Good video on exactly what makes a A cell vs a B vs a C cell

[Maast]

One of the manufacturers that I've been corresponded with put out this video on the manufacturing of LFP cells.

I thought it was very informative - if not exactly thrilling. It's about 18 mins long.

In short: what makes a A cell is it doesn't self discharge beyond spec, the internal resistance is normal but not so low it shows a internal short, and the capacity is up to spec. Also, there are a crapton of B cells out there being sold for cheap, you get what you pay for.

Youtube

 

[FilterGuy]

Thanks for sharing!! Like you say, not very thrilling, but very informative.

 

[erik.calco]

From the video, summary if why one should avoid B/C cells:

That is likely to stop most of us in our tracks from purchasing Bs.

The most important test is self-discharging rate since it effectively tests for a short, which is the easiest for any of us to test for, although it requires patience.

What would be your specification for 28 day self-discharging rate for an 18650? 0.1? 0.01? Or something even more minute?

 

[erik.calco]

Woke up, and my posts to this thread were in another unrelated thread. Not sure if a mod moved, or something else happened, but reposted the one above that was 100% about the OP, and another one was semi-offtopic.

In this post, I reference the next video YouTube recommended after the OP's video on B/Cs. While soldering 18650s is not directly related to A vs B, most of us looking to buy 18650s from China are building packs.

 

[electric]

That is likely to stop most of us in our tracks from purchasing Bs.

The most important test is self-discharging rate since it effectively tests for a short, which is the easiest for any of us to test for, although it requires patience.

What would be your specification for 28 day self-discharging rate for an 18650? 0.1? 0.01? Or something even more minute?

Except that hardly anyone is selling A cells in retail market, especially not on Ali or even Amazon.
My crystal ball tells me that most sellers will soon claim their cells are all A grade, which of course is impossible to prove because grading is not an industry standard, it is up to each manufacturer to decide. Many factories in China can only make B or C cells because their equipment is old and not well maintained and they cut corners on source materials to make a cheap product. I bet they consider their cells A grade, because it's the best quality they can make.
You don't need to wait 28 days for self-discharge. Cell charged to 3.65V with C/20 cutoff should rest at 3.40V or higher after 7-10 days of open circuit storage at room temperature. Less than 3.40V indicates some self-discharge. Must be open circuit, no BMS attached, as it creates a small load and skews the test results.
In my experience bad cells are almost always swollen, with various degree of swelling indicating severity of internal damage.
Here is how I test my cells:
1. No swelling.
2. Holds at or above 3.40V OCV for a week after a full charge.
3. Gives at least 95% of rated capacity at C/2 or less.

 

[FilterGuy]

Except that hardly anyone is selling A cells in retail market, especially not on Ali or even Amazon.

That is most likely a true statement, but how does one know? You certainly can't tell from the listings. Do you have any way of telling if a cell you received came from a 'B' sorting?

 

[electric]

That is most likely a true statement, but how does one know? You certainly can't tell from the listings. Do you have any way of telling if a cell you received came from a 'B' sorting?

I don't care about grading or sorting as long as cell is working and passes my 3 steps criteria I stated above. I try to research the supplier as much as I can, check technical quality of their product listing, check reviews, make sure price is not too good to be true, etc. Just common sense to minimize the risk. There are plenty of perfectly acceptable "B grade" cells, nothing wrong with them for 99% of use cases.
I also want to point out that reference to "explosion" is a complete nonsense, added for dramatic effect.
Extreme caution is needed when working with NCA cells ( Tesla ), some caution is needed for NMC cells, but LFP cells are completely safe. The worst outcome for LFP cell is a dead cell. They can create a bit of a mess when disintegrated due to severe external abuse, and a fire is a small possibility due to external causes, but not from the cells alone, and certainly not due to poor manufacturing.

 

[erik.calco]

I don't care about grading or sorting as long as cell is working and passes my 3 steps criteria I stated above. I try to research the supplier as much as I can, check technical quality of their product listing, check reviews, make sure price is not too good to be true, etc. Just common sense to minimize the risk. There are plenty of perfectly acceptable "B grade" cells, nothing wrong with them for 99% of use cases.
I also want to point out that reference to "explosion" is a complete nonsense, added for dramatic effect.
Extreme caution is needed when working with NCA cells ( Tesla ), some caution is needed for NMC cells, but LFP cells are completely safe. The worst outcome for LFP cell is a dead cell. They can create a bit of a mess when disintegrated due to severe external abuse, and a fire is a small possibility due to external causes, but not from the cells alone, and certainly not due to poor manufacturing.

I just looked back and saw OP put LFP. I don't remember LFP in the video. So, I presumed the "explosion" part to other lithium cells, such as the more popular 3.6Vs.

A short is definitely the number one cause of the explosions, as evidenced by the early 18650 DIY pioneers... vapers. I'm curious what percent are internally related versus being completely externally caused. Plenty are definitely external. But, then that raises the question, are there cases where external shorts are causing explosions that might not of happened if they didn't also have an internal short amplifying the impact?

That said, you are right. There is no evidence that LFPs are explosive. They would probably just bulge when shorted or over charged.

Shorting any cell or battery has the capacity to unleash a contagion in a system that could result badly, simply because of the power they hold and can discharge very quickly. Not sure how you'd demonstrate that with a failed LFP cell in a pack, but, you always want to think through cascading impacts of failure in packs and banks and throughout your circuits and construct them with safety in mind.

 

[electric]

I would guess that 100% of fire/explosion is caused by failure of the external circuit, such as thermostat on the heating circuit, which leads to runaway heating, overheating the cell to the point of failure. So it's not even a short in case of vapers, but a runaway load. This is just my educated guess. I suppose a poorly designed control circuit can actually short out across the battery and cause the battery to overheat, but this is again an external cause, not internal.
I have never seen or heard of internal short causing any kind of danger or contributing to such danger. Internal shorts are very weak, so they just drain the battery flat. If internal short was strong enough to cause the hot spot, then it would fuse the short and stop, resulting in a weaker or dead battery, but nothing else.
The only exception is physical damage, crush, puncture, etc. But then it's not a cell manufacturing defect.

 

[erik.calco]

Very hard to collect data on the contribution to a runaway from an internal short, particularly since the end result is a complete burn.

Regardless, I don't want cells dying or underperforming in my series. Safety issues aside, consistency in a series is a #1 priority for all packs and banks. I'm spending days prepping a LFP bank now before putting in series. I just went downstairs and found a lagger.

 

[electric]

Regardless, I don't want cells dying or underperforming in my series. Safety issues aside, consistency in a series is a #1 priority for all packs and banks. I'm spending days prepping a LFP bank now before putting in series. I just went downstairs and found a lagger.

Depends on your definition of a lagger or underperforming. No matter how you match the cells, there will always be one that hits the top first and one that hits the bottom first. It is not necessarily the same cell.
As long as you get expected use of the battery between those 2 events, then it can be considered acceptable.

 

[erik.calco]

Depends on your definition of a lagger or underperforming. No matter how you match the cells, there will always be one that hits the top first and one that hits the bottom first. It is not necessarily the same cell.
As long as you get expected use of the battery between those 2 events, then it can be considered acceptable.

The point is I care enough to not want a short in my cells no matter how tiny. I'll pay more if there is a way to be sure they don't have shorts. It's something I place a high value on in the context of so much love and investment going into it.

In parallel, my standards are not as high.

 

[Maast]

You never know what you're going to get buying LFP cells from ebay/amazon/alibaba/etc. Even if they're labled as a 'name' brand they could be fakes.

The only way to tell is buy a sample cell and test it before you drop real money on them. It's not hard or terribly expensive to do either but it may take a lot of reading on your part and some up-front investment on a benchtop cc/cv power supply and how to apply a load.

Its why I spent the money to make my tester rig before I buy 48 280AH cells because if it turn out to be a scam thats big money I can't easily replace. I've been saving for years to upgrade the battery bank.

As a bare minimum you need to check for self-discharge and capacity. Any decent cell should hold itself above 3.4V for more than 5 days.

 

[erik.calco]

@Maast can you share specs on your tester rig? e.g., parts list with links?

 

[Maast]

There are some details in this thread https://www.diysolarforum.com/threads/my-diy-300a-cell-amp-hour-tester-and-cycler.1662/ and I'm working on a diagram and parts list.

With the arrival of my meanwell HRP-600-3.3 power supply in parallel with my HRP-300-3.3 it'll charge at 180A and discharge at 20A, 30A, 40A, 70A, 140A or any combinations of those up to 300A. It'll charge to anywhere I select between 2.9v-3.8v and discharge down to anywhere I select from 2.7V to 3.8V. I have it set up for 3.65v and 2.8v. It'll also cycle the charge/discharge cycles and keep count of the number of cycles.

Other than the power supplies its all just relays and common amazon parts and pieces. The HRP 300 and 600 are from ebay. Believe it or not finding a low voltage alarm was the hardest part.

To keep the power supplies from running at their max amperage and overworking them and over-amping the cells I've reduced the HRPs to about 3.45v to lower the charge rate to 140A and I let my benchstop adjustable power supply do the last bit to bring it up to the 3.65v - so there are 3 power supplies feeding it.

You have to use a current limiting power supply or it'll just shut down (or burn out) and the HRP line is the least expensive line of power supplies I could find that'll do that. I've got about $250 into it which is nothing compared to what the new bank will cost.