18650 DIY powerwall 48v

Maybe I hold the soldering iron on the cell for too long?
Generally using the iron on a cell isn't going to make it 0V. You would need to have the iron on the cell for quite some time to do that level of damage.
I'm more inclined to think that you used too much solder and there's a slight bridging causing a short. An example of a possible overflow would be the Grey cell in the 2nd from the bottom row. It looks like it is awfully close to over flow. Same with the bottom left Orange cell. There's a few other cells that are suspect as well.
Which cells are reading 0V out of the ones in the image?
 
Generally using the iron on a cell isn't going to make it 0V. You would need to have the iron on the cell for quite some time to do that level of damage.
I'm more inclined to think that you used too much solder and there's a slight bridging causing a short. An example of a possible overflow would be the Grey cell in the 2nd from the bottom row. It looks like it is awfully close to over flow. Same with the bottom left Orange cell. There's a few other cells that are suspect as well.
Which cells are reading 0V out of the ones in the image?
The two cells with red ring:
20230911_121435.jpg


20230911_122015.jpg

20230911_122003.jpg
 
Last edited:
I have to get out the cells with voltage drop.
The plastic cell holders are really stuck to the cells so I doubt it will go off the same way in went on.

Do I have to pry off the plastic tab for getting the cell out? Or is there some tricks to get the cell holders off?
20230911_125702.jpg
 
Last edited:
The pics show good solder blobs to me - congratulations! Here's one of mine - the NCR18650A (green cells) take a bit less solder than others.
1694469215768.png


I've never caused 0v by soldering with 100w iron - and I've done over 12,000cells. I have gotten solder down inside the + side a few times (just being sloppy/in-a-hurry) and had to discard because the pressure switch might be compromised but I've never caused a short.

I'm guessing 0v would be due to bad cells or pretty good self-dischargers.
 
Last edited:
I'm guessing 0v would be due to bad cells or pretty good self-dischargers.
When I use cheap tin I noticed it sometimes creates quite a bit of residual tin; and that it flies everywhere around. Twice I got tin in my eyes (I suppose real tiny spherical blobs). From then I started using protective glasses. And I resolved the problem by buying good make tin, more expensive, yes, but a better quality soldering.

So, if the tin isn't very good quality, I could suppose that tiny boiling blobs of tin could fly under the cap...
 
And this is the reason i made me a ridicule extended tester.
It wil catch all the unwanted cells.
Unwanted:
To high or to low IR
Self dischargers.
SOH (only keep what i want and divide 50 to 60, 60 to 70, 70 to 80, 80 to 90, 90 to 100%soh in different packs)
Heaters
Slow chargers, to fast!, this translates into capacity remaining)
And i am going to add one new:
Cells that will not full charge in a certain amount of time.(too slow chargers)
I found sells of 2200mah with a charge of 800mah that were not full in 6 hours, but tested on all other values oke?!?!?!?
Btw those were not HV cells aka 4.5 or even 4.8v (li ion 18650)

With kind regards Igor
 
When I use cheap tin I noticed it sometimes creates quite a bit of residual tin; and that it flies everywhere around. Twice I got tin in my eyes (I suppose real tiny spherical blobs). From then I started using protective glasses. And I resolved the problem by buying good make tin, more expensive, yes, but a better quality soldering.

So, if the tin isn't very good quality, I could suppose that tiny boiling blobs of tin could fly under the cap...
Boiling point of tin is beyond the capacity of a soldering iron---->2200C

Even with cheap solder aka more lead or resin or lack of resin can create blobs(small) big enough to go under the insulator ring and cause a small short--->0v cells.
That is why most people here recommend cell level fusing, and not with a wire, but glass fuses.
I had those problems in the beginning also, solder was sometimes running away.

Yes resin core can have sometimes some water in it and that will sparkle a little bit, no need for worry.
Test your cells better and you will be oke.
It seems to me you are to eager to put cells into a pack.
Let them sit fully charged for a minimum of two weeks measure the v again...0.2v drop in 2 weeks...bye.
 
All the cells has been sitting for a half to one year. 😊

Is it important to have + and - at opposite ends so current flows thru the packs instead of pulling at the end of a pack?
 
All the cells has been sitting for a half to one year. 😊

Is it important to have + and - at opposite ends so current flows thru the packs instead of pulling at the end of a pack?
it depends on the amount of current which will be determined by the overall load. Let's say you have a 14s100p battery providing 20a @ 48v = 960watts. You get the per-cell-current by dividing 100 cells (the 'p') / 20a = 200ma/cell thru-out the 1,400 cells to get ~1000w of load. At this level of current, the issue is small and can be ignored per an electrical engineer friend of mine.

Ebike cells may be higher but the 'standard' discharge for more common laptop level cells is usually 400ma-500ma. Here's a common example - https://secondlifestorage.com/index.php?threads/lg-lgabb41865-cell-specifications.1789/
1694962979708.png

Most of us design 18650 powerwall batteries that peak at <500ma/cell because laptop cells don't do well over this level of current and so +/- at one end is not a major concern when averaging 200ma or 300ma or even 400ma/cell.

However, a big plus for + and - at opposite ends is it's really hard to accidentally short the pack. I use this method :)
 
Last edited:
However, a big plus for + and - at opposite ends is it's really hard to accidentally short the pack. I use this method
Also helps the current flow more balanced between the +/- thru the cells.

If both leads are on the same end of the pack, then the cells closer to the leads will see higher current flow than those furthest away.
So even if napkin math says that every cell will see 200mA/cell (as per OffGridInTheCity's example), this actually isn't true electrically if the leads are on the same end. You could say, for example, that the ones closest could see 20% higher current flow than those furthest away.

Remember, flow follows least resistance. If every cell is at 50mOhms, the buss wire also has resistance. So the ones closest to the leads will have the lowest overall resistance than those further away
 
However, a big plus for + and - at opposite ends is it's really hard to accidentally short the pack. I use this method :)
Yes that's a big plus. 😊
I suspect I'll end up with more than 200ma/cell.

Also helps the current flow more balanced between the +/- thru the cells.

If both leads are on the same end of the pack, then the cells closer to the leads will see higher current flow than those furthest away.
So even if napkin math says that every cell will see 200mA/cell (as per OffGridInTheCity's example), this actually isn't true electrically if the leads are on the same end. You could say, for example, that the ones closest could see 20% higher current flow than those furthest away.

Remember, flow follows least resistance. If every cell is at 50mOhms, the buss wire also has resistance. So the ones closest to the leads will have the lowest overall resistance than those further away
This is a good point and it make sense. I think I will make the + and - at opposite ends. Thank you.
 
Another aspect of running a busbar with only 1 cell on each side (instead of 2 cells on each side) is for easier fuse wire / solder management. You don't want fuse wire from one cell dropping across a 2nd cell on the way to the busbar.
Is 2 cells on each side of the busbar not recommended?
Screenshot_20230917-211742_Gallery.jpg
 
Two considerations with 2 rows of cells per bus-bar / pic above.
1) The fuse wire shouldn't span 2 cells - else 1 cell could take out 2 cells. Each cell should have it's own, direct, connection to the buss. Having 2 cells makes this difficult as the outer cell wire has to cross-over the inner cell to reach the bus.
2) My busses are 6awg twisted copper and I need only 1 cell per buss to carry the overall current of my design spec.

It's DIY so I won't say its wrong - just explaining why you don't see this often.
 
To go along with what OffGridInTheCity mentions, just like with the terminals being connected at the same end, the cells closest to the buss bar will see more current than those further away. Granted, at such low mA, this may not really be an issue. But something to take note of

I also would not recommend taking a fuse wire across more than 1 cell at a time. It increases the likelihood of one of the cells still causing damage to its mate. The fuse will pop from the buss line first. So the remaining cell(s) will still be connected for a time
 
Ok, thanks a lot. This make sense as well.

I've bought 30m with 12awg wire and my plan was 4x12awg, but 30m isn't enough for 2 x busbars running 4x12awg with only 1 cell on each side of the busbar.

But it's maybe enough (the same?) with 2x12awg?
Like this?
20230918_140747.png
 
Last edited:
In my case (pic above) I have 4 x runs of 6awg which is effectively >1AWG ( >119a transmission / >211a chassis) - https://www.wirebarn.com/combined-wire-gauge-calculator_ep_42.html In practice, I run 150a peaks routinely for several hours and can vouch that 6awg x 4 per pack + 4/0 AWG inter-battery-connects will sustain up to 300a (15,000w) loads with 0 heat.

And since I do + to - opposite ends per pack and overall battery - a 300a load is running thru the entire system (including busbars) as I understand it - but perhaps someone can confirm/deny this concept.

4 x runs of 12awg (pic above) is equivalent to 1 x 6awg - e.g. 37a transmission to 101a chassis (https://www.powerstream.com/Wire_Size.htm) so say ~60a max in practical terms. If this meets you're load / design needs then great.
1695048735126.png
 
Last edited:
In my case (pic above) I have 4 x runs of 6awg which is effectively >1AWG ( >119a transmission / >211a chassis) - https://www.wirebarn.com/combined-wire-gauge-calculator_ep_42.html In practice, I run 150a peaks routinely for several hours and can vouch that 6awg x 4 per pack + 4/0 AWG inter-battery-connects will sustain up to 300a (15,000w) loads with 0 heat.

And since I do + to - opposite ends per pack and overall battery - a 300a load is running thru the entire system (including busbars) as I understand it - but perhaps someone can confirm/deny this concept.

4 x runs of 12awg (pic above) is equivalent to 1 x 6awg - e.g. 37a transmission to 101a chassis (https://www.powerstream.com/Wire_Size.htm) so say ~60a max in practical terms. If this meets you're load / design needs then great.
View attachment 30653
Let's see if I understand you correctly.
Do you have twisted 4x 6awg wires in each busbar? So it's 4+4=8 wires on the positive side and 8 wires on the negative side?
20230918_190350.png
 
I don't know the actual gauge of the individual 7 wires but the combiner calculator shows that 7 x 14awg = 6awg.
Nice combiner calculator, thanks for the link. 😊

If I use your busbar configuration and use 4 x 12awg wire the combiner calculator shows that 4 x 12awg = 6awg.

It will be exactly the same as yours?
20230918_220936.jpg

20230918_221509.png
 
Last edited:
Back
Top