Begginers Guide to Processing 18650 Cells | Cell Database


Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
G'day from WA
#1
Hi Guys,
I held off introducing myself until I had something to show.
So, the easy part is now done, I've cracked open a few lappy packs and have harvested a few cells pictured below. Now it's time to test and sort.

[Image: IMG_6218_crop.jpg]

Small steps for this beginner; my primary motivation is to replace my aging camping solar batteries.
Our family camps 3 or 4 times a year and for the last 8 years or so I have been lugging around two 38Ah LA batteries in parallel.
They have served us well but are now starting to get a bit tired, especially with the electric blanket when camping in winter!
Also, at 26kg (57lbs), I'm tired of moving the damned things! even if it is only 3 or 4 times a year.

I had been looking at a replacement LA battery, a Fullriver 120Ah costing ~$400 and weighing a mere 37kg(82lbs), but then started searching for Lithium alternatives. Seeing the prices for a 120Ah Lithium(Fusion) floored me, I had expected prices to have dropped a little over the years but ~AU$1500 is insane.

That's when a little more research and learning about 18650's, well, evil plans started to brew.

Current plan is a 4s60p battery, so ~14.8 and ~120Ah if I'm lucky, and ~15kg (33lbs).
If I can keep costs below $1 a cell, so $240 for the battery (a little more than half the LA price) I'll be happy.
If I get more good cells from my harvest, all good, go for 80p or 100p, but I'm figuring conservatively at ~50% good/bad cells after testing.

What's after this?
Well, my old man is building a motorhome, which currently has 800W solar on the roof, if I can build an 18650 system for him he might save some money on fuel (instead of lugging around 200 kilo's(440lbs) of lead!) 

And after that?
An EV to drive to/from work would be really cool.... (I currently spend ~$4000 a year on petrol) ..... but let's not get ahead of myself!

Breakdown of my first haul in the photo:
There's ~430 cells ready for testing
- 157 (36%) of the cells are Sanyo, so will need careful scrutiny,
- 110 (26%) are Sony
- 71 (17%)  are Samsung
- 58 (13%)  are Panasonic
- 28 (6%)  are LG
and just 6 cells were no-name chinese cells.

So far that's 47 cents per cell, if I can get 240 usable cells after testing, which would be 84 cents per cell, I'd be happy with that.

Not pictured:
70 leaky/rusted cells
95 cells that measured 0.0v
71 cells that measured below 1.0v
So ~35% attrition before testing!
Some of the <1.0v cells might be recoverable but I'm not sure if I'll bother.

Enough blah blah.....
Reply


#2
Awesomeness Simon get into it. My testing regime went through 3 upgrades. Once I started testing I then calculated the time required to complete batches with the existing equipment and process. Lessening time required, looking for bottlenecks in the process.
Reply
#3
Cheers mate!
Cheers also for the packs, I have a feeling you'll be getting another call after I get through testing these Big Grin
Reply
#4
Welcome! Will be nice to see your progress
Follow me! >> YouTube / Forum system setup / My webpage  Diy Tech & Repairs

Current: 10kW Mpp Hybrid | 4kW PIP4048 | 2x PCM60x | 83kWh LiFePo4 | 10kWh 14s 18650 |  66*260W Poly
Upcomming: 14S 18650~30kWh | Automatic trip breakers, and alot more
Reply
#5
Thanks Daromer.

When I got home from work this evening I had a package waiting for me... more packs!
It was an ebay purchase so not the most economical method, but I think I did ok.

47 more cells ready for capacity testing.
[Image: IMG_6222.jpg]

Mostly LG and Samsung.
One 4-cell pack (Sanyo UR18650ZY 2500's) was dead, every cell was 0.0v
and a single 2800 cell (ICR18650-28A) was <1.0v, but the remaining 47 cells were all above 3.0v, and most are 3.5-3.8.
Feeling confident about these!

This is addictive Big Grin
Reply
#6
(09-01-2017, 03:38 PM)SimonW Wrote: Thanks Daromer.

When I got home from work this evening I had a package waiting for me... more packs!
It was an ebay purchase so not the most economical method, but I think I did ok.

47 more cells ready for capacity testing.

Mostly LG and Samsung.
One 4-cell pack (Sanyo UR18650ZY 2500's) was dead, every cell was 0.0v
and a single 2800 cell (ICR18650-28A) was <1.0v, but the remaining 47 cells were all above 3.0v, and most are 3.5-3.8.
Feeling confident about these!

This is addictive Big Grin

I have had quite a bit of luck with cells below 1v. Some of them have been my better cells. Just finished charging some cells that measured in the mV range. 47mV and there about. They charged just fine, no heat, no sign of rapid discharge. We'll see about self discharge in a week or so but.

Just charge them real slow, about 100ma, until they reach 3v, then 300ma until full charged. Cycle them, watch closely, some may get hot, and need to be discarded. Even on the first discharge, some can get a little warm, quite often the internal resistance will be high for the first cycle. When finished allow to settle for a week, check for self discharge, check internal resistance, charge, and capacity check.

If you don't want to mess about with cells below 1v, send them my way.   Wink  Ill be happy to pay postage, and give you a few $$ for your trouble.
Reply
#7
I have infact increased this to 1.5v.
I'm also weeding out everything between 1.5v and 2.0v.

I'll be focusing on testing everything above 2.0v for now.
Once that is done, I'll test the 1.5v to 2.0v cells, then everything between 1.0 and 1.5v.
For the <1.5v cells, maybe they'll be okay, but I'm not going to hold out much hope for them.

The manufacturers test labs are full of scientists and engineers that are much much much smarter than me, and have done a huge amount of research into this.
If their literature says anything that has spent considerable time below 1.5v is a risk, due to copper dissolution, then who am I to argue?

Quote:Just charge them real slow, about 100ma, until they reach 3v
Yep, have been pre-charging everything under 2.5v, so far so good.
Reply
#8
(09-04-2017, 04:46 AM)SimonW Wrote: I have infact increased this to 1.5v.
I'm also weeding out everything between 1.5v and 2.0v.

I'll be focusing on testing everything above 2.0v for now.
Once that is done, I'll test the 1.5v to 2.0v cells, then everything between 1.0 and 1.5v.
For the <1.5v cells, maybe they'll be okay, but I'm not going to hold out much hope for them.

The manufacturers test labs are full of scientists and engineers that are much much much smarter than me, and have done a huge amount of research into this.
If their literature says anything that has spent considerable time below 1.5v is a risk, due to copper dissolution, then who am I to argue?

Quote:Just charge them real slow, about 100ma, until they reach 3v
Yep, have been pre-charging everything under 2.5v, so far so good.

May as well give them a go. That's my policy. I usually keep notes. Often cells that are below 2v probably aren't worth putting in a power wall. Although, time will tell. I shall certainly be scrutinizing these cells more than others known to have been above that.

Edit:

However I do have far fewer cells available to me. So anything extra is a bonus
Reply
#9
Just been testing an OPUS and, well, so far I'm not impressed.
How can anyone trust these things?

I just picked a cell at random, a LGDAS31865, rated 2200mAh by the manufacturer. I capacity tested at 1A draw until 2.8v
My iCharger tells me it's 2023mAh which I can believe.
Same cell in the OPUS, 2303mAh.... what?!

In addition to that, the iCharger should have a figure higher than the OPUS because the OPUS is 1A CC drain with a hard cut-off at 2.8v.
The iCharger on the other hand is 1A CC until 2.8v, and then it switches to CV until the current drops to 500mA. In my experience this adds another few tens of mAh to the final figure compared to a hard cut at 2.8v.

I think these things were meant to be named UPOS not OPUS. Utter Piece Of Shit.
I'm thinking I'll knock-up a quick LM317 constant-current regulator and test the same cell the old fashioned way that I used to do with Sub-C sized NiCad's about 20 years ago, with a multimeter and a stop-watch.... will determine which is telling me porky's, the iCharger or the UPOS....
Taking bets?
Guess where my money is....

Edit: Forgot to mention, with the iCharger the cells don't even get warm but in the UPOS they're pretty toasty.
I'm thinking this is two things, first probably has to do with how the UPOS uses PWM at around 3.2 amps to average a 1A discharge... this high 3.2A current draw is likely heating the cells. I also suspect the load resistors are buried somewhere under cells 3 and 4 as they seem to get much hotter than the others.
Reply
#10
First of all the current drops to mA not mAH. Smile

No one that have talked to me trust the number as actuall capacity.

Yes the Opus does show a different valyue but the value is consistent. The consistant number is what you look for and then dont forget what you get for the money.

I have taken my numbers from my Opus chargers. Tested >2000 cells that are in production. I have then tested the same cells in the Powerwall and the capacity was damn close to reallife when done in terms of %. I dont take the numbers from the Opus as mAh and I have said that 100x on the forum before. All chargers does it differently and in this case only take the number as s number.
I would love to have 20 iCharger Duo on my table but I have 1 Duo sitting there for ref meassuring. I instead I got 10 Opus chargers with some minor calculation are able to sort my cells into same bin as the iCharger would have done Smile

It is what they say. You get what you pay for. But as long as the numbers are close enough to continous it will work and as long as you are aware of it.
Follow me! >> YouTube / Forum system setup / My webpage  Diy Tech & Repairs

Current: 10kW Mpp Hybrid | 4kW PIP4048 | 2x PCM60x | 83kWh LiFePo4 | 10kWh 14s 18650 |  66*260W Poly
Upcomming: 14S 18650~30kWh | Automatic trip breakers, and alot more
Reply


Who read this thread?
41 User(s) read this thread:
Matthew Oliver (Yesterday, 03:03 AM), karrak (09-12-2017, 03:21 PM), JamesD (09-05-2017, 08:36 AM), lucapgt (09-01-2017, 12:25 PM), Learjet (09-10-2017, 01:25 PM), vtpolak (09-01-2017, 04:27 PM), paulsramon (09-09-2017, 11:28 AM), SimonW (09-13-2017, 08:53 AM), floydR (09-01-2017, 11:01 PM), BJcell (09-04-2017, 11:38 PM), Swake (09-05-2017, 09:23 PM), Stephen Jobson (09-20-2017, 12:59 AM), priba83 (09-13-2017, 08:54 AM), jens18 (09-12-2017, 05:42 AM), NicZa (09-12-2017, 03:52 PM), osweld (09-11-2017, 03:54 AM), danipg (09-08-2017, 12:32 PM), Grumplestiltskin (09-06-2017, 01:44 AM), nautal (09-14-2017, 05:10 AM), Con Magourilos (09-10-2017, 01:02 PM), Geek (09-13-2017, 02:17 AM), Chris Davis (09-02-2017, 03:46 AM), Darmok (09-12-2017, 02:04 PM), Batrium (09-01-2017, 09:39 PM), A + A (09-12-2017, 01:04 PM), ruby2 (09-13-2017, 09:26 AM), Majorphill (09-13-2017, 01:32 AM), ken (09-04-2017, 01:42 PM), owitte (09-05-2017, 10:41 AM), daromer (09-12-2017, 02:07 PM), HogensHero (09-09-2017, 06:01 AM), jdeadman (09-12-2017, 06:59 PM), KtB (09-02-2017, 10:53 AM), Erik Smith (09-09-2017, 12:34 AM), Stefaan De Ridder (09-05-2017, 07:58 AM), MetalFang (09-05-2017, 04:53 AM), Hanssing (09-12-2017, 05:27 PM), watts-on (09-12-2017, 09:50 PM), frnandu (09-14-2017, 05:33 AM), wim (09-12-2017, 05:41 PM), hbpowerwall (09-14-2017, 12:11 AM)

Forum Jump:


Users browsing this thread: 1 Guest(s)