PDA

View Full Version : Soldering copper to aluminum



Evan
11-04-2011, 12:45 AM
I am making some 1.2 ghz yagi antennas for radio astronomy, in order to detect the emissions from bursts emitted by solar flares. I wish to make the antenna elements from 4032 aluminum welding rod as it is corrosion resistant, stiff, cheap and easy to bend. The only issue is making good connections to the driven element that won't corrode quickly. I will be using 300 ohm twinlead for the phasing harness and soldering is the most secure connection possible. However, it is impossible to solder copper to aluminum directly with regular electronics solders.

I dropped in the local Canadian Tire store and checked out the welding supplies. They carry a product from Bernzomatic that is advertised for soldering aluminum to aluminum. Checking out the label it is an alloy of copper, zinc and magnesium and doesn't require flux.

I was skeptical but bought some to try. Surprisingly, it works well to solder 4032 to 4032 but unfortunately it will not flow to copper. Then it occurred to me that perhaps it would act as an intermediary for regular lead solder. Yes it will.

By "tinning" the bare aluminum with the Bernzomatic rod and then tinning that with the tin/lead solder it is easy to solder copper wire to the aluminum part. The absence of any highly corrosive flux insures the connection will not corrode. It requires only a propane torch to reach soldering temperature of the special rod.

http://ixian.ca/pics9/alsolder2.jpg

http://ixian.ca/pics9/alsolder1.jpg

http://ixian.ca/pics9/alsolder3.jpg

lakeside53
11-04-2011, 12:52 AM
Good tip. Could be useful.

I find you can usually "tin" solder onto aluminum with standard flux core solder if you use the iron to scrape the aluminum surface (to get though the oxide) as you apply solder. I've also used a stainless brush though a puddle of solder. Once it's "tinned" the rest is easy.

The benzomatic rod makes torch soldering viable.

I've also used HTS- 2000 rod for aluminum brazing. http://www.aluminumrepair.com/index.asp

I wonder of the benzomatic product is similar?

darryl
11-04-2011, 12:56 AM
Interesting. I'll keep that in mind.

Evan
11-04-2011, 12:56 AM
I have never been able to get an actual bond to aluminum with any ordinary solder. You can get it to stick to the aluminum but it isn't flowed to the metal. It is not a true solder connection where the metals mix at the interface to form an alloy.

MichaelP
11-04-2011, 01:21 AM
Interesting. Thank you.


It is not a true solder connection where the metals mix at the interface to form an alloy.
I have to respectfully, but strongly disagree. Metals (parent and the soldering alloy) do not mix and form alloy as a result of soldering or brazing. Parent metal does not melt. This is the difference between those two processes and welding.

Your Old Dog
11-04-2011, 01:34 AM
Thanks for the post. It could be of some use to Ham radio operators.

jnissen
11-04-2011, 01:48 AM
True standard solder will just ball up and not stick to aluminum. The two step approach is a good tip. We ran into this about a two years ago. My son was building a fuel cell and used an aluminum perforated plate in place of traditional graphite electrodes. Anyway he punted at the time and used a brass contact to make a mechanical connection to the aluminum. Not as reliable but it worked for the test at the time. This would have allowed him to directly make a good soldered connection.

lakeside53
11-04-2011, 01:54 AM
If you work the surface as you heat and apply lead solder (you will need a decent iron to get the aluminum up to temp), it will stick nicely. if you don't, it will "ball up" like you say. I once soldered hundreds of wires to aluminum (way back). I suspect they are still there;)

Pre-tining with benzomatic or other modern materials is an attractive alternative.

I re-read the HRS-2000 site. They say that rod can be used to braze to both copper and aluminum (and others). I have a bunch of that rod - I'll go try it as a "pre-tin".

beanbag
11-04-2011, 03:48 AM
Interesting. Thank you.


I have to respectfully, but strongly disagree. Metals (parent and the soldering alloy) do not mix and form alloy as a result of soldering or brazing. Parent metal does not melt. This is the difference between those two processes and welding.

Then what do you call it when a gold wire dissolves into a puddle of tin/lead solder, even though it is below the melting point of gold?

The Artful Bodger
11-04-2011, 03:57 AM
The old technique for soldering to aluminium was to make the joint under a puddle of oil.

Apply the oil and rub the oil covered surface with a big soldering iron (hot of course) then apply the solder. The requirement is for the soldering iron to clear the oxide off the aluminium and to complete the join without air getting onto the metal.

Evan
11-04-2011, 06:40 AM
I have to respectfully, but strongly disagree. Metals (parent and the soldering alloy) do not mix and form alloy as a result of soldering or brazing. Parent metal does not melt. This is the difference between those two processes and welding.

The metals form a eutectic at the boundary. Solders act as a solvent for the metals with which they are compatible. The metals form a eutectic at the boundary. The result is an alloy at the interface. It is not the same as welding. In particular the welding process refers to joining the same metals together by melting both parts at the junction.

Boucher
11-04-2011, 07:46 AM
There used to be a company that specialized in joining difficult materials. They had eutectic in their name. Their products were a little pricey but they really did work.

Tony
11-04-2011, 09:28 AM
Good tip!

Just curious though, how does this soldered connection compare (electrically)
to a mechanical connection? meaning using a small screw.. or perhaps drilling
in the end of the aluminum rod, inserting the wire, and crimping it?

Tony

Weekend_Scientist
11-04-2011, 10:28 AM
Hey Evan,

That's a great tip. Thanks for the details.

I'm going to have to make up some aluminum-copper connections just so I can her my other builder/tinkerer friends say "How'd you do that!?!?"

Speaking of difficult to solder materials, does anyone know a material that can be used to solder to tungsten? I'm talking tungsten metal not tungsten carbide.

Tony
11-04-2011, 10:42 AM
Weekend.. steel will work just fine.. put some tungsten in a tig torch and
try to weld some steel without touching it.. in no time flat you'll have a very
strong tungsten-steel bond! :) :)

JABoyce
11-04-2011, 11:36 AM
The soldering process does not form a eutectic. The soldering alloy (not the base metals being joined) are eutectic. Eutectic merely means that an alloy of more than chemical compound that transitions from solid to liquid (and the reverse) at a single temperature. Non eutectic alloys solidify (or become liquid) in stages, at the solidus temperature of each constituent of the alloy. Because of this they have a plastic stage.

Eutectic alloys are useful for soldering as they do not have a plastic phase. The transition from solid to liquid and back is a single specific temperature.

The topic of whether or not soldering causes the solder alloy to alloy with the base metals is still debated. Some sources claim the joint is a "glue joint", while other claim the alloyed joint.

So at this point, I don't think the science conclusively proves the glue joint or the alloy joint...

Just my .02

Jeff

macona
11-04-2011, 11:42 AM
Harris sells an Aluminum solder kit:

http://www.weldfabulous.com/Filler-Metals/SolderFlux/Aluminum/Harris-Al-Solder-500-Aluminum-p5918946.html

This works at normal soldering temps and will stick to a lot of things. It work especially nice for soldering Lithium Polymer battery packs since one terminal is aluminum and the other is iron. It is a Zinc-Tin solder with a special flux.

RWO
11-04-2011, 11:49 AM
The topic of whether or not soldering causes the solder alloy to alloy with the base metals is still debated. Some sources claim the joint is a "glue joint", while other claim the alloyed joint.

So at this point, I don't think the science conclusively proves the glue joint or the alloy joint...

Just my .02

Jeff

Then what, in your opinion, causes a bare copper soldering gun tip to slowly disappear when used with lead-tin solder.

Now, if you iron plate the tip, it seems to last nearly forever. Could it be that iron is nearly insoluable in molten lead-tin alloy while copper is just the opposite?

RWO

Chris S.
11-04-2011, 12:02 PM
Then what, in your opinion, causes a bare copper soldering gun tip to slowly disappear when used with lead-tin solder.

Now, if you iron plate the tip, it seems to last nearly forever. Could it be that iron is nearly insoluable in molten lead-tin alloy while copper is just the opposite?

RWO

The casing you see on soldering iron tips are not iron, it's a silver/nickle alloy. An iron tip would never tin properly, if at all, and if you can't tin an iron it makes for very poor heat transfer and crappy solder joints.

Chris S.
11-04-2011, 12:05 PM
Good tip!

Just curious though, how does this soldered connection compare (electrically)
to a mechanical connection? meaning using a small screw.. or perhaps drilling
in the end of the aluminum rod, inserting the wire, and crimping it?

Tony

Two very unlike metals = electrolysis corrosion. Ask any boater. ;)

lakeside53
11-04-2011, 12:10 PM
There are billions of mechanical copper to aluminum connections - electrical cables... rated devices simply use a plated surface on the connector and aniti-oxidant grease on the aluminum.

Evan
11-04-2011, 12:20 PM
The soldering process does not form a eutectic. The soldering alloy (not the base metals being joined) are eutectic.

Not all solders are eutectics including not all tin/lead solders. So called "radio" solder is a eutectic alloy since the eutectic alloy of tin and lead is close to 60/40 ratio. A eutectic alloy also has the property of having the lowest melting point of any alloy ratio of two metals. Because of that when an alloy is formed by dissolving a higher melting metal in a molten lower melting metal a eutectic alloy is automatically formed.

As for any "debate" regarding whether an alloy is formed, there isn't any debate in scientific circles. In order to solder at all the substrate must be soluble in the solder alloy and will form an intermetallic alloy at the joint.

Evan
11-04-2011, 12:23 PM
It never occurred to me that there might be anything controversial about this subject. :rolleyes:

RWO
11-04-2011, 01:12 PM
The casing you see on soldering iron tips are not iron, it's a silver/nickle alloy. An iron tip would never tin properly, if at all, and if you can't tin an iron it makes for very poor heat transfer and crappy solder joints.

You are wrong about iron plating. See http://www.all-spec.com/products/8125N.html

I know some tips are now nickle plated, but the original plating for long lived tips was iron. Haven't you ever soldered something made from steel? It's almost as easy as copper with the right flux.

RWO

Optics Curmudgeon
11-04-2011, 01:17 PM
Regardless of the controversy, I'm glad to know that there's a product out there that will make soldering aluminum to copper easy.

Allan Waterfall
11-04-2011, 01:58 PM
Then what, in your opinion, causes a bare copper soldering gun tip to slowly disappear when used with lead-tin solder.

RWO
In my experience that only happens with that crappy lead free solder,I won't use it anymore for that very reason.
Took me a while to realise that it was the lead free solder causing loss off copper from the bit.

Allan

JABoyce
11-04-2011, 02:06 PM
Some of the corrosion of the copper soldering tip is from galvanic action, some is from the acid in the flux, and some does end up alloyed into the joint.

The intermetallic bond is a result of the wetting. Surface tension and friction resist wetting. The interatomic attraction between the solder alloy and the base metal overcomes the resistance to wetting. While the metals at the intermetallic bond commingle, and form a very thin layer (1-3um) of substitutional alloy. Tin/Copper (SnCu) being the most common.

An intermetallic and substitutional alloy are not the same thing. Colloquially an alloy is a mixture of metallic elements in solution. Scientifically an alloy is a homogeneous mixture of metallic elements in solution. An intermetallic is not homogeneous.

On the other hand, there is also a colloquial defintion of intermetallic that can include all alloys. Complex, simple, homogeneous and non-homogeneous.

There are elements of chemical (atomic) bonding, and mechanical bonding of the crystal structures in the intermetallics of a solder joint. So many feel the bond to both an alloy and a "glue".

I guess the debate is not on the atomic structure of the solder joint, but on the terminology used to describe it. The definitions of "intermetallic" and "alloy", and how they overlap is where the debate is.

Just my .03

awemawson
11-04-2011, 02:22 PM
This is for a yagi aerial element as I recall, so presumably exposed to weathering. My concern would be electrolytic corrosion.

When aluminium first became available as a sheet at reasonable prices, a well known boat builder decided to make a lifeboat from it. He used the traditional coppr riveting that had worked so well for clinker built wooden boats. After launching and during tests they were puzzled by the fizzing sound and a little later by the spurts of sea water coming from the rivets !

Chris S.
11-04-2011, 03:39 PM
There are billions of mechanical copper to aluminum connections - electrical cables... rated devices simply use a plated surface on the connector and aniti-oxidant grease on the aluminum.

Yes, that can be done too.

Evan
11-04-2011, 04:32 PM
One of the problems of using a strictly mechanical connection is due to the dimensional constraints when dealing with gigahertz frequencies. There isn't room for a reliable physical connector that won't disturb the electrical properties of the antenna. The spacing of the elements, the dimensions of all parts and the connections to the driven element are all critical. For instance, to achieve close to the theoretical maximum gain the lengths of the antenna elements must be correct to within a tolerance of about +- .007".

Building such an antenna array isn't just a matter of bending up some wire. It actually becomes a machining exercise.

As for corrosion of the joints and the rest of the antenna it will be sprayed with clear lacquer for protection. I will post pictures of the antenna when I have something to show.

While on this topic does anybody have a suitable downconverter lying around that will work with a Yaesu FRG 7700 that they don't need and would like to trade? :D

The antennas are designed for the 1296 mhz amateur band.

Chris S.
11-04-2011, 09:56 PM
That's for sure. At that freq a screw terminal could look inductive, capacitive or both. ;)

George_Race
11-05-2011, 06:51 AM
I have used solder-it paste for years to solder to aluminum chassis while building electronic projects. Really works well. Here is a link to more information.
http://www.solder-it.com/shop/item.aspx?itemid=89
And yes, I am a Ham Radio Operator
George

Evan
11-05-2011, 09:05 AM
Never seen that here but this is a small and isolated town. Looks useful. I will have to see if I can find a distributor that will ship here and that doesn't use UPS.

Bill736
11-06-2011, 01:12 AM
One of the problems of using a strictly mechanical connection is due to the dimensional constraints when dealing with gigahertz frequencies. There isn't room for a reliable physical connector that won't disturb the electrical properties of the antenna. The spacing of the elements, the dimensions of all parts and the connections to the driven element are all critical. For instance, to achieve close to the theoretical maximum gain the lengths of the antenna elements must be correct to within a tolerance of about +- .007".

Building such an antenna array isn't just a matter of bending up some wire. It actually becomes a machining exercise.

As for corrosion of the joints and the rest of the antenna it will be sprayed with clear lacquer for protection. I will post pictures of the antenna when I have something .
While on this topic does anybody have a suitable downconverter lying around that will work with a Yaesu FRG 7700 that they don't need and would like to trade? :D

The antennas are designed for the 1296 mhz amateur band.

But Evan, we all know that, in the final analysis, your precision antenna will work well only after you start hanging pieces of aluminum foil on it !

darryl
11-06-2011, 01:23 AM
And don't forget to wrap up your smart meter in aluminum foil before firing up the receiver :)

Evan
11-06-2011, 03:24 AM
The smart meter is being dealt with. I am in discussions with BC Hydro right now over the potential interference problem. They don't know what to do. Nobody has complained that Hydro's meters will interfere with their radio telescope before. I am suggesting they do the same for me as they will certainly be doing for the Dominion Radio Astrophysical Observatory in Penticton, which will be a wired smart meter or even the old style. After all, I am a long time member of the Royal Astronomical Society of Canada and will be doing the same type of research. In the past I have contributed earthquake data from my seismograph to the Pacific Geoscience Centre. Amateur astronomers routinely contribute data to the professional astronomers.

Interference is a high probability since I also intend to build a wide band helical array to cover from about 800 mhz to 1200 mhz with a wideband noise receiver. We are in a radio hole here with very little interference in that range. If they don't make any accommodation then I will be building a stylish little shielded box for it with a screen door than can be closed when necessary. If they don't like that they can go suck eggs. They don't have a legal leg to stand on since the meters are not permitted to interfere with anything. If necessary I will get my amateur ticket which will give me legal priority to them on the 902 to 928 ISM band since Ham radio is a secondary licensed user of that band and they are unlicensed.

Evan
11-06-2011, 03:27 AM
But Evan, we all know that, in the final analysis, your precision antenna will work well only after you start hanging pieces of aluminum foil on it !

Precision is relative. +-.007 isn't a very demanding tolerance.

brickie
11-06-2011, 05:46 AM
Try this link. Ed


http://www.muggyweld.com/1clip7.html

Evan
11-06-2011, 06:06 AM
What I dislike about that and the Solder-it products is that they both rely on very active fluxes with the potential to produce corrosion after the join. This method I am using doesn't use any flux other than that found in regular radio solder.

It does look worthwhile for joining aluminum to aluminum for small projects.

Thruthefence
11-06-2011, 10:28 AM
Evan, you are "wise in the ways of science.....";

What is your take on the Lightsquared/GPS spectrum controversy?

Evan
11-06-2011, 11:43 AM
I really don't have an opinion on that particular issue. It's a clearly definable problem in technical terms, the rest is all politics. Spectrum is worth an incredible amount of money and they aren't making any more of it. Any time there are trainloads of cash involved science issues tend to become a victim of war along with truth.

How much trouble will it cause, if any? That depends on who you ask and who pays their salary.

The smart meter issues are the same even if you leave out the lunatic fringe. The law says that part 15 devices must not cause interference (part 210 in Canada, same rules). The problem with that is that anything that transmits will cause interference to something sometimes. So now you have an absolute legal paradox and that immediately throws the problem into the political arena since it is politicians that make the laws.

There are some very interesting aspects to the smart meter and smart grid program. One in particular is the absolute insistence on the part of the power companies that the system must be real time and bidirectional to to work. The companies have wrapped themselves in green cloaks claiming that the smart grid will save the world by reducing energy consumption.

Even if you buy that line it is easy to demonstrate that a smart grid doesn't require real time bidirectional communications to the household level to function. All the meters require is the ability to receive commands in real time and that can be easily provided on numerous existing parts of the spectrum from ground FM radio sideband signals to direct satellite radio. Return data on consumption can be stored in the meter and sent by many possible means on a monthly basis.

The decisions to change daily and hourly operating parameters on the grid in order to save energy will never be made based on the information provided by a single user. It will be based on overall energy use and availability in entire regions which the system operators already have. That information is available minute by minute.

So why the insistence on having the meters all connected in a high speed IPv6 network?

It doesn't take much imagination once it is pointed out to see where that leads. The power companies want to become information service providers. Once the smart meter system is in place it is tiny extra step to start selling you internet, telephone and related data services. The meters are even equipped to interface to your home wireless network directly.

Supplying you not just with electric power but all your communications needs is worth an enormous amount of money. And that is the rest of the story.

Chris S.
11-06-2011, 12:43 PM
Beware of liquid solder flux. For many years, when I've needed additional rosin flux, more than what's held in cored electronics solder, I've used solder paste. It just more of what's in cored solder, which I've never found to be conductive. Then one day, while designing a one off microcontroller board I grabbed some clear liquid flux that my customer had in their Bio-Med shop. When I did a test run the ADC input/output wasn't working anything like it did on the protoboard. I examined all of my joints under magnification and everything looked looked good. So what was the problem?

I took some voltage drop readings across my resistive divider on the ADC input and they were all wrong! I then took some Ohmmeter readings, with the power off, and they were all messed up too!

I then stuck the probes in the container of liquid flux and eek :eek: a half inch gap read <10K! The board required more than 4 thorough solvent scrubbing's, with a tooth brush, to get this stuff off. All worked fine afterwards though.

FYI: For those that will say that I should have defluxed it from the get-go... I usually do that operation after the basic I/O testing. ;)

Chris S.
11-06-2011, 12:58 PM
Spectrum is worth an incredible amount of money and they aren't making any more of it.

That's why the Amateur Radio motto has been "Use it or lose it.", ever since the loss of the 11 meter band to CB. ;)

Daminer
11-06-2011, 01:15 PM
Don't know if the flux would be an issue, but this stuff has been around for awhile and can be a lifesaver at times.....It's quite pricey, and I don't know if an alternative to UPS is available.....

http://www.muggyweld.com/index2.html

Evan
11-06-2011, 02:14 PM
I then stuck the probes in the container of liquid flux and eek a half inch gap read <10K! The board required more than 4 thorough solvent scrubbing's, with a tooth brush, to get this stuff off. All worked fine afterwards though.



That was a lesson I gained way back. I had several jobs in succession in electronics after I dropped out of Berkeley. One was at a company that built radiation detectors of all kinds. They hired me in quality control. They had one model that was used for detecting alpha by means of ionization in an atmospheric air chamber with an ultra thin conductive rubber membrane cover and a detector grid. The detection circuit input stage was a differential amp that used the then very latest in transistors, a dual mosfet. Because of the extremely high input impedance it was imperative that the chamber and all insulation for the detector grid supports be cleaner than Mr. Clean. Even a finger print would cause problems.

They were having problems with leakage and erratic operation of the finished units and nobody other than the long gone original engineer who designed it knew anything about mosfets. I developed a washing protocol for the detector chamber after assembly that required a three stage wash in an ultrasonic bath using acetone, distilled water and anhydrous ethanol to remove every trace of conductive contamination. That solved the problem completely.

Evan
11-06-2011, 02:16 PM
Don't know if the flux would be an issue, but this stuff has been around for awhile and can be a lifesaver at times..

You are the second person to mention Muggyweld so I guess it must be worth a try. I will have to see about getting some. I always like having as many options as possible in metal working and fastening.

beanbag
12-30-2011, 04:57 AM
Online descriptions of this aluminum brazing rod says it contains flux. How do you know it isn't corrosive to the copper?

Evan
12-30-2011, 09:28 AM
To which do you refer?

beanbag
12-30-2011, 04:09 PM
benzomatic

Evan
12-30-2011, 04:31 PM
It's plain metal alloy, no flux at all.

EVguru
12-31-2011, 06:47 AM
I have a roll of Aluminum solder at work. I think it has some silver in the alloy and the flux is quite aggressive.

It works very well for electrical/electronic work, but its mechanical strength isn't that good.

I mostly use it for making connections to PT100 temperature sensors. The wires on these seem to be Nickle plated Iron and can be very reluctant to wet with conventional solder. Once wetted with the Aluminium solder, I wipe them off and use the ordinary stuff.