*Dusting off the 1000lb of laptop batteries

[Skrzypas]

It would be convenient to make a bulk capacity test (balance charge, bulk discharge) on a laptop battery without opening it.

In this way you could save time on destroying a battery with all bad cells. Also this would severely increase safety of shorting bare cells lying on the table.

Anybody understands those schemes?

 

[agniusm]

Doesn't allow URL without www. Try this one (just copy paste):

sbworkshop.com

Hmmmm, but I doubt it will show a tested capacity figure? Only what is left with the current state of charge?
But I can see it to be helpful to know condition of pack before opening, like detecting all cells are under 2v etc

What do you mean you doubt? It shows remaining capacity of your cells, cycle count, manufacturing year, none of that information you will get by ripping apart those batteries and testing them with RC chargers which might be off a lot.
Voltage alone is nothing.

This is what you need to read battery state:

 

[Skalabala]

But it does not show the most important? What is the capacity from full to empty at a +-0.7C discharge?
Only show what is left as it sits in your hand?
I am not arguing, I just want more facts and info on this as it seems like a good idea

 

[agniusm]

What good that will do to you? Whats the point in having .7C discharge data when cells are meant for .5C and in most cases .2C discharge? For fact you will get less out of the cell the higher the discharge rate. There is no need to prove what is known. What you want is to sort out used cells and match them so you could make less dangerous battery pack. From my point of view, using laptop cells in ebike application is pure nonesense, i just thought why people dont use tools that are awailable and cost next to nothing

 

[dequinox]

After over a half year I'm back on working on these packs, and I got some help in the mail... and have more help on the way!




Having the Opus for this will be nice. I've played with it a little already... it does charge, discharge capacity test as well as an IR test! I'm able to discharge capacity test 4 cells at a time (8 when the foxnovo arrives), and that's a hell of a lot more than my Accucell will do. I'm hoping to build a pack in the neighborhood of 8s5p at first, then move on to something bigger.

I'd like to see what BMS Damien Rene is using... that series of build videos of his on youtube are an inspiring experience.

 

[Skalabala]

That is cool that it can discharge 4 cells at a time! At what current can it discharge?
I got me a few of these
http://www.ebay.com/itm/5V-Lithium-...542271?hash=item1c57b4aabf:g:4iAAAOSwP~tW3B6a

 

[dequinox]

That is cool that it can discharge 4 cells ...

It's discharge current is up top 1A per cell. I got this off gearbest.com, but you can also get it on Amazon and eBay. Opus BT-C3100 version 2.2. This one cost about $34 I think, and has free shipping. It comes from China so prepare for a little wait... I ordered April 14 and got it the 26th.

 

[Skalabala]

I wonder how it will get rid of +- 15W heat

 

[dequinox]

It's got an on-board cooling fan with a thermostat switch, so it turns on automatically. It's not too noisy for the most part. I don't hear it anyways since it's in my shop across the yard.

 

[Skalabala]

Thats awesome. A few of them will be great

 

[dequinox]

My final bit of equipment came in for processing all those cells I bought last June.
Equip list:
OPUS BTC-3100
FOXNOVO F4-S
2 LIPRO 6S CHARGERS
TURNIGY ACCUCELL8
SHITTY MULTIMETER (soon to upgrade)
BUTTLOAD OF 18650 HOLDERS
BUTTLOAD OF DOLLAR TREE BINS (perfect width for these cells, btw)


[youtube]HLgUEQ5pgSc[/youtube]

 

[Skalabala]

Great update bud Thanks!

 

[markz]

Love it, you mention one is $6 cheaper but never mention the price for the OPUS.
Love it. Thanks bro!

 

[dequinox]

Love it, you mention one is $6 cheaper but never mention the price for the OPUS.
Love it. Thanks bro!

Try Google... There are several different prices depending on where you're willing to shop, how long you want to wait, and what you're willing to pay. gearbest, Amazon, eBay, and I'm sure some other sites have it as well. I paid $34 on gearbest.

 

[dequinox]

Ok after a few days screwing around with the foxnovo I've concluded that I wouldn't recommend it for this type of battery recovery task. Its instructions are non-existent, it takes a LONG time to discharge, and it has some automatic features like where it will start charging while you have it on capacity-measure mode. If you play with it a few days you'll realize that it is filling the battery to 100% full before it does the full discharge test. None of this is really clear in the box-side directions.

I'll probably end up keeping it, as it does the job...albeit slower than the opus. I just wanted to warn anyone thinking of buying the Foxnovo off of it if they are just trying to save a few dollars. It's worth the extra $3-6 to upgrade to the opus... the features are a leap ahead of what the Foxnovo offers.

 

[dequinox]

So I've processed about 3/4 of a 5g bucket of these laptop batteries, and I've been IR testing them in pairs. The book on my Opus says to expect the real internal resistance of the battery you are testing to have a value of about 30mOhms less than the number it give you, so that is what I go by in my head, but i write down the number it gives me on each cell. My threshold for chucking a cell aside has been 100mOhms for a single cell.

Then I remembered my electronics courses... 1/R_total= 1/R_1 + 1/R_2 + ... so if you assume the cells have the same resistance in a soldered pair, then if you work out the algebra R_total=2*R_1. That means each cell has twice the resistance of the value measured by the opus...in theory.

I decided to check. I charged up & IR tested a pair of cells (LGDAS318650) and got these values:

100, 102, 100

Then I split the cells and got:

166, 161, 167 on one, and 171, 175, 175. Not exactly the same, but pretty close. This is not twice the measured value, but it is higher...

but then if you take that 30mOhms off for the real IR value of the battery, you get that the pair's measured resistance is about twice that of the individual cells (164.6-30)/2~=70... so that theory holds pretty closely.

Any thoughts from those watching this on what is an acceptable IR value for a single cell? I thought 100 milliohms seems like a logical cutoff for an ebike pack. It seems new NCR cells come in at around 50mOhms or less, so I kind of arbitrarily doubled that and went for it.

 

[DrkAngel]

IR value is not, by itself, a good measure of battery condition.

Typically, the higher capacity type cells will have a higher-worse IR!
That is why I recommend a "comparative IR" among same brand and model cells

 

[Ykick]

Any thoughts from those watching this on what is an acceptable IR value for a single cell? I thought 100 milliohms seems like a logical cutoff for an ebike pack. It seems new NCR cells come in at around 50mOhms or less, so I kind of arbitrarily doubled that and went for it.

For single “series string” DC IR (I do it the manual way with a 10 sec load) 20mOhm is about where I cutoff eBike use. Still useable for other low current projects/duties or parallel together for range extenders, etc.

So, if you had 5P (parallel) each cell about 100mOhm that would put you in the 20mOhm IR range for that parallel group.

I prefer effective pack DC IR around 5-10 mOhm. That’s a nice perky ride on 30A controllers.

IR is really just the amount of “voltage sag” so it does not relate to capacity or self discharge. Those being other important parameters which need to be tested/qualified on their own.

 

[dequinox]

Any thoughts from those watching this on what is an acceptable IR value for a single cell? I thought 100 milliohms seems like a logical cutoff for an ebike pack. It seems new NCR cells come in at around 50mOhms or less, so I kind of arbitrarily doubled that and went for it.

For single “series string” DC IR (I do it the manual way with a 10 sec load) 20mOhm is about where I cutoff eBike use. Still useable for other low current projects/duties or parallel together for range extenders, etc.
So, if you had 5P (parallel) each cell about 100mOhm that would put you in the 20mOhm IR range for that parallel group.
I prefer effective pack DC IR around 5-10 mOhm. That’s a nice perky ride on 30A controllers.
IR is really just the amount of “voltage sag” so it does not relate to capacity or self discharge. Those being other important parameters which need to be tested/qualified on their own.

Good to know, and I will be building about 8P sets to put in series. I plan on manually testing the IR of the whole series string at the end, and even testing each parallel string. My first pack has a 7 (in reality probably 10A max) controller it will be feeding, so hopefully IR won't be such a significant issue.

I need to become a little more diligent about comparing measured capacity with the original spec sheets, but I can't seem to find data sheets for everything. Happily they are available for most cells.

 

[cal3thousand]

The deeper in parallel you build these packs, the better the cells will perform. These could do well in a cargo bike that has the room for a ton of batteries.

 

[DAND214]

The deeper in parallel you build these packs, the better the cells will perform. These could do well in a cargo bike that has the room for a ton of batteries.

But this is only a half a ton :lol:

Dan

 

[dequinox]

The deeper in parallel you build these packs, the better the cells will perform. These could do well in a cargo bike that has the room for a ton of batteries.

But this is only a half a ton :lol:

Dan

I know not nearly enough right?! Would you believe I was contemplating going to this recycler to get another load? I still have plenty to work with...

 

[DAND214]

have done that but not that much. Found NOS HP 12 cell laptop batteries a couple years ago. Was a bitch taking the case apart. Good thing was all the cells were all usable and still like new. Got 2 - 12s packs 1 12ah and the other is 16ah. Not used much as the wife doesn't ride much. 16ah wasn't enough for me as I like a little more umph than they can put out. Tested them over the winter and they are still at full ah, 12 and 16ah.

Good luck with your build.

Dan

 

[DrkAngel]

I recommend a large enough pack to limit:
surge demand to ≤1C
continuous demand ≤0.5C

But, especially important with recycled cells, 1C should be rated at actual (present) capacity, rather than oem rated (new) capacity.

I lucked into some HP 12 cell packs, at nearly the same time, and was able to build a 9s12p 33.3V 31.2Ah battery, inside an eZip RMB (ReMovable Battery) pack.
Cells tested at >95% of new capacity for 30Ah+ for use with my typical (cheap) 28A controller.

 

[DrkAngel]

I recommend a large enough pack to limit:
surge demand to ≤1C
continuous demand ≤0.5C

But, especially important with recycled cells, 1C should be rated at actual (present) capacity, rather than oem rated (new) capacity.

I lucked into some HP 12 cell packs, at nearly the same time, and was able to build a 9s12p 33.3V 31.2Ah battery, inside an eZip RMB (ReMovable Battery) pack.
Cells tested at >95% of new capacity for 30Ah+ for use with my typical (cheap) 28A controller.

• 
  Regeared 1 eZip for 20mph "legal" 20mph with 33.3V battery.
• OEM eZip 1 eZip (9T motor x 20T wheel) for ~16mph @ 24.0V (450w motor output) ... 100% Torque - 100% speed
• "My use" is 1 eZip (9T motor x 16T wheel) for ~20mph @ 25.9V (485w motor output) ... 80% torque - 125% speed
• Regeared 1 eZip (9T motor x 22T wheel) for ~20mph @ 33.3V (625w motor output) ... 150%+ torque - 125% speed

Nearly double the torque of my typical cruising mod!

Reasonably projected MY1018z (or XYD-16) 24.0V vs 33.3V when geared for 21mph

("gearing" accomplished by using comparative wheel size function) - wheel size not actually used, eZip rigs allow multiple gearing options!

Just for fun ...
33.3V with "my use" gearing (9T motor x 16T wheel) for ~26mph @ 33.3V (625w motor output) ... 110%+ torque - 162% speed
but ... tried it a few times to get it out of my system and rejected as being abusive of battery and attracting bad attention ...