Testing 18650's for performance

lionman

100 W
Joined
Jun 7, 2018
Messages
182
So i'm planning a battery build, a 14s6p pack, and so ordered 84 HG2s which arrived yesterday. I took a bit of gamble and ordered them from dhgate. :pancake:

I made the mistake of researching fake cells after ordering them and found out that HG2s are pretty much the most faked cell out there. Woops! :oops: The cells look new and appear of reasonable physical quality from the exterior but they do have a tell tale sign of not being the real deal in that the ring insulator is very shiny plastic, where the originals apparently have a matte/textured one. It's not definite though. I haven't removed any wrapps yet but I may do to check the can itself.

I tested the voltage quickly on about 20 of them and they were all within .01V of eachother. I have an Opus BT-C3100 arriving soon which is capable of a discharge capacity test so it will be interesting to run a few through that.

If they have a reasonable capacity then I will probably still use them. My controller is 45A so each cell should only see about 7.5A. I was hoping for a lot of headroom in terms of power so the cells had an easier life but as it turns out, a lot of the fake HG2s are more comfortable at 10A.

What would be the easiest way to test a cell at 10A?

I could hook up a voltmeter and temp probe to a cell and put a 0.37 ohm load on it? In the absence of a data logger, I could monitor closely and record the temp and volts each minute until it reaches 2.5v? Would take less than 20mins I think, (18 would be good :wink:)

I could do it first on a light load, maybe 1 ohm.

I just want to be sure they are reasonably happy at 10A so I can have some faith in them performing ok in a pack.

Is there another way short of spending $$$?

Am I overthinking it?
 
Since 12 VDC lights and accessories are commonly available at a variety of watts, I'd probably make a simple 3s2P pack or something similar and do a discharge test at the standard discharge rate (600ma per "p") first. In the case of a 3s2p pack, that would be something like a 15 watt light bulb. If the cells test pretty close to their 3000 mah capacity, then I'd look into testing at a higher discharge rate. For this same pack, that would mean something a 12 volt light bulb of about 120 watts.

Of course, if you have the proper test gear that can apply specific loads, then you wouldn't have to rely on using commonly available 12 volt stuff. But I'd probably still do a first test to check on the overall capacity at the standard discharge rate. It is easier to do that safely and might give you a quick answer to whether or not these LG cells are legit or not.
 
Commonly available means something aittle different in Perth Western Australia. Haha.

I have some spare headlight bulbs for the car though so that could work. I think they are 55w each so I could test at 55w first and then run two in parallel for a second test. ~110w at 12v will be more than the cells would have to deal with in my bike. If they handle that ok then they should be good enough.

Hard to get affordable cells here.
 
So not looking good. Tested 4 cells in the Opus. 1800mAh! So pissed right now! :evil:

Does anyone know how accurate this charger is? I noticed it stopped the discharge test at 3v, which is better for the cells, but there must be 500mAh or so left between 3.0v and 2.5v.

I also set it to 1000mA setting which may have bought down the capacity a bit. I might try again after resting the cells to see if the numbers are different. Doubt it would have that much impact though.

IMG_20180616_164128.jpg
 
Did you fully charge them before discharge test? Your bms and/or controller will cut power around 3v, so you won't be able to use the remaining capacity down to 2.5v.
 
Yeah, in the test mode the charger charges them to 4.2, discharges and measures capacity then recharges to 4.2 again.

I realise the bike will only see capacity to 3v but trying to work out what the likely capacity of these cells really is. I might test some known quality cells I have from my vape mods and see what they test as. I have some VTC6 and 21700A to try out.
 
I have used an imax b6 for most of my battery testing and I would say I've seen pretty close to full capacity down to 3 volts, but worth you doing the same with some other cells for a comparison.
 
The stickers on those cells scream fake looks like you got screwed. Fortunately you paid with a credit card, paypal, or some other legit service right? In that case you would be able to do a chargeback / claim and get a full refund.
 
lionman said:
Yeah, in the test mode the charger charges them to 4.2, discharges and measures capacity then recharges to 4.2 again.

I realise the bike will only see capacity to 3v but trying to work out what the likely capacity of these cells really is. I might test some known quality cells I have from my vape mods and see what they test as. I have some VTC6 and 21700A to try out.

If you look at the chart I've linked to below, the HG2 has about 200-250 ma capacity between 3.0 and 2.5 volts at the standard discharge of 600 ma or .2C. The capacity per volt of most Li-Ion 18650 cells starts to become non-linear, reducing capacity below about 3.4 volts and really starts to reduce at around 3.25 volts. By the time they reach 3.0 volts, the reduction per volt is pretty dramatic. Judging by the HG2 graph and my own tests of my LG MF1 cells, there's probably 2.5% to 5% of the cells capacity left below 3.0 volts and whatever the cell's specified cuttoff is. It really isn't worth worrying about.

https://www.nkon.nl/sk/k/hg2.pdf
 
I agree those stickers make me think fake!
Is there actual printing on the cells?

I've tested hundreds of recycled HG2 among others, on the opus. You should be definitely getting more than 1800.

Most used cells I test @ 1 amp still have around 2900mah left.
 
Here is a shot of my known good performing VTC6s. The pickle ricks are about 2 years old and have copped a flogging. The pizzas are about 1 year old and have also worked pretty hard. The wraps where installed by me after the original ones where damaged during regular use.

IMG_20180617_080022.jpg

These seem to still have a decent capacity after a lot of use. VTC6 are a 3000mAh 19A cell. I bought these locally though for $10AUD each.

The sticker on the HG2 is no indication of them being fake. Even gunuine cells have them, they are put on there by battery exporters, something to do with import/export laws. They have marking like HG2s under the stickers.

IMG_20180617_081643.jpg

I have handled plenty of 18650s and these look very convincing. The only indication they are shit is where it counts. Performance!
 
So I got a bit over a 50% refund for these cells. Bit of a bummer I have ended up paying for a bunch of pretty crappy cells. You live and learn I guess. Thinking or using them in a 4S21P config in a solar generator for camping. Would be great for lights and ac adapters etc. Might even run a toasted sandwich machine. Glamping!

I have ordered some 35E's instead from Liitokala which seem to have a decent reputation in terms of cell quality. Hopefully these ones work out a bit better... Wish me luck! :shock:
 
lionman said:
I made the mistake of researching fake cells after ordering them and found out that HG2s are pretty much the most faked cell out there. Woops!

How much did you pay per cell?

If it is to cheap to be true it most likely is not true.

Use a resistor and test at 10A or better 20A discharge. Check for Ri (switch load on and off and Monitor voltage) and heating of the cell.

Builing a cheap fake cell with 3000mAh is at least possible, building a cheap fake cell with a DC Ri of 22mOhm is very unlikely.

So check for Ri and capacity.
 
flat tire said:
The stickers on those cells scream fake looks like you got screwed. Fortunately you paid with a credit card, paypal, or some other legit service right? In that case you would be able to do a chargeback / claim and get a full refund.

They most likely are. Imho IMR cells (LMO cathode) with 3000mAh capacity do not exist.

LG HG2 cells are quite expensive, so why should anyone sell them cheaply?
 
lionman said:
So I got a bit over a 50% refund for these cells. Bit of a bummer I have ended up paying for a bunch of pretty crappy cells. You live and learn I guess...

At least you suppored another fake seller. If the buisness selling fake cells works fine why should they ever stop it?

If your cells have only 1800mAh I assume we are talking about cells worth less than 0,50USD per piece.
 
Cephalotus said:
lionman said:
I made the mistake of researching fake cells after ordering them and found out that HG2s are pretty much the most faked cell out there. Woops!

How much did you pay per cell?

I paid about $2.70USD and got $1.40 back.

It's obvious now. As I said, live and learn.

They would be worth less than I paid even after the refund but I'll live with that for being a fool. :oops: That being said, 18650s are difficult to get under $8 a cell here. They are sold in low volumes for vapers.

The 35E cells where $4 each. Hopefully that's a realistic price...

I'm not going to do any more testing on these cells. My charger does IR testing but I don't think it's very accurate. I don't have the gear to test it accurately. I'll use the crap cells in a big parallel groups so each cell doesn't see a lot of current. I'll still get some use out of them.
 
Back
Top