Help me solder 14k bezel and balls to silver. Struggling!

Hello everyone! This is my first post. (Please ignore double post-struggling with signing up to ganoksin too lol)

I really hope there are some brilliant minds that can help me with this. I have read a couple of the posts on here about this topic but I’m still struggling, so I thought I would ask a specific question or two (or 5) to get some advice.

I’m making a pendant. There’s a silver stamped wire (2mm) running 2/3 around a 14k bezel with 5 14k balls completing the perimeter of the bezel. All of this is on a 1.2mm thick 925 silver backplate.

I have tried a few different times and had “almost success”. We all know that almost success doesn’t cut it.

I feel like my issue has to do with heat management. On this latest try one of the five balls melted a bit more than the others and now the piece has to be remade again.

I am using an acetylene Smith Torch that uses atmosphere, not bottled oxygen. I have all of the tips available for it, should I be using the smallest one? A middle sized one? I believe I have been using a size one which is the middle one. My speculation is that is too big but I’m not sure.

I’m using boric acid in alcohol as my flux.

I have been heating the piece from below to heat up the silver, but maybe I should be heating the gold from above.

I am adding the gold parts first, then the silver wire. I am using 14 K hard solder and then the silver stamped wire with soft solder later.

I used a ball bur to create seats for the balls in the backplate and sweat soldered them to make soldering them easier.

I have no problem with silver to silver or brass to silver, I’m just struggling with gold to silver. Thank you for any insights!

Boric acid and alcohol protects the entire piece from firescale, it is not a good soldering flux. In addition to the boric acid and alcohol, you need a flux like borax powder (not boric acid) and water, batterns, Handi Flux and others. An acetylene plumbers torch like you have can be very hot and, depending on the flame shape, you can over heat one spot and under heat another. Practice bringing the entire piece, or at least the area that you are soldering, up to temperature at the same time. Also, it is very important that everything is clean and free of any polishing grit if you did any prepolish. It just takes time, but look into your fluxing operation. There is a lot in the archives about the differences between a barrier flux and a flow flux. One protects against firescale the other helps keep a joint clean and free of O2 while the solder flows. Good luck…Rob


When you say that you’re having problems connecting the gold to the silver, can you be more specific. For instance are you getting it so hot that you’re melting the silver and gold together into a blob or is nothing melting at all and things are falling off because they’re not actually soldered?

Also it might be helpful to take a quick photo or two and upload them if you can.



Thank you to you and rmeixner very much. I think I somehow made an assumption that the barrier flux performed both functions. So dumb, but I guess that’s learning!

I am definitely being fastidious about cleaning. All pieces get cleaned and ultrasoniced before soldering.

This last effort was SO close but one ball slumped more than the rest: here’s an image attached.

Would you add silver or gold first to the backplate?

Would you sweat solder the balls before adding them?

Would you heat from the bottom or top for this piece?

I’m so grateful for your help.

1 Like

There’s a number of different paths to approach this. Others might suggest doing it differently, but here’s what I would do.

First though, let’s talk about soldering gold and sterling silver as they tend to melt together fairly easily.

There’s two reasons. The first is eutectic bonding. Certain kinds of similar metals can bond together under heat (without solder). That’s the core concept of Keumboo and Mokume Gane. It’d be worth you doing bit of research on those techniques to find out what eutectic bonding is all about.

Next a common list of ingredients for gold solder is gold, silver and copper. Since sterling silver is 92.5% & 7.5% copper, if you solder sterling to gold and if you get it too hot, you’re basically making solder.

I can’t count the number of times in my life where I’ve melted gold into sterling when doing gold to silver overlay.

With all the in mind, when I solder silver & gold together I use silver solder, not gold solder. I have better luck that way. Others may disagree, but that’s what works for me.

In your case, I’d solder the gold bezel together with 14 kt hard solder. Then I’d solder the gold bezel to the silver base plate with medium silver solder. I’d put the solder on the inside of the bezel and would put 1 mm x 1 mm squares approximately every 5 mm. I’d coat everything with the same boric acid mixture, and would also add a bit of flux where the bezel connects to the baseplate. I’d heat primarily from the bottom, but would move the torch around a lot.

Then to solder the balls, I’d do what you did. Grind a place for each ball to sit with a ball bur. Inside each ground section, I’d pre melt a tiny bit of easy silver solder. Then add your boric acid mixture and flux like before. Add the balls where they’re supposed to go and solder together, probably heating primarily from the bottom again.

The trick with this is to be as observant as possible. Before the gold will slump into the silver because of overheating, you’ll see it start to happen. As soon as see a hint that might happen, pull the torch away. But if all goes well, the silver solder will melt before then.

I forget the numbers of the Smith torch tips, but I’d use a medium size tip. Not a big one and not the smallest.

Hope this helps!! I think you’re going to get it this time!



Oh my gosh! I’m SO grateful for your kind and informative response. I’ll definitely do some deep diving to strengthen my understanding of metallurgical processes.

I’m digesting everything you wrote. Just understanding what’s going on makes making mistakes much more tolerable.

Thank you!


whenever working with mixed metal, be very careful not to overheat… heat just enough for the solder to melt…

Jeff mentioned eutectic bonding… the eutectic point is at lowest temperature that bimetallic alloy melts according to its composition. For Silver/Copper the eutectic is about 30 weight% copper and 70 weight % silver, any more copper or any more silver in weight percent will increase the melting point…if you are trying to solder copper and especially brass which contains zinc, you almost don’t need to use solder since the two different metals in contact will seek out the eutectic themselves. The Gold/Sliver eutectic is at about 20 wt% gold and 80 wt% silver…either more gold or more silver raises the melting point… this is actually the way solders are made…
From the practical standpoint, use the lowest melting point silver solder to attach your gold balls to the back plate…and be very careful not to overheat…mixed metals in contact with each other basically want to make their own solder…as soon as the solder, that you put on melts, stop heating immediately…


just as an addendum: tri metallic alloys such as commercial solders that contain zinc in addition to copper and a precious metal are far more complicated as to liquidus/solidus phases that vary with composition.
They do not have a single eutectic point but curved surfaces that are near what would be the eutectic melting point of a bimetallic alloy. These are plotted on a triangle, with the apices being 100% pure metal. Melting point curves are superimposed on the composition plot…the lowest melting points of an trimetallic alloy can be further extended on to a 3D tetrahedral diagram with composition and temperature represented at the apices… I haven’t checked out the phase diagrams for a trimetallic alloys because they are so many of them. I do know that this is how we can look up the melting point of minerals that have a varied composition such as feldspar…which can contain sodium, potassium and calcium. All of this has been experimentally determined. For those who remember a post that I wrote sometime ago, in response to a sapphire cracking with heating, that went off the rails into deep earth processes and deep diamonds, color centers in gems, etc…all of which tested the limits of this amatuer geochemist’s knowledge, 3 composition phase diagrams of common minerals that form igneous rocks are what geologists and petrologists refer to all the time…what makes it more complicated is that ambient pressure makes a big difference in melting surfaces. Fortunately, these have all been experimentally worked out at pressures that are equivalent to going from the surface of the earth to the core itself, using a diamond anvil cell. These devices are made of gem quality diamonds for clarity. Mineral or metal samples such as iron, nickel and sulfur can be squeezed to pressures of 770 gigapascals or about 113 million psi, and heated to 2000 degree C with a laser…also fortunately we only have to deal with atmospheric pressure, or 14.7 psi… sorry for the digression but this is just FYI…

Returning back to the earth from outer space, or rather returning from the depths to the surface of the earth, there is a online free course on gold/silver/copper alloys with both the color of the alloys and melting points by Aldo Reti… just google silver/copper/gold phase diagram and Aldo Reti… I t’s useful if you want to make your own 18 or 22K gold of the desired color by adding silver or more copper to 22K coin gold…just adding silver alone gives a more greenish color at 18K…