18650 best soldering technique part2

Redpacket

Active member
Joined
Feb 28, 2018
Messages
1,534
So following up on soldering and effects on cells, from this thread:
https://secondlifestorage.com/showthread.php?tid=9482
I wanted to observe & "over-do" soldering heat to see what might show physical signs of damage, so I destroyed a cell for science.
It was water damaged & self discharged to 0.7V but otherwise physically OK, not corroded or punctured.
The cell was a Samsung INR18650-30Q from a power tool pack. Purple jacket, white top ring.
Ambient temp here at time of test: approx 16degC

The soldering iron I used is this heavy dutyone, the tip is approx 45mm x 18mm solid copper & measured at approx 330 degC with a non-contact temp meter. I think it's 80 or 100W (label is toast sorry)

image_fgmnnz.jpg


For the positive end terminal, I soldered to & held the iron on it forleast 15 secs.
The negative end was soldered to & held the iron on it for at least 20 secs.
(both timed with phone)

At the positive end I found this:

image_ptkkig.jpg

Note the connecting strip. IMHO, this strip would reduce heat conduction from the pos terminal pretty well.
The white seal edge partaround the CID/PTCterminal metal piece seemed to be very heat resistant (even tested with a gas torch, didn't soften, barely reacted)
Theblue insulator ring (sorry bit out of focus) also seemed to be completely undamaged. It was easily melted if touched directly with the iron (gas torch made it shrivel up immediately).

image_ichcdu.jpg

I cut the cell length ways like so:

image_ciqbmu.jpg

image_bqeepw.jpg

Then soldered the neg end per above, then let is cool for ~ a minute then peeled it open & found this:

image_qrqyjd.jpg

Basically no damage, maybe some softening of the blue disc but not significant, still intact, pic looks worse due to me peeling open roughly.

But what about the white layer? To see what happens I soldered more:
So I unrolled some of the cell & found the copper sheet. I soldered to that with light pressure on a still tightly rolled part & it collapsed inwards noticeably (obviously not good):

image_wsabhc.jpg

Unrolling a bit more to see the result:

image_csfwzd.jpg

Pretty badly destroyed separating layer ....
So the separating layer is not good with soldering heat & will be easily damaged.
Even though easily damaged, I didn't notice any damage or changes at the edges, eg from neg case


I unrolled more & checked the pos strip area, no visible damage at all.

image_ynijgc.jpg



So given the above and the connecting strip at the neg end is much shorter so I decided to heat that more by directly soldering to the internal strip to see if heat from external soldering would/could do any harm.
Before internal soldering(note white insulation layer is 100% intact after previous external soldering for 20 secs):

image_zqixfw.jpg

After soldering directly to tip of internal strip (note insulation melted back):

image_exxlrk.jpg

This is the first obvious damage I've observed so far. But is even this a problem?
There are extra layers if the white & kapton tape(high temp) there.

image_ljiuuy.jpg

image_kmcizl.jpg

And I had to solder to the internal strip directly only a few mm from that. So I don't believe it is a problem.

My observations &summary:
- soldering for 20 secs with a large hot iron did no visible internal damage to either end of the cell
- at cell pos end, heat from soldering won't conduct to cell case much due to plastic seal.
- the pos end plasticseal seems tough & unlikely to be damaged.
- cell construction at pos end seems less likely to cause damage from soldering heat due to long connecting strip.
- the top & bottom (blue) insulator discs are susceptible to heat but seemed undamaged bysoldering.
- the internal separator layer is easily damaged by soldering.
- cell neg end is more likely to suffer damage than pos enddue to more direct contact of separator layer with cell casing
- soldering at the center of the neg end (not at edges) would be better (more likely to heat separator layer at edges of casing)

Caveats:
- I only tested on one cell
- other cells might be different
- there might be damage that's not visible, eg chemistry, subtle separator layer changes at neg end rolledge, etc
- air (from cutting cell open)might have influenced results (unlikely)
- my iron was ~330degC,hotter irons would increase damage probability
 
Redpacket said:
[...] Caveats: there might be damage that's not visible, eg chemistry, subtle separator layer changes at neg end rolledge, etc [...]

Yes, that's a crucial point. For example, soldering can thermally damage the separator, e.g. it can fuse its pores, which can (locally) limit rate capability, possibly leading to Lithium plating then dendrites then internal shorts. It can also cause the electolyte to decompose, leading to various contaminant byproducts wreaking havoc.

To detect and analyze these and other microscopic changes requires high-end lab equipment, e.g. SEM (scanning electron microscopy) and XRD (X-ray diffraction), combined with solid knowledge of Li-ion electrochemistry in order to asses the impact of such damage.

That no damage appears visible to the naked eye does not imply that there is nodamage. In fact, generally,almost all damage and degradation is not visible by naked eye in a teardown as in the OP.

Again, recall that all respected Li-ion battery manufacturers explicitly warn not tosolder directly to Li-ion cells.
 
The biggest concern is the inside that you dont see. You need a proper xray to be able to determine anything. Or an area where you can open it without pollution.
I have seen some d
Papers regarding this but only for normal testing and not soldering.
 
gauss163 said:
Redpacket said:
[...] Caveats: there might be damage that's not visible, eg chemistry, subtle separator layer changes at neg end rolledge, etc [...]

Yes, that's a crucial point. For example, soldering can thermally damage the separator, e.g. it can fuse its pores, which can (locally) limit rate capability, possibly leading to Lithium plating then dendrites then internal shorts. It can also cause the electolyte to decompose, leading to various contaminant byproducts wreaking havoc.

To detect and analyze these and other microscopic changes requires high-end lab equipment, e.g. SEM (scanning electron microscopy) and XRD (X-ray diffraction), combined with solid knowledge of Li-ion electrochemistry in order to asses the impact of such damage.

That no damage appears visible to the naked eye does not imply that there is nodamage. In fact, generally,almost all damage and degradation is not visible by naked eye in a teardown as in the OP.

Again, recall that all respected Li-ion battery manufacturers explicitly warn not tosolder directly to Li-ion cells.

I understand totally why manufacturers say to use the lowest risk method - to protect their reputation & liability (especially in today's super litigious world).
While I agree none of us want bad outcomes, I don't agree we should be unreasonably scared about this.
We (you guys more than me!) have a large base of soldered cells and don't seem to be seeing incidents with cells.
This alone is a solid measure of actual issues.
Measurement of actual results is the best proof of anything.

There's a balance between theoretical, risk adverse manufacturers & the real world.
I agree spot welding is preferable but I also suggest soldering seems to be OK to.
Let's not forget that home DIY done spot welding is not nearly as well done as factory systems (we hope factory ones are done right).
Even re-using cells we don't know the history of is a risk.
The process of tearing of the old spot welded tabs is a risk - eg holes, pressure on seals, stretching of neg internal strap.
No process is perfect.
But as a group, we've developed techniques to test & looked for ways to do things that work & don't cause real world issues.
In my tests above I left a large hot iron attached for 20 seconds & there is no visible damage.
Areas likely to be damaged (neg end strap & neg end edge) have extra layers & materials present there.
Given the combination of info we have on soldering, I believe it's OK to use.
 
^^ It iscertainly your prerogative to prefer cheaperhigher risk methods. But please be aware that not everyone is willing to make the same risky choices, so they deserve to have access to all that is known so they can maketheir own decisions on such risks.

The thermographic images from the study that I posted here make it plain as day why soldering is much higher risk. Thathas nothing at all do with "corporate liability" and everything to dowith basic scientific facts.
 
It's great we have multiple methods to connect cells but not everyone is going to use a spot welder, so we need to understand soldering properly too.
My interest is the question "how much of an actual risk is soldering" ie so we understand this method from as many angles as we can.

We respect your knowledge & scientific inputs, thanks for these, they're helpful & interesting.
Yes, soldering clearly gets cells hotter than other methods, not denying that or the science.
The numbers of cells soldered in the second life reader base is scientific data too (maybe not perfect, but data just the same).
The good thing about science is it's based on observations & measurement of actual as well as theory.

So my experiment was observing/measuring "can we show definite damage from soldering?"
So far, I haven't found anything definite & the cell appears undamaged from some deliberately heavy handed soldering.
I get it we might have to get Xrays, microscopes, etc out, but hey we've cleared the "oh look it obviously melted badly" stage.
Apparently only subtle possibilities are left.

Anything can be killed with too much caution....
 
Redpacket said:
[...] Anything can be killed with too much caution....

Alas, people have been killed by too little caution with Li-ion batteries.

And in some cases that was because they were never educated about the innate risks.

Education is not something that should be discouraged - esp. when it comes to matters of safety.
 
Mine are all soldered. 100w, light touch (2-3secs)once to stick somesolder and second time (later) to melt the solder around the wire instead of sticking the solder to the cell - also 2-3secs.

I remember watching a @Daromer youtube with 300w? *huge* soldering iron - a few years ago now :) He was doing (what seemed like)<1sec per cell with quick touch, touch, touch. Because of that I started with a 300w iron but it was too heavy (for me) to manipulate, the tip crumbled within a few hours,and I found that 100w was OK for me.

Its interesting that some cell tops (+ side) solder quicker than others as in 1-2sec vs 2-3secs. The NCR18650A tops take asolder dabwith 1sec - e.g. quick dab - as they seem tinned?. The grey cell tops take the longer couple of secs. The bottoms are moresimilar in my experience - e.g. 2-3secs. Note: I don't do any cleaning or scraping - just solder to the cells 'as is' with flux in the solder.

I'm not planning to unsolder any of the 11,500 cells to date or change my technique - so eventually I'll be able toadd areport on the long term affects of the 6 cell types I've used so far.

Firstbattery (14s122p) wentonline May 13, 2018.Its now completed752 charge/discharge cycles with46.3% averageDOD (within 3.5v -> 4.0v range). Nodetectable degradation or heat or problemsso far.

If I live long enough, I may be able to report more on the long term affects as the years go by:)
 
gauss163 said:
Redpacket said:
[...] Anything can be killed with too much caution....

Alas, people have been killed by too little caution with Li-ion batteries.

And in some cases that was because they were never educated about the innate risks.

Education is not something that should be discouraged - esp. when it comes to matters of safety.

Could you link any of that to anyone soldering cells?
Sure mobile phones, laptops & vape devices have all had issues (manufacturer original & untouched) but what about soldered cells?

All for education, hence the "solder for science experiment".
 
gauss163 said:
^^ It iscertainly your prerogative to prefer cheaperhigher risk methods. But please be aware that not everyone is willing to make the same risky choices, so they deserve to have access to all that is known so they can maketheir own decisions on such risks.

The thermographic images from the study that I posted here make it plain as day why soldering is much higher risk. Thathas nothing at all do with "corporate liability" and everything to dowith basic scientific facts.

I'm having a hard time understanding why a video posted that shows the ends of cells getting hot but not damaged is more dangerous than the possibility of blowing holes in the metal casing with a spot welder. Basic scientific facts? What would those be? I don't see what problems soldering heat is causing? I played the game of testing a group of cells multiple times for capacity, then soldering heavily on each end twice then multiple capacity tests again. I could find no cell degradation in any and all received way more heat than I normally use when soldering. Not plain as day to me anyway.....
 
DiggsUt said:
[...] I don't see what problems soldering heat is causing? [...]

Excessive heat is known to cause various problems, e.g. it can cause decomposition of the electrolyte, and it can cause changes to the separator, etc. These may eventually lead to dendrites, which can (after many cycles) grow large enough to pierce the separator, causing internal shorts and possible thermal runaway.

Unfortunately there is usually no early warning signs of such damage that is detectable by end users. That may occur only when a dendrite tumor grows large enough to cause a short - which could take days, months, or years.

What's plain as day from the linked study is that soldering transfers much more heat to the cell than does spot welding, hence it increases such risks. That's a primary reason why soldering is never used by (reputable) professional pack builders.
 
Everyone agrees that correctly spot welding is preferred for many reasons, but I have yet to see anything quantitative that shows correctly soldering yields reduced performance of the cell. Pictures everywhere that say "See! The end of the cell gets hot during soldering!". Um - ya. That's what soldering does, but how does that soldering quantitatively affect the performance of the cell? I can't find a difference after heavy soldering but I'm a hack. In turn, I can't find anything published that shows a difference.

@add - I build bicycle battery packs and dendrites are more of a concern in powerwalls that have low charge and discharge rates. Cycle packs that have high bursts of discharge tend to burn off any dendrite formations. I'm not sure where I saw this discussed. Can't quite seem to find the reference right now.
 
^^^ Yes, I too wish there were some quantitative results to chew on. But it's probably not very likely that we'll ever see them since such studies would be complex and quite expensive. Who would have any motivation to fund them? Not the manufacturers since they never solder and have little motivation to consider doing so. Perhaps there will eventually be some university funded studies as recycling receives greater emphasis in the near future. But I wouldn't hold my breath for that.
 
DiggsUt said:
I have yet to see anything quantitative that shows correctly soldering yields reduced performance of the cell. ... how does that soldering quantitatively affect the performance of the cell?

gauss163 said:
I too wish there were some quantitative results to chew on. But it's probably not very likely that we'll ever see them

So at this point, the statement that soldering to lithium ion cells is dangerous and should never be done is basically hearsay and just theory. Until solid evidence proves that soldering to cells destroys them, then it is considered a viable option.

With everything, do it correctly. Use just enough solder to make the connection, and just enough heat for just long enough to flow and bond.
 
Was doing some research and came across this site: https://www.samsungsdi.com/lithium-ion-battery/safe-information.html

In there it is stated:
DO NOT charge individual, cylindrical Lithium-Ion Batteries if you are a consumer or end-user.

DO NOT place charger and Lithium-Ion Batteries or Battery Packs under your pillow, on your bed, or on your couch.

This is an example of why we can't "always" trust what their "safety" guidelines are. So, according to Samsung, none of us should be charging Li-Ion cells. We are all consumer/end-users.

And I suppose we are allowed to charge them in other locations, and it's ok. So, looks like you can charge them in garage, on the loveseat, on the table, on the counter, on the carpet, in the clothes hamper, in the dryer, etc.

This falls back on a statement I made earlier. A hair dryer warning label states: DO NOT USE WHILE IN SHOWER.
 
^^ Those poor analogies were already debunkedi inthe prior thread so I see no need to repeat such here.

Korishan said:
So at this point, the statement that soldering to lithium ion cells is dangerous and should never be done is basically hearsay and just theory. Until solid evidence proves that soldering to cells destroys them, then it is considered a viable option.

Qualitative scientific knowledge is not"hearsay". Since you seem to love analogies,let's apply your "logic" in another subject domain

K-logic said:
So at this point, the statement that injecting Lysol to cure covid is dangerous and should never be done is basically hearsay and just theory. Until solid evidence proves that injecting Lysol harms humans, then it is considered a viable option.
 
gauss163 said:
^^ Those poor analogies were already debunked in the prior thread so I see no need to repeat such here.

Korishan said:
So at this point, the statement that soldering to lithium ion cells is dangerous and should never be done is basically hearsay and just theory. Until solid evidence proves that soldering to cells destroys them, then it is considered a viable option.

Qualitative scientific knowledge is not"hearsay". Since you seem to love analogies,let's apply your "logic" in another subject domain

K-logic said:
So at this point, the statement that injecting Lysol to cure covid is dangerous and should never be done is basically hearsay and just theory. Until solid evidence proves that injecting Lysol harms humans, then it is considered a viable option.

You apparently are misreading my comment, either mistakenly or on purpose.

I was "agreeing" with needing qualitative results. Because it appears that there currently is none. Perhaps you should fully understand the comment for what it is before jumping to conclusions and to your limited narrow view of the world. Not everyone is always trying to counter your arguments. If you re-read my comment on results, I quoted you as stating that you hadn't seen any and I was agree we need it.
 
I soldered 80 cells to a skateboard pack, took at most 3 seconds per cell on + side and probably little more on bottom side.
heat doesn't last long enough to do any damage.

3 years later, board is still rocking hard with no cell damage. get over it.
 
@Korishan I see nothing I misread. Rather, my disagreement regards the (unfounded) method of inference you employed above to attempt to deduce that soldering is a "viable option". As you can see from the analogy, that's not a wise way to make deductions.
 
My analogy was not about safety of using lithium cells and soldering. It was in reference to that manufacturers will put warning labels about anything, even if it seems "stupid" to everyone because they are covering their own butts.

Again, you misread what I was stated and inferred my statement was in reference to your comments and safety. Which they were not.
 
Back
Top