S-150-12
Infa-red meter

Check List for building
~14 & ~10ga tinned copper braid (dependent on pack size and power demand)
≥50W soldering iron, 60/40 rosin core solder & soldering flux
Small chrome plated probe (screwdriver?) to stabilize soldered components
Clear 2" boxing tape
Duct tape - 3" preferred

Will update - fill-in after coffee!

PTR! (Points To Remember)
Test for self-discharge!!!
Leave cells with tabs connected! (wires too!) - makes testing and building much easier!!!
Cells of 0V are "iffy" at best. You can try to charge ... but monitor closely!
 
jaunty said:
1000 lbs? That's a boat-load. I might be interested in some of those depending on the price and types of packs there.
Do you have any idea what you'll be getting? :lol:

Out of a sample box it looks like a fair portion of them will be Dell, but I can't say I know more than that at this point. I"ve seen some prismatic formats like Apple batteries and such, but haven't dove into the deep end yet.

I take delivery today so I will let you & others know via what will be my new thread in Battery Tech. I figured if I'm going to be building some big battery packs I should probably document a little.
 
dequinox said:
jaunty said:
1000 lbs? That's a boat-load. I might be interested in some of those depending on the price and types of packs there.
Do you have any idea what you'll be getting? :lol:

Out of a sample box it looks like a fair portion of them will be Dell, but I can't say I know more than that at this point. I"ve seen some prismatic formats like Apple batteries and such, but haven't dove into the deep end yet.

I take delivery today so I will let you & others know via what will be my new thread in Battery Tech. I figured if I'm going to be building some big battery packs I should probably document a little.


Pictures on first sight please! :D
 
In order to stop polluting DrkAngel's DIY battery pack thread (Sorry DrkAngel!), to discuss the "1000lb of laptop batteries" topic please go HERE.

If you want to get a piece of the pie, click HERE.
 
DrkAngel said:
dequinox said:
Thanks, I had not seen that part of the thread. I also briefly looked at your bleed-down testing post. That's good information to have!

I'm thinking I'll make up a testing rig with spring-connectors out of wood or some-such for easy cell removal/replacement.... perhaps for 20P or so.
I'm planning on Neodymium magnets and paper clips ...


I'm guessing you know this.. but just in case you (or others) dont, you can salvage Neodymium magnets from dead disk-drives -- often avaible at the same places one gets old laptops.

http://www.instructables.com/id/Pulling-apart-a-desktop-hard-drive-to-get-rare-ear/step8/Getting-the-magnet-of-the-brace/
 
DrInnovation said:
DrkAngel said:
dequinox said:
Thanks, I had not seen that part of the thread. I also briefly looked at your bleed-down testing post. That's good information to have!

I'm thinking I'll make up a testing rig with spring-connectors out of wood or some-such for easy cell removal/replacement.... perhaps for 20P or so.
I'm planning on Neodymium magnets and paper clips ...


I'm guessing you know this.. but just in case you (or others) dont, you can salvage Neodymium magnets from dead disk-drives -- often avaible at the same places one gets old laptops.

http://www.instructables.com/id/Pulling-apart-a-desktop-hard-drive-to-get-rare-ear/step8/Getting-the-magnet-of-the-brace/

Yeah I was tearing hard drives apart for these back in 2000, in high school.... Gack! :? I can't believe it's been 14 years since then...
 
Finally re-retiring my re-re-re-recycled 33.3V 31.2Ah 18650 pack.

33V_31.2Ah.JPG


31.2Ah capacity has degraded to ~10Ah.
I had intended on saving for my Snow Beast but couldn't restrain my need for speed this Summer and wore 'er down.
10Ah actual capacity, with >25A peak draws (2.5C from laptop cells *) was too hard on the old gall ... damaged 1 bank.
Going to retest remaining cells and build into 12V inverter power pack.

Good news is ...
I can drop my NOS Sony & Samsung cells into same eZip RMB pack for a quick replacement.

file.php


For a true - metered 33.3V ~30Ah capacity!

* Observation, (heat production & deterioration rate etc.), seems to prove out that C rate is more directly related to (a function of) actual capacity, rather than original rated capacity!
 
DrkAngel said:
* Observation, (heat production & deterioration rate etc.), seems to prove out that C rate is more directly related to (a function of) actual capacity, rather than original rated capacity!
An attempt to clarify my observations:
33.3V 31.2Ah build consisted of 90 severely used cells and 18 nearly full capacity rated cells.
Build capacity was an actual ~20Ah.
~100% capacity cells in parallel with ~60% capacity cells look to have deteriorated the same percentage!
Based on cells labeled capacity (from pre-build capacity test) ~2500mAh cells now meter ~1250mAh matching the entire packs deterioration, indicating the same 50% deterioration for the severely used cells.
Basing C rate on actual capacity ... the cells functionally discharged and recharged at the same relative C rate and so deteriorated identically!

"31.2Ah (~20Ah actual) capacity has degraded to ~10Ah."
"10Ah actual capacity, with >25A peak draws (2.5C from laptop cells *) was too hard on the old gall ... damaged 1 bank."

Am in the process of retesting remaining cells for 12V inverter pack and it looks that Comparative IR, among same model cells is a reasonable indicator of capacity.
48 x 2p cells will be charged and re-tested for self-discharge then sorted into 3s32p, based solely on IR;
Tested by discharge;
Carefully measure resultant voltage of each bank;
Results to be posted.

I expect banks to prove out as of nearly the same capacity ... but will tune banks by swapping or adding cells, if necessary.
 
Does anyone have any recommendations on sourcing laptop batteries? Is it better to go with the new (non-oem) replacement packs, used packs, or?? So far I have found some new ones, 6 and 9 cell, for around $10 - $13 each. There is 11.1v 9 cell (it was rated 8200mah IIRC) but someone who actually tested the capacity of the cells reported it is truly 5400 mah (1800 mah per cell true capacity). I'm not sure if that is the best way to go because the 6 cells might be the same way, they rate them at 5200 or 5600 mah but they actually test significantly lower.

I haven't been able to find dirt cheap used batteries like a lot of you have done. It seems the lowest I can find a used pack for is about $7 shipped. They claim they are "slightly used" or "like new", but who knows. I'm not sure if a 30 to 40% savings is worth it if more of the cells are dead or low capacity and the cycles might be very high.

It seems like I would need to get them for more like $3/each or less to make it worth it.

I'm trying to build a 48v pack for my 1000w direct drive hub. After reading everything it seems like a 12S might be the way to go. I've read my LVC on my controller is probably 41V, which would mean I might want a 11S setup. Others have recommended a 13 or 14S which probably means I would need to add a BMS or something to prevent deep discharge then. My controller is the typical silver chinese one sold with the direct drive hub kits on ebay (which is what I bought). I don't have it yet to know if it has adjustable voltage on it.

What do you all think on the series size I should use? Not so much worried about getting the max top speed as much as just keeping the batteries in good shape.

Thanks
 
It takes some patience and searching. Back in January I found 3M projector batteries on Amazon. It took weeks of poking around and reading to find them but 3 batts for $1.06. I bout ten packs for $10 shipped. BUT they aren't good for much. I'm using them to learn welding techniques and making my F ups with cheap batts before I jump into the 29E for $3 each from Tumich.

Patience.
 

Attachments

  • Screen Shot 2015-09-11 at 2.46.00 PM.png
    Screen Shot 2015-09-11 at 2.46.00 PM.png
    9.3 KB · Views: 5,914
Relegating, not retiring, Spring of 2014 Lipo builds to lighter duty applications.

2 - 25.9V 25.92Ah (7s12p 2160mAh cells) = 671wh
Both packs were built from recycled Dell CD bay auxiliary batteries of very good condition. Actual metered charges, from modest discharge depths were 90% + of new rating.
Each has cycled hundreds of times for resultant thousands of miles on eZips geared up for 21-22mph from motor only.
Initial 13T motor sprocket upgraded eZip of near 25mph capability was too stressful for these packs, so I "relented" and moderated back to the 11T.

file.php


file.php


This still exceeds my recommended .5C discharge rate for these cells and actual capacity has degraded ~20Ah ... which nears a 1C discharge rate when cruising at ~21mph.
Rather than running till dead at this high, for laptop cells, discharge rate I am relegating packs to my oem geared eZips. ~16mph motor only speeds
25mph requires 600w motor output (~750w battery input) - 28A
22mph requires 425w motor output (~530w battery input) - 20A
21mph requires 380w motor output (~475w battery input) - 18.4A
20mph requires 340w motor output (~425w battery input) - 16.3A
19mph requires 300w motor output (~375w battery input) - 14.4A
18mph requires 260w motor output (~325w battery input) - 12.5A
17mph requires 220w motor output (~275w battery input) - 10.6A
16mph requires 190w motor output (~237w battery input) - 9.1A
I anticipate many years of satisfactory service, as opposed to the comparatively rapid degradation at present demands.

Watts_%3D_Speed.jpg


Replacement builds will be:
(All in eZip RMB battery cases)

#1 7s12p (25.9V 25.92Ah) = 671wh of NOS Sony laptop Lipo cells
As everyday battery compatible with all my eBikes except the Snow Beast.
And to evaluate the Sony cells, of which I have several hundred.

#2 9s8p 2160mAh (33.3V 17.28Ah) laptop Lipo
paralleled with
9s1p 2000mAh (33.3V 2Ah) RC Lipo
for a total
33.3V 19.28Ah = 642wh of mixed hybrid cells
For PA (Pedal Assist) use on several of my multi voltage capable eBikes and light duty on 36V controllered Snow Beast.
And to test evaluate RC + Laptop Lipo discharge capabilities.

#3 9s12p 2600mAh (33.3V 31.2Ah) 1038.96wh (1.038kWh)
Mix of NOS Sony and Sanyo cells
Dedicated for Snow Beast Winter use and performance "runs" on several multi voltage capable eBikes.

I also have a fair quantity of NOS 3s 5700mAh Winforce RC packs
Not sure if I will build a :

#4a 9s3p (33.3V 17.1Ah) pack = 567.43wh
As HD high speed or Snow Beast pack
or
#4b 3s9p (11.1V 51.3Ah) pack = 567.43wh
As serial pack to Turbo switch 25.9V >>> 37V (eZips hold 2 packs with switch that can be jumpered to provide Turbo)
and\or
As auxiliary pack for 12V Lighting, power tools, inverter etc.
 
I've got a question about the 1C discharge. Does it still hold for 18650s as they age? Will a battery that has 50% capacity be stressed by 1C vs a new battery?

Also, if 13s5p is officially spec at 13ah, but via cycle fatigue it is 9ah, is 13 amps truly safe to run continuous, or is 9 amps safe to run continuous?

I've built a 13s5p pack from recycled 18650s. I estimate it is around 10ah. Ran around with it for a few weeks, and have realized I've been pulling 20 amps near instantaneous and 15amps for 1 minute going up some hills. The battery pack was getting hot. Thus, I'm adding another 4 banks to make it 13s9p which is speced (9x2.6ah) out at 24ah, but I think it will be about 18ah. I would like to run it at 24 amps on the road, but if it is only 18, i need to know.
 
Never used an ESR meter.
I rate cells by:

#1 - Self discharge - eliminates most defective cells!
I charge cells precisely to rated, or my determined working voltage, usually charge a large batch in parallel.
Then I separate and let set for several days, looking for cells that lose voltage.
Any cell that self discharges is eliminated as defective and possibly dangerous.

#2 - "Comparative IR" (Internal Resistance comparison) - Gives rough capacity estimate
I apply a specific discharge (.5-1C?) for a specific period (1-5min?) and measure voltage sag before removing discharge.
Each cell or pair etc. is labeled with sag as a comparative IR.
With like cells, the comparative IR is typically a reasonable indicator of actual capacity.

#3 - Capacity
I either:
a. rig in series (28-30s) and discharge with 120w (2-60w light bulbs) for a 1000mA discharge rate
to determine cell comparative capacity. (EG 3h = 3000mAh + 3.78 remaining voltage)
b. rig in series ≤ 20s and discharge through a cheap 90V mAh meter.


This allows me to build packs without defective self discharging cells and banks of reasonable equal capacity.
Paralleling cells of unequal capacity seems to do no harm ... but banks must be equal capacity!

At moderate charge\discharge rates, and restricting peak charge and discharge depth ...
Banks remain equal, at charged and discharged voltages.
No cell properly intact has ever failed me in 8 years and 20,000+++ eBike miles.

Failures!
#1 cracked solder joints resulted in one bank excessively discharging = cooked 10 of 12p. Various similar occurrences.
#2 Broken parallel lead over-volted 4 of 12 Lipo cells causing hazardous puffing during bulk charge. Various damaged Lipo ribbon cables caused less severe-quickly noticed unbalances. Charging to ±4.10V instead of 4.20V left a large safety margin ... and prolongs battery life.
 
Sorry ... I personally do not use any form of BMS.

After stringent elimination of self-discharging cells and carefully capacity matching all banks, there is no need for any re-balancing function, banks stay balanced.
I use a simple bank level voltage meter w/alarm to monitor pack ... for nearing empty or any defect developing.

I have come to regard a BMS ( ± w/balance function) as an expensive security blanket?
or
As an expensive Band-Aid for a defective pack.
 
DrkAngel said:
Sorry ... I personally do not use any form of BMS.

After stringent elimination of self-discharging cells and carefully capacity matching all banks, there is no need for any re-balancing function, banks stay balanced.
I use a simple bank level voltage meter w/alarm to monitor pack ... for nearing empty or any defect developing.
So, how would you make a 10S10P pack using batteries from http://www.nkon.nl/ ? Where would you buy the rest of the electronics?
 
Golyo said:
So, how would you make a 10S10P pack using batteries from http://www.nkon.nl/ ? Where would you buy the rest of the electronics?

Since you are using new batteries run don't walk from this thread. Nobuo and others have created threads on battery building using new batteries. The advise here is more appropriate to playing with and recovering used batteries. The advise given for using BMS, for example, should be laughed off. If you are building a good pack with good batts, it's best to do the job as a professional would. All the material links and information is in the welder threads and in the Battery Technology section.


This thread has great links...
https://endless-sphere.com/forums/viewtopic.php?f=14&t=68005&hilit=Nobuo

Good luck!

Oh and be sure to look for Nobuo's youtube channel for blow by blow builds. Great stuff!
 
Back
Top