18650 spot welding -how to- ULTIMATE REPOSITORY

[mistercrash]

And THAT is THAT! We all good now? How about a discussion on what is the best way to connect the cells together in series and parallel? Most batteries I see that people make themselves have strips of Nickel welded to make the parallel connections and then more strips of Nickel to make the series connections. I remember dmun brought up a different way to make those connections that takes a lot less Nickel strips and despite my ignorance on the matter, it just makes sense to me. Make the series connections with small lengths of Nickel strips and then the parallel connections on the strips with a small gauge wire soldered on. Solder heavy gauge wire on the Nickel strips for the positive and negative poles of the battery. Below I included a couple drawings of a 4S14P battery, first the all Nickel strip battery I often see and second, the way dmun suggested many times. Sorry for the bad choice of colors representing some materials.

 

[ridethelightning]

thanks mr N!
this is an awesome bit of info!

 

[litespeed]

There's no need for tone like that. I just made a few points and asked a few questions, but guess you got a bit to upset to answer them, and just started bashing.
Please don't lie, my welder wont cost 350$. And while it might be "just battery short circuit device" (LOL), what is yours, capacitor short circuit device? What do you think, which one has nicer pulse shape and more controllable pulse?

So, my point was that this is ultimate 18650 spot welding repository, and your welder is total overkill for 18650 cells. That's because you can't weld material that is too thick for 18650 cell, because you will destroy them. So, if you want to weld 18650 cells (not some thick sheet metal), your welder is too expensive, and demands labor and skills. Especially for someone who just wants to make a pack for himself.
I also didn't mention my spot welder, let alone compare it to yours, because they are whole different leagues.

So, ease up on your tone, and contribute to the topic by answering questions you have been asked. Bashing someone with lies and ad hominem is not the right way to contribute on forum. Thanks.

You set the tone,

I think it's an overkill and waste of money if you are only going to weld 18650 cells. It's 600$ (with electrodes?) and you still have to invest your time by assembling it? And it's way to powerful.

I just followed it!

To your first question, Yes with electrodes and for the record the second question was retoracal since I already mentioned it was a "kit!" Just incase the language barrier is up "yes you must invest your time by assembling it."

Unlike yours this is not a "capacitor short circuit device" mine is actually a voltage adjustable charge circuit that precisely charges 3 computer grade 1 Farad capacitors that then discharge into 2 individually precisely controlled pulses and then near instantly recharges the circuit ready to go again. This one is fan cooled, plugs into any 110 outlets and completely portable for weighing in at 40 lbs so you can easily take it to your buddies house, no disassembly required.

Yes mine has a better pulse shape and more controllable pulse. Feel free to prove me wrong.

Also I did not lie....car batteries are 75 bux each which you recommend 2 plus the needed plug in power supply and your unit with shipping is $350 and this does not include a way to actually charge the power source of yours....just saying.

To sum up I followed the tone you set and I did not lie. All I was making was a point to add a capacitor discharge welder to the Ultimate Repository list but you were bothered by it.

I'm done with this discussion.....

And THAT is THAT! We all good now? How about a discussion on what is the best way to connect the cells together in series and parallel? Most batteries I see that people make themselves have strips of Nickel welded to make the parallel connections and then more strips of Nickel to make the series connections. I remember dmun brought up a different way to make those connections that takes a lot less Nickel strips and despite my ignorance on the matter, it just makes sense to me. Make the series connections with small lengths of Nickel strips and then the parallel connections on the strips with a small gauge wire soldered on. Solder heavy gauge wire on the Nickel strips for the positive and negative poles of the battery. Below I included a couple drawings of a 4S14P battery, first the all Nickel strip battery I often see and second, the way dmun suggested many times. Sorry for the bad choice of colors representing some materials.g]

I believe the way maddin88 did it is the best way and I followed in his foot prints is to have all the cells in parallel in a straight line and then the series next to it in a + up - up scenario like this picture so you don't have to have so many small pieces. Also trying to keep any one of the small pieces from falling across the batteries creating a short with so many to install would take a lot of concentration. To each their own though,

 

[riba2233]

How about a discussion on what is the best way to connect the cells together in series and parallel? Most batteries I see that people make themselves have strips of Nickel welded to make the parallel connections and then more strips of Nickel to make the series connections. I remember dmun brought up a different way to make those connections that takes a lot less Nickel strips and despite my ignorance on the matter, it just makes sense to me. Make the series connections with small lengths of Nickel strips and then the parallel connections on the strips with a small gauge wire soldered on. Solder heavy gauge wire on the Nickel strips for the positive and negative poles of the battery. Below I included a couple drawings of a 4S14P battery, first the all Nickel strip battery I often see and second, the way dmun suggested many times. Sorry for the bad choice of colors representing some materials.

You are right, second picture is better option, since on the first one there's a lot of wasted nickel. However, that single soldered copper wire in the middle of the pack is not so practical because it protrudes over the plastic brackets, so it would be better to replace is with welded thin and narrow nickel strip.

Unlike yours this is not a "capacitor short circuit device" mine is actually a voltage adjustable charge circuit that precisely charges 3 computer grade 1 Farad capacitors that then discharge into 2 individually precisely controlled pulses and then near instantly recharges the circuit ready to go again. This one is fan cooled, plugs into any 110 outlets and completely portable for weighing in at 40 lbs so you can easily take it to your buddies house, no disassembly required.

Look, I haven't said that you welder is not great, but it's not fair to bash on mine like that. Mine also discharges precisely controlled pulses, more controlled because battery doesn't drop voltage as capacitor does. And what do you think happens after "precisely " charging your capacitor? It gets short circuited the same way as mine does with the battery. Absolutely same principle, but difference is that battery holds voltage better. Also, I can weld with one 45 Ah batter, which weighs 11 kg, making whole system 5 kg lighter than your unit. And it's more portable since it doesn't require power outlet (you can use small battery for logic).

Yes mine has a better pulse shape and more controllable pulse. Feel free to prove me wrong.

This is picture from sunstone page, pulse shape for capacitive discharge dual pulse welder:

http://sunstoneengineering.com/wp-content/uploads/2013/12/150dp-graph3.jpg

Doesn't look splendid, but ok. I will take some shots on mine so we can compare.

Also I did not lie....car batteries are 75 bux each which you recommend 2 plus the needed plug in power supply and your unit with shipping is $350 and this does not include a way to actually charge the power source of yours....just saying.

Well, I can weld with one 45 Ah battery that costs 45$ locally (you don't have to buy the most expensive one), and even two of them are 90$ which is not much. Power supply is like few bucks, and car batteries are easy to charge, you can use really simple circuit to do that, or find used battery charger, whatever. But, in reality, most people already have battery and power supply, so it's really convenient.

And be honest, my tone was nowhere near to yours, you upped the ante. I just made few objective points, and I would like to keep discussion like that, objective and without negative tone. Thanks.

I'm sorry but this isn't strictly off topic, and I can't let someone to bash my design with inaccurate claims.

 

[John in CR]

2 pages of the "ULTIMATE" and still zero "how to" other than some different cell configs. If I'm going to DIY spot weld, then I want to do a better job than a typical hand assembler in China. I want to know how to make good connections that don't go deep enough into the battery to reduce its life.

 

[riba2233]

2 pages of the "ULTIMATE" and still zero "how to" other than some different cell configs. If I'm going to DIY spot weld, then I want to do a better job than a typical hand assembler in China. I want to know how to make good connections that don't go deep enough into the battery to reduce its life.

There are some advices in other thread (high power cells, samsung25r/lg he2/sony vtc5), maybe should be moved.

My advice - don't used nickel thicker than 0.15, that way you are on the safe side. Then you can weld something on top of it if needed.

 

[Nobuo]

2 pages of the "ULTIMATE" and still zero "how to" other than some different cell configs. If I'm going to DIY spot weld, then I want to do a better job than a typical hand assembler in China. I want to know how to make good connections that don't go deep enough into the battery to reduce its life.

You are right, I started this because a united resource was needed. I take all into account, I will compile and make sums about the -how too's- soon with info and videos. It is going slow but solid.

There are some advices in other thread (high power cells, samsung25r/lg he2/sony vtc5), maybe should be moved.

My advice - don't used nickel thicker than 0.15, that way you are on the safe side. Then you can weld something on top of it if needed.

nice, if you spot some essential info you can post the ES tread link here, and i will link it with a title on the first post.


Thanks, you are both two of the battery experts needed here

 

[riba2233]

Here, from this post:

http://endless-sphere.com/forums/viewtopic.php?f=14&t=57291&start=550#p1034783

 

[Nobuo]

i'm not expert but I think tungsten is used in TIG welders because is the metal conductor that has a reliable electric conductivity with highest melting point.

I'm sure it would work good in spot welders as electrodes but there is two reasons I think is not the most used option.

- copper is cheaper and work good welding the nickel thickness is needed for 18650 cells

- Tungsten is only slightly better conductor than nickel, so it would need higher amps and thicker coppers for the same welding with copper poles

Now, Let's hear what the experts say

 

[ridethelightning]

just remember that some of the the tig electrodes are~2% radioactrive thorium. i believe this makes them last longer.

be careful, better not start grinding them...

 

[ohzee]

What are the suggestions for holding the cells together until I can shrink wrap the entire pack ? I was thinking I would just use hot glue with a high melting point.
I also have some pretty good double sided tape that may work just as well , but be easier to deconstruct in the event it's needed. About have everythign I need to
start building my 20s15p battery just trying to decide the best way to do it. Thanks

 

[Nobuo]

In my opinion if you use completely new brand cells, high quality ones (LG, Panasonic, Samsung) with the same batch date, you use a BMS and you take discharge continuous rates at 1C máx, 0.33C charge rates, all is spot welded with pure nickel and shrink insulated, the chances of a failing cell are extremely low. In any case also exchanging a failing cell with a new one wouldn't be the best for the pack.

if you make a complete paralleling/serialling between cells a bad one won't make a difference in such big packs (depending on what kind of fail).

(in any case glue let things the most stone-strong and right secured)

 

[ohzee]

Thanks Nobuo and I tend to agree with your points.

Rarely will I use all that capacity so even if I happened to loose some it should not be an issue.

It is just so much more fun to be cruising along with my pack at 80+ volts vs starting to dip down into the high 70's.

I have 400 of the tesla cells 1st batch so I plan on stuffing as many as I can into the frame. Because of how the frame is built Ill either have to have plugs between the groups or spot weld them in place. Now I just hate to have my bike down while I do this.

Another good point is when I get that done I will have the current 20s9p 25r battery to use on something else.

 

[litespeed]

What are the suggestions for holding the cells together until I can shrink wrap the entire pack ? I was thinking I would just use hot glue with a high melting point.
I also have some pretty good double sided tape that may work just as well , but be easier to deconstruct in the event it's needed. About have everythign I need to
start building my 20s15p battery just trying to decide the best way to do it. Thanks

Just regular hobby hot glue is best since it's easier to remove a cell if needed with just a little heat. If your pack is getting hot enough to melt glue you have bigger problems to think about.....

Double sided tape would make the pack a lot fatter than it needs to be and that stuff holds way to good.

Tom

 

[litespeed]

I am curious what you experts think about using Tungsten electrodes such as used by TIG welders? Of course not cut to such a needle tip but maybe just a 90deg. tip? I figure 2.4mm diameter for example or the thickest I can find mounted to the thicker wires and not protruding more than necessary from the brass block through which they are fed. I figure tungsten has the highest melting temperature, and should keep its shape for a long time, while not being so bulky as some of the copper electrodes I have seen. The idea is to rather weld many small spots than only 2 or 4 big ones!

I am planing to build a microwave oven transformer based spot welder, But I am not going to try to squeeze one big gauge wire through its secondary position but many parallel 1.5 or 2.5 mm°2 for maybe 5 or 6 windings.
Any comments welcome!

Sunstone engineering recommended the Glidcop.....

"Thank you for contacting Sunstone Engineering. The electrodes that we recommend using for welding nickel battery tabs is Glidcop, it is a copper/chromium mix. We have special electrodes that we sell for battery welding, part number EL125BATT. These are 1/8” shaft with a 1/16” welding surface on either end. You can find them on our website here:
http://sunstoneengineering.com/products/accessories/electrodes/ "

They work perfect on my welder!

Tom

 

[spinningmagnets]

I would like to add several things to take into consideration. The entire bottom and sides of an 18650 cell are the negative electrode. Only the "button" tip of the positive end is the positive elctrode. there is a small area of insulation between them on the positive end. The sides and even a small portion of the negative electrode that is at the edge of the positive end, is typically covered in a plastic sleeve, upon which the name and type is usually labeled.

Due to vibration over time, a short can develop across the positive central "button" electrode, and the circumferential negative "shell" that is very close to it. For this reason, many pack builders add some type of additional insulation. In the pic below, the 4-cell Paralleled string has a tan-colored paper insulator. Individual silicone "washers" are also available, at very affordable prices.

In the pic below, the cell on the right has an additional "washer" style of insulator on the positive end.

______________________________________________

The next item is adding a space between the places where the "two-probe" spot welder will create the welds. Due the variability of the thickness of the nickel strips, and also the resistance of one strip to the next, some of the spot-welding current will travel through the strip itself, rather than only through the places where the strip touches the exposed end of the cells. This is why production spot-welded packs have a short and thin "split" between one side and the other. This air-gap forces all of the spot-welding current to travel from one side, through the end of the cell, and then up through the other side. Doing it this way can create a more-consistent weld.

This slit can be added to plain nickel strip material with a dremel tool, using a thin abrasive disk.

______________________________________________
Individual cell-fusing:

If one cell in the middle of a pack begins to short out and become very hot, very rapidly...having the connection between the cell and the rest of the pack formed as individual fuses can (in some cases) stop a runaway short from affecting the rest of the cells. The pack might still be ruined, but having only one cell overheat, rather than all of them....is a much more desirable situation.

The pic below is a design of a laser-cut nickel strip that connects one end of a collection of 12 cells, to form a 2S / 6P sub-group. The six tabs on the left side have the central split (described above), and those would be spot-welded onto the negative ends. The six tabs on the right side not only have a central split, they also have the connection from the center out to the rest of the cells formed with four thin "S" shaped strands. These are the fuses. All four strands would get very hot during a shorted cell, and...as soon as one of them melted completely through, the current flowing through fewer fuses would cause an immediate cascading failure of the remaining three strands. The fuse "strands" are to be located on the postive "button"

 

[mlt34]

I'd like to submit a how-to I wrote up on building 18650 ebike packs - I took a lot of pictures along the way of one of my builds to try and make a very clear walk through: http://www.ebikeschool.com/how-to-build-a-diy-electric-bicycle-lithium-battery-from-18650-cells/

 

[mlt34]

This slit can be added to plain nickel strip material with a dremel tool, using a thin abrasive disk.

The DIY slit is a cool idea. I usually buy just normal pure nickel strip, but there is the option to buy some pre-split strips like these: http://www.aliexpress.com/item/0-15-7mm-forming-nickel-tape-high-quality-nickel-strip-18650-4P-pure-nickel-tape/1565779974.html?spm=2114.32010308.4.2.4YFfRP

Kind of pricy, but it certainly accomplishes the benefits of what you're talking about.

 

[riba2233]

I'd like to submit a how-to I wrote up on building 18650 ebike packs - I took a lot of pictures along the way of one of my builds to try and make a very clear walk through: http://www.ebikeschool.com/how-to-build-a-diy-electric-bicycle-lithium-battery-from-18650-cells/

You should have used extra washers on cell positives, and extra insulation between parallel groups connected in series.

 

[mlt34]

You should have used extra washers on cell positives, and extra insulation between parallel groups connected in series.

Thanks for the feedback!

The Panasonic 18650PF cells actually already come with an extra layer of white insulation at the top end. I've used the insulating washers on some cells without an extra layer of built in insulation but these cells don't really need it. In fact, I've never actually seen a failure of the heat shrink there in all the cells I've built with, even the ones without an extra layer of insulation built in. I know some people use them, but I've never seen them to be necessary. Are there any documented cases of cell shrink wrap wearing away and failing there? I've never seen it in any of my packs, but I mitigate the vibration by bonding all cells together anyways.

I've never heard of anyone insulating the sides of the cells. Do you mean that this would protect against vibration damage to the green shrink wrap on the sides? In all of my experience I've never seen any damage to the sides of cells, and plus they are all glued as one block (and in practice the glue actually provides a bit of gap between cells) so I don't think insulating between cell sides is actually necessary. I guess you could add foam in between each set like I did around the whole thing, but it would just be wasted space in my opinion. No manufacturer I've ever seen does that. By mitigating individual cell vibration, you mitigate cell vibration damage.

 

[ohzee]

I'd like to submit a how-to I wrote up on building 18650 ebike packs - I took a lot of pictures along the way of one of my builds to try and make a very clear walk through: http://www.ebikeschool.com/how-to-build-a-diy-electric-bicycle-lithium-battery-from-18650-cells/

wow nice mlt good work.

 

[meelis11]

I'd like to submit a how-to I wrote up on building 18650 ebike packs - I took a lot of pictures along the way of one of my builds to try and make a very clear walk through: http://www.ebikeschool.com/how-to-build-a-diy-electric-bicycle-lithium-battery-from-18650-cells/

Couple suggestions/ideas:
* I added fuse and connector before BMS for pack I built. Then connected BMS to this connector. That way I can bypass BMS if it somehow becomes bad or something. Or if mosfets fail shorted, fuse burns. Call it extra line of defence (fuse before BMS) and redundancy (skipping BMS).
* these strips on positive should be little-bit shorter - to avoid sharp nickel-strip corners near negative case. If enough compression from top these may go thru shrink wrap and you have shorted and sparking cell.

 

[mlt34]

I added fuse and connector before BMS for pack I built

interesting! I normally add the fuse when I connect it to the bike, but putting one in before the BMS, that I never thought of! I guess you could run the wiring outside the pack for the fuse, so you can still access it but keep the whole pack shrink wrapped.

these strips on positive should be little-bit shorter - to avoid sharp nickel-strip corners near negative case

That's a good point. I like the idea of shorter tabs for the top. The scissors cutting the nickel sometimes gives the corner a bit of a bend. I always try to put that bend up if I can't flatten it completely, but shorter tabs would solve that too.

 

[Electric Junk]

Just a tip i was having problems stacking nickel strip for more amps so i sanded the top layer and it stuck better with less heat.

 

[Nobuo]

I would like to add several things to take into consideration. The entire bottom and sides of an 18650 cell are the negative electrode. Only the "button" tip of the positive end is the positive elctrode. there is a small area of insulation between them on the positive end. The sides and even a small portion of the negative electrode that is at the edge of the positive end, is typically covered in a plastic sleeve, upon which the name and type is usually labeled.

Due to vibration over time, a short can develop across the positive central "button" electrode, and the circumferential negative "shell" that is very close to it. For this reason, many pack builders add some type of additional insulation. In the pic below, the 4-cell Paralleled string has a tan-colored paper insulator. Individual silicone "washers" are also available, at very affordable prices.

In the pic below, the cell on the right has an additional "washer" style of insulator on the positive end.

______________________________________________

The next item is adding a space between the places where the "two-probe" spot welder will create the welds. Due the variability of the thickness of the nickel strips, and also the resistance of one strip to the next, some of the spot-welding current will travel through the strip itself, rather than only through the places where the strip touches the exposed end of the cells. This is why production spot-welded packs have a short and thin "split" between one side and the other. This air-gap forces all of the spot-welding current to travel from one side, through the end of the cell, and then up through the other side. Doing it this way can create a more-consistent weld.

This slit can be added to plain nickel strip material with a dremel tool, using a thin abrasive disk.

______________________________________________
Individual cell-fusing:

If one cell in the middle of a pack begins to short out and become very hot, very rapidly...having the connection between the cell and the rest of the pack formed as individual fuses can (in some cases) stop a runaway short from affecting the rest of the cells. The pack might still be ruined, but having only one cell overheat, rather than all of them....is a much more desirable situation.

The pic below is a design of a laser-cut nickel strip that connects one end of a collection of 12 cells, to form a 2S / 6P sub-group. The six tabs on the left side have the central split (described above), and those would be spot-welded onto the negative ends. The six tabs on the right side not only have a central split, they also have the connection from the center out to the rest of the cells formed with four thin "S" shaped strands. These are the fuses. All four strands would get very hot during a shorted cell, and...as soon as one of them melted completely through, the current flowing through fewer fuses would cause an immediate cascading failure of the remaining three strands. The fuse "strands" are to be located on the postive "button"

Thank you very much with this extended considerations. I really appreciate how well explained it is. I actually had an idea of all you mentioned except about the small gap some nickel tabs have in the middle of the welding spots. That's really interesting.

About the insulation "washers" I would add, that some cells like Panasonic or LG ones had a double layer of insulation around the anode (it is really like a plastic spacer inside the insulation wrap) so if the pack is clever spot welded and then is consciously covered with anti-vibration pad, I think it would be made enough secured for years without the need to add spacers, although it is always a good improve.

The nickel gap I'm sure it make a difference on >=0.15mm thick strips, where sometimes nickel disintegrate making a hole, or stick harder on the pole than on the cell. For thinner strips as 0.1mm I have a stone-strong clean weldings with 20A - 25A welding current and standard pure nickel strips.

I really like the individual cell-fusing method for a very appreciable increase on security for a hard fail/short cell. The well-known discussed thing is that it makes resistance points, and it only would work for a specific max continuous amps cell range. (I don't know if that tabs are made in different thickness/fuse points sizes). Also are not a very reliable tab depending on your configuration / shape cells method, like offset "pyramidal" hot glued cells. I think that a pack with factory new, high quality, same batch, spot welded cells and a conservative max discharging rates with a good BMS, had a very low fatal fail rate.

Despite my small thoughts, this info is gold and essential for what a Ultimate repository must be!

I will add the information in the head post if you are agree.

I am, and sure the whole ES battery scene, truly grateful for your contribution spinningmagnets