No luck soldering aluminum cell tabs. What am I doing wrong?

[jag]

I finally putting together a pack from my cellman A123 pouch cells purchased earlier this spring:
View attachment DSCF7542.JPG
Before risking the real cells I decided to practice soldering on some other aluminum pieces. Tried both thin alu sheets and household aluminum foil.

I sanded and applied "solder it" aluminum and pot metal paste:

Then soldered with a pen torch:


While the solder wetted the copper of the wire fine, it didn't stick to the aluminum at all. Here I used a hot flame and got a bit of burnt flux.
View attachment DSCF7547.JPG

Here I used a less hot flame and didn't burn the flux, but still no wetting of the aluminum:
What am I doing wrong??

I tried the same procedure with a soldering iron instead of the torch with exactly the same result: "Solder it" paste would stick fine to copper wire, but not to any of the aluminum pieces I tried on.

 

[liveforphysics]

I get good tinning/wetting from this type of solder-it paste:

http://www.goodmart.com/products/solder-it-aluminum-solder-paste-alp-21.htm?source=gsGM

I think the moment you touch flame to the aluminum, you've got an oxide layer on it, and you're toast for soldering.

This is my method. Get some course wet/dry sand paper, like 120grit. Put a little dab of aluminum solder flux on the tab. Take a few strokes with the sand paper to scratch off the oxide layer, while the flux is acting as a shield to keep it from being able to re-oxidize. Now you've got a surface ready to be soldered. I personally would go with a focused tip attachment on a heat gun rather than something with flame, but I think either could work if done correctly.

Aluminum is all about defeating the oxide layer, the moment it's clean and hot metal exposed to air, it oxidizes in a split second. It's about making sure that doesn't happen.

If you've got a bunch of cells to do, I made an iron that uses two flat bars of copper mounted to a pair of high powered soldering irons on a pair of BBQ tongs. It's sized just perfect for soldering the tabs on the cell-man cells. I made it as a tool to freely pass around for folks who want to solder up a pack, then pass on to the next guy.

 

[jag]

I get good tinning/wetting from this type of solder-it paste:

http://www.goodmart.com/products/solder-it-aluminum-solder-paste-alp-21.htm?source=gsGM

I think the moment you touch flame to the aluminum, you've got an oxide layer on it, and you're toast for soldering.

This is my method. Get some course wet/dry sand paper, like 120grit. Put a little dab of aluminum solder flux on the tab. Take a few strokes with the sand paper to scratch off the oxide layer, while the flux is acting as a shield to keep it from being able to re-oxidize. Now you've got a surface ready to be soldered.

I'm using exactly that same paste. So I guess I have to work on my technique. What you say about oxide makes perfect sense. I sand first, then flux and solder. I'll try your suggestion to rub flux in while sanding and see how it goes.

Saw a post where someone suggested soldering the alu tabs submerged in kerosene. Seems justa bit risky. Have to keep soldering iron temp below flame point of kerosene or else...

Thanks for offering the tool. Any pics? I'm going to build a pack that can be reconfigured between 16s and 4s4p, so will do groups of 4 cells, then double 10 gauge wire. I'm most concerned to get good contact between tab and wire and was thinking to try to make a jig where the tab would be pressed around the wire during soldering.

Pack will eventually run an Astro 4t at 100A, but peak amps may be much higher.

 

[myzter]

yeah it is a pain.. I eventually gave up on the soldering and bolted the wire to the cells tabs

 

[webfootguy]

I suggest researching how to join to Aluminum. Here are some links:

http://vimeo.com/2721680
http://www.youtube.com/watch?v=OLvDHwQFvAM
http://www.advrider.com/forums/showthread.php?t=384969
http://durafix.com/
http://www.wattflyer.com/forums/showthread.php?t=33076
http://www.ccis.com/home/hn/index_files/Page4.htm
http://blogs.indium.com/blog/mario-scalzos-tech-support-blog/0/0/soldering-to-aluminum


If you can get the joint hot enough without putting the rest of the cell in danger, then any of the Solder-It products with a good flux in it will probably work. A product like Alutite might be better as it has a lower melting point than raw Aluminum but you must break the oxide layer during the welding process by hand. Doing this under oil might be best if you figure out how to get it hot enough without setting the oil on fire. I advise caution here.

My opinion after researching this a bit that is for Aluminum tabs, it is just best to clamp it under high pressure after cleaning with a decent flux. The makes it easier to replace any cell that might fail later and it is easier to assure high pressure clamping than a good solder joint. Why do they design cells with aluminum tabs to begin with? Would not copper make a better tab material? What am I missing here?

Hope this help.

 

[myzter]

there is a copper layer under the aluminum, If I recall correctly

 

[jag]

So I checked all the links, saw all the videos and tried again today. Still no luck. I mostly worked with alu pieces cut from a roll of aluminium flashing. I cut tabs roughly the sice of the battery tabs to practice with. I also carefully tried a bit on one of the real a123 cells; no luck with that either. B.t.w. I didn't see any copper under the alu when sanding the a123 tab...

What happens is the following:
I sand (tried both dry and with solder it paste on the sand paper)
Apply solder it.
Heat (tried both a small torch and a solder iron)
First flux flows to a clear liquid.
Solder remains suspended in flux.

With a bit higher temp solder starts melting. However it does not wet the aluminum. It just balls up despite that it is in a flux bath.
At this point I've tried to scratch the alu simultaneously as heating. Didn't help.
If I have a tab and a piece of copper wire, the solder will wet and adhere to the copper wire, but not to the aluminum.

Any other ideas of what I can try?

 

[cell_man]

Hi Jag,

I never had much luck with the soldering although I did manage to join a few series cells using just standard solder using a really big iron and rubbing the tab repeatedly with the iron and it did eventually wet. You need lots of heat to get those big tabs up to temperature. Peter Perkins said he also soldered them using standard solder, but just soldered along the edge of the tabs.

When I was soldering I would clamp copper or alumium both sides to act as a heat sink as close to the tab as possible. This seemed to stop the heat getting into the cell.

A coated copper sheet and an aluminium sheet are pretty standard for most/all pouch cell tabs as far as I have seen. Many spot weld a Nickel tab to the Alumium tab to make soldering easier. It still sounds like you're not getting enough heat into the tab. If all else fails I can advise how I went about terminating the cells mechanically. Oatnet (JD) terminated the cells by rolling the tabs and compressing them and hasn't reported any problems with this method.

 

[Jeremy Harris]

One trick that might help is to get a bit of thin (maybe 18g) stainless steel wire and scratch through the molten solder when it's on the surface of the aluminium. I've found that following p[rocess works OK:

1. Clean the aluminium really well with abrasive paper
2. Wipe with a solvent like propyl alcohol to degrease
3. Quickly coat with flux/solder paste, scratching through the flux with the stainless wire to spread it and get it in good contact with the aluminium
4. Apply a lot of heat very quickly with a big fat soldering iron with a lot of heat capacity
5. Rub through the molten solder with the stainless wire to get a good tinned surface on the aluminium, but don't take too long doing this
6. Solder the wire on as normal whenever you like after the aluminium is tinned with solder - even a few days later is OK

If you get the aluminium tab tinned nicely using the aluminium flux/solder paste, and the copper wire nicely tinned with ordinary solder, you will find that you can bring the two together with a fairly quick touch of the iron and get the two solders to fuse nicely. In my experience, once the aluminium is tinned I almost always use ordinary solder for the joint, as it tends to flow better than the aluminium solders and the flux doesn't spit so much. I've never had a problem making joints like this, although it is all about preparation and technique.

Jeremy

 

[jag]

Thanks for additional tips.

I didn't do the solvent wipe, so maybe my tabs had a grease film on.
I'll work more on the "scratching while soldering" also.

I'll try more today.

 

[texaspyro]

The problem with the scratching technique is that you only get solder adhesion where you scratched. It is hard to get decent coverage, particularly if you are scratching around in molten solder.

 

[jag]

3rd day trying to solder aluminum. Still no luck with the tabs.

I also tried to replicate the demonstration in this video that webfootguy pointed out:
http://vimeo.com/2721680
Solder will go into the joint and clog it, but it is not adhering to the aluminum. (So the solder just falls out when applying force.)

I've tried high temp (to the point of the flux burning)
Low temp (flux stays clear, solder just melts.

Quick with a torch
Slow with an 80W soldering iron.

None works.

Not sure what do do next...

 

[flip_normal]

I'd forget about soldering the tabs and buy some clips like these.View attachment u clips.jpg
Fold the tabs together and slip some clips over them. Cover in kapton tape. Job done.

EDIT: Balance wires can be soldered to a thin strip of copper which the tabs are then folded on to. It maybe a good idea to put the kapton tape on top of the folded over tabs before adding the clips thus making a thicker tab for the clip and therefore possibly greater pressure and also avoiding any potential electrolysis issues between the steel and aluminium.

EDIT:EDIT: Maplin, and other places, sell Electrically Conductive Silver Paint which if you applied when you folded the tabs over may give a better connection when it dries. I haven't tried the stuff.

 

[Jeremy Harris]

The problem with the scratching technique is that you only get solder adhesion where you scratched. It is hard to get decent coverage, particularly if you are scratching around in molten solder.

In practice, what seems to happen is that the solder creeps under the oxide layer from the scratches, rather than only fusing just where the scratches are. When I was first shown this method I was pretty sceptical, until I saw for myself just how well it worked. It appears to lift the very thin oxide layer and then flow over the surface.

It's also the method that the various DIY aluminium brazing systems seem to use too, presumably for the same reason.

Jeremy

 

[peterperkins]

I managed to solder the tabs on my 50 cells fairly easily. I used an 80W el cheapo iron with large flat tip.

I posted some details in the termination thread.

I used a power file on the tabs just before soldering them. I only joined the top 3-5mm edge to the next tab. It's actually pointless joining more than couple of mm as the tab thickness where it enters the cells is the final limit as to current capacity.
You don't need to join the entire tab face to the next tab.

Seems to be fine with my 100A drain.

 

[jag]

So I read both Peter Perkins and the "how to terminate" thread all through. Lots of neat projects and good info.

I also found a nice technical description of Aluminum soldering electrochemistry:
http://www.aws.org/wj/2004/02/046/
http://chestofbooks.com/home-improvement/workshop/Handy-Man/How-To-Solder-Aluminium.html
Even got some ZnFl flux inspired by above articles.

Yet still no luck with wetting the aluminum with the solder-it paste. Not sure what else to try.

In desperation I got some paper spring clamps and was considering mechanical termination similar to oatnet's. But while probably fine for 50A, I was worried about long term all weather durability as well as heat and possible fireworks if I try to draw 100-200A from a wire rolled into the cell tab and then spring clamped.

Despite trying for 2 weeks results look identical to the pictures I made for the first post in this thread.
Should I give up, and do the spring clamps? Or is there anything else to try to get the soldering working? It seems silly to not be able to learn to solder if it indeed can be done.

 

[liveforphysics]

Aluminum can be soldered.

Years ago, every rc lipo pack had to be hand soldered with aluminum tabs. I was able to solder these tabs fine. You can too.

 

[cell_man]

I was able to solder the tabs with normal solder so it should be possible with a big and hot enough iron. I used a massive 200W iron set to it's max temperature of 500degC to give you some idea, but you need to make sure that you are taking precautions to prevent the heat getting into the cells themselves. I clamped copper or aluminium both sides of the tab where it enters the body of the cell. If you do not do that you can/will damage the cell. Prepping the cell tabs will help a lot. I never found the tabs wetted properly but it was reasonably good. I would scratch the hot iron on the cell tab and it did eventually take and then spread.

 

[Icewrench]

I`m in the same boat as Jag. I have had success soldering a few 123 cells that lost the tabs. So I figured this would be easy.
I have not hit the right combo yet. If I do I will share.
Luke could you post a pic of the solder tongs you mentioned? Thanks

 

[cell_man]

I`m in the same boat as Jag. I have had success soldering a few 123 cells that lost the tabs. So I figured this would be easy.
I have not hit the right combo yet. If I do I will share.
Luke could you post a pic of the solder tongs you mentioned? Thanks

Soldering A123 26650s is really not so difficult, it is much more tricky to solder the pouch cell Alloy tabs. A123 26650s have a nickel lamination on the positive and niclel plated steel on the negative. It is only the case that is aluminium, so as long as you get them hot enough, they do solder ok. You cannot isolate the heat on a cylindrical cell like you can on a pouch cell, so at least you have more time to work on the pouch cells.

If you really are not having any joy with the soldering I can try to put together a set of terminations for you but I will not be able to do this until I return from an overseas trip. Shipping on small parcels is a bit picey so might be better to combine it with some parts if you need anything. I also prepared some termination instructions and will forward them by PM. If you have a few tools the terminals can be DIYed if you are careful, I use a small lathe to get the holes positioned accrately. The dimensions need to be fairly accurate but not rediculously so.

I'll try to forward the instructions by pm. It's not worth damaging the cells with excessive heat if you can't get them to solder. Most manufacturers now tend to spot weld a nickel tab to the alloy tab to get around this problem.