Where is the fault in this battery ?

rio9210

100 mW
Joined
Sep 29, 2015
Messages
49
Look at the picture, and try to find the error in this battery assembling setup:

l3WQl
 

Attachments

  • bad_battery.jpg
    bad_battery.jpg
    52.9 KB · Views: 1,625
Six 10P groups plus two 5P groups, instead of seven 10P groups?
 
Chalo said:
Six 10P groups plus two 5P groups, instead of seven 10P groups?

No, the battery is a 14S 5P.

Anyone smarter ? 8)

(suggestion: this battery should manage 40A+ requests...)
 
Aside from far to little cross section for the series connections and drain wires i cannot see an assembling fault.
If used like this the lifetime won't be high because of the non-equal current flow through every single cell, but probabaly something will melt before this might become a problem.
 
Nothing.
The pattern looks ok. I see there's no gaskets and the tabs are wide. May short over time. It should work anyway. Btw I can see only one side of the battery. The other link is down.
 
I see gaskets on the positive terminals. Don't need them on the negatives.

Well, you said 40 amps, which I think would beat 5p to death.

Some of the parallel groups have two strips interconnecting, and some have only one. So there will be increased resistance on some of the parallel groups, driving the pack out of balance. The worst might be the group on the positive end, that has only one cell with the single strip. That cell might take a beating sending its current through more resistance.
 
dogman dan said:
The worst might be the group on the positive end, that has only one cell with the single strip. That cell might take a beating sending its current through more resistance.

The negative end is even worse as there is only one nickel strip for the current of 3 cells.
At 40A continuous it will start glowing.
Most of the connection should be beefed up by soldering copper wires on top of.
 
dogman dan said:
I see gaskets on the positive terminals. Don't need them on the negatives.
I do not see them. But I must admit i am going to visit the ophtalmologist next week.
 
Well, lots of smart guys here, congratulations. Immagine that that battery was assembled and...sell to a friend. Lot of people think "I have a spot welder, I'm going into business" without having the necessary knowledge to do that kind of stuff. Knowing that A stands for Ampere and V vor Volt isn't enough to put friend's butt to risk :lol:

This is the battery with the positive and negative end soldered, they're clearly soldered on the single nickel strip bended behind the plastic support, for a "neat setup"...

b5JzA83.jpg




madin88 : you are right, current flow isn't equally distributed through the cells, so lifetime is questionable. And you are right again, the negative end it's a candidate for "Santa Klaus glowing star" battery reality show.

Joachim : probably some cinese spot welder was used, and while they (somewhat) work, you don't have a lot of control on what you're doing. Very "fine" observation -wasn't intended to be remarked, but the more the better :wink:

dogman dan : Maybe you mean the "standard" gasket used under the PVC protection layer of the cell. Over-the-top adhesive gaskets (like this https://eu.nkon.nl/accessories/welding-material/isolatie-papier-18650.html have a cost (in money) so probably that friend doesn't deserve them, or the "assembler" isn't aware of their existence... And you're right about the positive end (same for the negative) with only one strip, shrinking those 40A in 0,15-0,3mm nickel strip mean looking for troubles.

By the way, that battery was providing energy to...TWO MOTORS, with peak requests around 80A. I've finished to read that post, and the more I've read the more I've wonder how "innocent" some people are. (I think it's something called "the Dunning-Kruger effect")... :oops:

And least but not last, that post began with the client publicly thanking the assembler...
 
What would be a stellar thread, and post, are designs that are correct!
 
Re the insulators, looks to me like they built them into the cells. little black strip on the + ends under the shrink, so there is more than just shrink covering the - part of the + end, the can itself.

Might be wrong, but looks like something more than shrink at those + ends.
 
dogman dan said:
Might be wrong, but looks like something more than shrink at those + ends.
All I've seen do have plastic rings, but paper non-conductive sticky back cover even more.
 
dogman dan said:
Re the insulators, looks to me like they built them into the cells. little black strip on the + ends under the shrink, so there is more than just shrink covering the - part of the + end, the can itself.

Might be wrong, but looks like something more than shrink at those + ends.

Sony VTC5 have that black plastic ring between cell body and external PVC shrink sleeve

IooAVWK.jpg



Sanyo 18650GA have them, too, although they look less effective

dpsyySR.jpg


The VTCs - very rigid plastic rings - look and feel pretty tough, however putting the adhesive ones over the top wouldn't hurt - safety isn't enough, never.
 
rio9210 said:
Look at the picture, and try to find the error in this battery assembling setup:

l3WQl

How about a marked up and corrected layout as a learning tool?
 
tomjasz said:
rio9210 said:
Look at the picture, and try to find the error in this battery assembling setup:

l3WQl

How about a marked up and corrected layout as a learning tool?

There's not just one and perfect layout for that (and any) battery. In this case, if you've just finished the battery and realize "Damn, my friend asked for 80A, those one-two-three strips of nickel (even if 0,3mm) wouldn't be enough" - later in that post the owner of the battery stated that he couldn't pull out all that Ampere from the battery - in this case the only wise thing to do is to pump up the Ampere way with copper wire. It's not a neat solution, but it works.

Like a general rule, I consider the max Ampere I want from the battery and divide it by 6A (if I'm using 0,15x8mm nickel) or by 8A (for 0,2x8mm nickel strip).

Example: BBS01 250W, 15A controller: 15A:6A=2,5 so minimum three strips for 0,15x8mm nickel (allowing 18A safely, you can even "overclock" it to 350W...) and minimum 2 strips (15A:8A=1,87)for 0,2x8mm nickel. Numbers are rounded to the next higher integer.

The faulty battery was a 14S5P, so for the ideal condition of a rectangular battery, there are 5 strips between each parallel. Using 0,2x8mm, that mass of nickel can carry 40A. So a quick math count says that for 80A, you should have 10 (ten) nickel strips. You have to use 0,3mm nickel to get the job done with just five strips.

Another funny thing: an 80A fuse was implemented in that battery, just for safety :? Bad things can happen way before those 80A will be able to see the daylight...
 
Back
Top