[Ted Yapo] shared a method of easily and conveniently soldering to aluminum, which depends on a little prep work to end up only slightly more complex than soldering to copper. A typical way to make a reliable electrical connection to aluminum is to use a screw and a wire, but [Ted] shows that it can also be done with the help of an abrasive and mineral oil.
Aluminum doesn’t solder well, and that’s because of the oxide layer that rapidly forms on the surface. [Ted]’s solution is to scour the aluminum with some mineral oil. The goal is to scrape away the oxide layer on the aluminum’s surface, while the mineral oil’s coating action prevents a new oxide layer from immediately re-forming.
After this prep, [Ted] uses a hot soldering iron and a blob of solder, heating it until it sticks. A fair bit of heat is usually needed, because aluminum is a great heat conductor and tends to be lot thicker than a typical copper ground plane. But once the aluminum is successfully tinned, just about anything can be soldered to it in a familiar way.
[Ted] does caution that mineral oil can ignite around 260 °C (500 °F), so a plan should be in place when using this method, just in case the small amount of oil catches fire.
This looks like a simple technique worth remembering, and it seems easier than soldering by chemically depositing copper onto aluminum.
thanks – that’s going in my bag of tricks, who knows when that may come in handy.
Sounds good but what solder. No one tells what solder. 60/40? SnAg? special alumi weld?
Cool Don, BTW, any thoughts on titanium oxide pealing? Write something?
Wonder how well such a thing can be done to a AL heatsink extrusion/cast, those thing have huge surface area to dump heat, but if you’re 3dprinting a case while pushing that the big cooling fin out the back you really should make sure its grounded (at least usually).
if it has a large flat surface, then just lay in on a hot plate, to keep the whole part hot while you scratch the oxide away.
Nothing to stop you filling a pan on the hotplate with sand to try that if it is a less convenient shape either – I’m just not sure how well you can really achieve such a thing, might be too much effort and energy expended compared to tapping that little screw hole.
Just screw a terminal. The increased area will compensate for the very thin (<10 nm) oxide layer, assuming the thread doesn't scrape it off.
Aye screws are the goto, but it could be useful sometimes to solder the earthing lead to the heatsink when its too thin to screw too, or there isn’t clearance etc.
https://hackaday.com/2020/06/11/solder-to-aluminum/ Also look for “electroplating” in the comments.
That’s why a “star” washer is always used on earth screw. It punctures through the oxide layer and makes a low resistance connection.
That’s extremely helpful. I didn’t know that. Thanks!
Abrasively polish the area first.
White mineral oil can boil between 218-643°C so your results can vary. Some fraction of the oil will vaporize, so don’t breath it in – lung damage is guaranteed with enough exposure (fire breather’s pneumonia).
Other than that, why not put the aluminum plate in a frying pan, cover it in oil and heat up the whole thing on a hotplate to the point it melts the tin? 220°C is enough for SAC305 which is well below the 260°C auto-ignition temperature and probably below the boiling point of the mineral oil you have.
you’d still need to scrap some oxide layer off. hopefully while the oil is still cool so you don’t make a big hot mess of it.
The oil makes a film that stays on, so you can scrape it in a shallow dish on the side and then just lift it into the pan for tinning.
I would go with the sand idea for heating and pass the oil idea altogether.
Many heated oils emit carcinogens when heated to a high enough temp. It can’t be too high a temp because there’s warning labels on containers for transmission oil and crank (engine) oil.
So if your going to use oil when soldering then use an air extractor / filter.
I would expect the sand to wick the oil away.
the oil is to prevent the forming of a new oxide layer, not to help heating up the part.
I wonder if refined avocado oil could work? Its smoke point is 520F 270C
Buuut…. Avocado supply faces ‘big concern’ as inflation, scarcity send prices on a tear:
https://www.msn.com/en-us/money/markets/avocado-supply-faces-big-concern-as-inflation-scarcity-send-prices-on-a-tear/ar-AAUnGq6
That sounds like a good way not to be concerned too much about what’s happening to the oil and what it may be doing to your lungs and your body.
I will be looking into for cooking. Cooking oils cause trans-fats when cooking at higher temperatures.
I have used he oil method for 40 years. You don’t have to use mineral oil. Any high boiling point oil will do eg rice bran oil. The heating problem is no more than soldering copper so no big deal
Can someone independently confirm this please?
I have tried this method in the past and I’m reasonably confident it doesn’t work. With all the fake videos on tik-tok and YouTube of people appearing to do impossible things I guess I’ve become cynical when I see people with “tips” to do things that are widely accepted to be not possible.
I’ve done it. Smaller parts aren’t too difficult, but larger parts can be frustrating because the aluminum dissipates the heat so quickly. You will need a pretty hefty soldering iron, and preheating the part and keeping it insulated while tinning will help. Don’t forget the oil evaporating will also carry off heat.
Guys, you don’t need hefty. You needa Metcal, 900 degree induction iron. I have yet to find anything it won’tsolder with proper treatment.
FWIW I replicated this with a metcal and a sttc838 tip but did not at all find it easy. It did work better with no-lead solder, specifically tin/zinc/silver, than with leaded solder. This was on aluminum flashing, like 0.4mm thick.
A Metcal absolutely qualifies as “hefty”.
Depends on which tip your using. To the Metcal, it doesn’t matter though
I’ve done this and it does work. However, I’ve found it to be possible to get a joint that looks good but has almost zero bonding. I just use a screw and lug, it’s not worth the gamble.
I would be more convinced with a normal solder joint with a small amount of solder. That big blob just doesn’t look convincingly like a well connected solder joint.
Agreed. Should look the shape of a drop off water, not a drop of mercury.
I have done this with aluminum foil and regular cooking oil. It worked well. This was several years ago. A critical point is to apply the oil, then scratch at the foil surface under the oil (where there’s reduced oxygen availability to re-oxidize the surface), then solder the surface immediately. I only use lead-free solder.
I’ve been able to do it with cooking oil. But it only worked like 1 out of 14 tries.
It certainly felt like a strong bond when it did work that one time though.
Hm…yes. I think 30years ago my grandfather told me they used this trick before 2.world war. But if you realy try it you will find it is not so easy in real live. :)
I thought this was the purpose of flux. Maybe flux in electronics multicore solder only works with copper oxide not aluminium oxide. Ah I see special aluminium flux is a thing.
You can buy chemical (flux like) liquids and pastes to “weld” various dissimilar metals together at relatively low temperatures – lower temps than using a filler similar to the welded metals. Like Aluminium (Aluminum) to Stainless. They’re for capillary (lap) welds, not but welds, they don’t have the bond strength for but welds.
They’re amazing to use. If you can handle an iron and or a small flame then your good to go. They can be welded using butane / LPG / LNG so around 1800 Celsius or less.
I wonder if this would work with Stay-Brite 8 solder as a more foolproof way of repairing aluminum evaporator and condenser coils. Aluminum brazing rods exist for that purpose, but the melting point of the brazing material is only slightly lower than pure aluminum. If you’re not careful, you’ll melt the coil and ruin it.
I don’t when you last checked. Now you can get ultra low melt brazing rods with a special flux. As always, surface preparation will make or break the weld.
I had a fridge used in a very hot area and it started leaking from the factory weld off the compressor tube. Some goo seemed to be dripping out, it looked like some form of glue. And it was a flammable refrigerant so perhaps they didn’t like the idea of welding, perhaps the compressor was pre tested with the same refrigerant.
As for evaporators / compressors. If the leak is from mechanical damage like something hit it or rubbed against it for a long time then it’s probably woth repair. But if it just happened over time then the rest of the unit is ready to fail so it becomes and endless process of repairs.
Using an oil film to exclude surface oxides is a great trick. The telephone industry is so obsessive about grounding, they even specify this when working with copper ground busbars:
Smear the surface of the bar with petroleum jelly, then scour with Scotch-Brite until all visible traces of oxidation are removed. Wipe off the dust and smear with more fresh PJ and give another little swirl with the abrasive pad. Similarly prepare the mating face of the terminal being attached to the bar. Then install the hardware and torque to spec, which should squeeze the PJ out of the interface and result in true metal-to-metal contact.
Depends what the grounding is for. If it’s for lightning strikes this is fine as any remaining petroleum jelly would vaporize as the plates weld together during a lightning strike. PJ would provide some hydrolytic resistance but if you have one bolt in the middle it will be forced out (mostly) and any heat would do the rest because of the very low melting point.
This is going to create a poorly bonded joint. You need a shielding gas. Dry ice in a bowl of water in a plastic tub should do. Clean the oxide off and solder in the bottom of the tub where the CO2 has displaced the oxygen
The CO2 is fine. Water contains oxygen so give it a pass.
And you dont know dry ice
Definitely a good trick to remember.
I wonder if something similar could be done when welding aluminium, to mitigate the headaches when you run out of the appropriate gas.
you can oxy weld aluminium with the appropriate flux. I have used this flux to tig aluminium with CO2 one very desperate evening at about midnight when I had no gas left. it was not pleasant, atall, but it did work, Gave off a very bright orange flare (sodium) when welding too.
Why go through all the trouble when a screw, terminal and star lock washer works well?
And with a screw, terminal and star lock washer there’s no mess to clean up!
how about using paste flux, with a scrubbing pad?
Why? There is a special flux for aluminum: LaCo Aluminum Flux paste. Used here for example https://www.instructables.com/Hybrid-solar-panel-photovoltaic-and-thermal/
In the 1960s I used to solder to aluminum using a short, stiff fiberglass brush, similar to what is sold today as a “Jewelers Fiberglass Scratch Brush” (Google on eBay). You melt a blob of solder on the aluminum surface with a torch and, keeping the heat applied, stick the fiberglass brush through the solder blob and scratch away the oxide under the solder blob, which is now sealed from oxygen exposure. Works perfectly.
I’ve been doing this for many years, learned it from my grandfather. In my experience leaded solder will not wet aluminum no matter how you prepare it. It must be lead free, majority tin solder.
That’s why a star washer is always used on earth screw. It punctures through the oxide layer and makes a low resistance connection. Abrasively polish the area first.
Long ago I soldered aluminum using a fluoride based flux. It ate away the oxide layer. I think those items with toxic chemicals have been banned or restricted so they’re difficult or impossible to get anymore.
Guys the oil is to take the place of shielding gas, drill a small hole run the wire through it and loop and twist it together, then solder it with the oil there is also alumiweld rods at harbor freight same thing with them though you have to get through the oxide layer the oil is an awesome idea can’t wait to try it