cell capacity - how precise is precise enough?

Hi everyone,

This is my first post after lurking for quite some time. I have learned so much from so many of you, and gotten really interested in batteries and solar. Its not much, but my first 18650 battery pack replaced an 8 Ah SLA that powered the fish-finder on my fly-fishing pontoon. Weighs 1/4 as much and lasted well over a week in the bush without any need for charging.

Ok. So I have a question about recording cell capacities on cells like we can see in pretty much every 18650 youtube video. It seems like people always record the full 'exact' capacity (e.g. 1981 mAh; 2022 mAh). I'm wondering if that level of precision is even needed. Would it not be OK to simply write capacity to the nearest 100 mAh (or maybe 50 mAh). That would make sizing packs much simpler.

Here's why I'm thinking this way. If you look at a battery spec sheet, the capacity is almost always tested in factory at a CONSTANTdischarge rate of 500 mAh or less in a tightly controlled environment. With any cell, discharging at different rates will produce different capacity #'s. Different temperatures will affect capacity #'s (and other factors influence it as well I'm sure). Then, in our real life use, the discharge rates are rarely constant/stable, and using full capacity isn't normal.

So, long to short, given all the variability, is the common practice of 4 digitprecision needed?

Thanks,

Eric

Personal preference is the key. When I build my system I'll probably have a 300mah +/-. For example low/high; 2000/2300. That should be more than adequate for "precision".

Obviously everyone is capable of choosing their own range. Which will depend on their available cells and cash availability.

When testing cells writing down the ~capacity~ should only be used as a source reference to quickly put together similar cells per pack.

But if your look for simplicity, I suppose you could use a shorthand reference: instead of 2854 or 2495; you "could use" 2.85 or 2.49.

"Personal preference"

I don't see any reason why not to round to the nearest 50mAh. Even 100mAh would probably be ok. The reason being, when building a pack, it's usually built with a range anyways. Like, adding all the 1800mAh - 1900mAh to a pack. So at that point, it's kind of a mute point to be "accurate" to the 10's or 1's of a mAh.

I can see your point, here. 100mAh would probably suffice. However, I would probably round down if the number is at least 60 or 70 (1860mAh rated at 1800mAh). It'd be better to have a margin/buffer. Might even knock off all the way to 80mAh, even.

Thanks for the insight. I can see the logic in rounding down - having a buffer is smart. Given my starter level of knowledge/experience, I'd rather 'lose' a bit of capacity than possibly fry things.

Capacity is one but select the same internal resistance is the second factor to get the best battery!

Yep. I really don't care much about capacity, since I don't usually fully charge batteries, but internal resistance at LOAD is the most important factors in my eyes, since it's the one that can make or break a pack.

I don't think it matters either. You're mostly testing for good vs bad cell anyway. The important part is consistency. If you test 500 cells writing in the full 4 numbers like 2518 then you switch to rounded numbers for the next 500 and have things like 2550, 2600, etc, you'll probably find it annoying that you now have 500 tested 2 different ways I know I would.

I run per 100mAh. Because I test with the opus chargers and they can easily diff 10% per test and device used.

Then i just random it up when building and that should give me somewhat even packs.


As said the important part is to test all cells equally.

ESR ESR ESR! Then capacity. Will it work if you'd don't match the ESR, yes..will it work 100% better if you do, yes. The internal resistance is literally responsible for letting the current flow in and out...if you want a balanced pack, match them by ESR and the lowest capacity cell. So many people don't bother with ESR and that's nuts! You'll notice how much faster your battery charges and that it will stay balanced all by itself if you use like ESR cells.


Check out my new thread to trade your cells for matching cells!

But ir depends. If you go with 50 mA or 5A. I dont care much about ir in powerwall since i have so damn low current. If they pass my 1A test they will work with 50mA continious.

If you have decent equipment and time then test ir. So far my laptop powerwall have worked without switching 1 single cell. Its now 1.5 year.

I think that random will spread it evenly. Time will tell

But for those sceptic that want 100% ir is a must. But then dont use cheap ass tester

Inverted18650 said:

ESR ESR ESR! Then capacity. Will it work if you'd don't match the ESR, yes..will it work 100% better if you do, yes. The internal resistance is literally responsible for letting the current flow in and out...if you want a balanced pack, match them by ESR and the lowest capacity cell. So many people don't bother with ESR and that's nuts! You'll notice how much faster your battery charges and that it will stay balanced all by itself if you use like ESR cells.


Check out my new thread to trade your cells for matching cells!

Click to expand...

Sorry...noob here. ESR?

Also, is there a reasonably efficient and effective way to test resistance? My hobby charger will do it, but with several hundred cells, it will take many hours. I kinda like Daromer's thinking.

Having said that....

Once cell resistance is known, what is a reasonable range between cellswithin one pack? A few mOhms, 5-10 mOhms, 20+ mOhms? How about between packs?

So much to learn....but I'm enjoying it.

Thanks.

This is mine. HP 4274A. Costs about as much as a small house. Lol..Sorry, ESR is an engineering term for equivalent series resistance and it's probably not the appropriate term for the 18650 "internal resistance" is a better term.

---

The datasheet tells you "the initial internal resistance." That's your baseline. As cells age that figure rises.


You do not need specialized equipment to make the measurements. The cell needs to be under load, again it's in the datasheet and most manufacturers are different. If your tool has an internal checker now, it will work fine, and it WILL be a little different every time and that's okay too

I am working on including an easy calc into my new Arduino based 18650 tester that I'm going to roll some videos out for soon.

---

Are you building 12/24/48V packs and what range cells (mah) are you getting? As far as how accurate they need to be to go into a pack, your original question, I think the personal preference reply was spot on. Just know that whatever the minimum capacity cell in that bank is, that is the entire banks new minimum. I will commonly use 2200mah cells in a 2000mah build, but INEVER put a 1950mah cell into it.