Application · watch case & bracelet
Watch Case Welding: The Layer You Can See Is Not the One That Decides
The surface is finished. The gasket is already pressed into its groove. This one has been worn. The layer you can see is rarely the layer that decides the answer.
Micro welding on finished watch cases, bracelet links, clasp springbars and lugs — and a straight answer about how far down the damage goes before we stop.
Most of this page is about where we stop, because that is the half of the answer a bench actually needs. Not that a joint can be made. Which cases go back untouched, and why the reason for that is almost never the thing you can see.
There is usually a first time. A case arrives where the damage is not on a face at all — it is in the gasket groove, or on the sealing face itself. Everything that worked until that day worked because the damage had stayed in the top layer.
What damage on a watch case looks like, and which layer it sits in
Three layers deep, and only the first of them is visible with the case in your hand.
Damage on a watch arrives in one of three layers, and only the top one is the layer you can actually see.
Cases, bracelet links, clasp springbars, lugs — stainless and precious metal, and a finished surface sitting right next to wherever the joint has to go.
A lug knocked out of line or torn through at the springbar hole. A bracelet link cracked at the pin. A clasp that has worn through where it folds. Also on this layer, though nobody thinks of it as damage until it appears: heat tint on a polished surface, which does not buff back to the same finish it had. And one position on this layer is not inside a face at all — the line where a brushed flank meets a polished face.
The gasket groove and the sealing face are the only parts of a case that do their work by staying exactly the shape they were made in, and they are also the parts nobody looks at once the case is closed. A case whose seal has been lost and a case whose seal is intact are, from the outside, the same case.
Everything behind the case: the movement, the crystal, the seats they sit against. Nothing about this layer is a surface, and nothing about it can be refinished afterwards.
What puts all three in play at once is the state the piece arrives in. It has been machined, finished, assembled, sold and worn, and only after all of that did anything go wrong with it. The damage is the newest thing on the piece; everything it is sitting next to is older and already finished.
Where the weld sits in a watch's order of operations
By the time the piece is on your bench, every other station has already been passed.
On a watch, the weld sits at the far end of the sequence.
The order runs like this. Parts machined. Surfaces finished — brushed, polished, PVD. Assembly, with the gasket pressed into its groove and the tube seats and pins fitted. Then the watch leaves, and is worn. The damage happens out there, on a wrist, at some distance from anyone who makes anything. Only then does the piece come apart again — gasket out, and where the job needs it, movement and crystal out too. Only then is there a weld. And after the weld: the one face the heat reached is brought back, and the case goes together again.
Everything a maker could have done to this case while it was still bare has already been done to it. There is no station left upstream to move the joint into, because upstream ended before the watch was ever sold. The usual advice for a joint sitting next to something that cannot take heat — change the order of the operations rather than the machine — has nowhere to go here. The order is finished. The weld is the last operation, and it is being made on a piece that is complete.
This is also the change that happens to a bench, rather than to a piece. The day the work shifts from making new parts to repairing finished ones — a batch of service-centre returns, say — it stops being the occasional case that arrives worn, finished and closed. It becomes all of them.
If the gasket can come out, it comes out
The second time the same watch is on your bench, what has gone wrong is often not the joint that was made — it is the gasket that was still in place when the joint was made.
At that last station, the first thing that happens is that the second layer gets opened: the gasket comes out before any heat goes near the first.
A gasket is an elastomer, and heat takes it out of shape without giving the shape back. So it comes out first, it is kept on its own while the work is done, and it goes back into the groove it came from — the same seat, not an equivalent one, because a gasket and its groove have already worn to each other. After the weld and before anything is called finished, the sealing parts are checked back into their seats.
We do not weld next to a gasket that is still in place, and we do not publish a distance from the groove that is meant to be safe to work at. There is no such distance that holds across case shapes, groove depths and the age of the rubber sitting in them, and a number that fails on one of those is worse than no number at all.
Some gaskets will not come out — pressed in, bonded, or aged until they have gone hard against the groove. The same happens with a movement or a crystal that will not lift out of the case. Those pieces do not get a standard answer; they get looked at one at a time, and we do not tell you the outcome before the piece has been in front of us.
There is a second reason for getting the movement and the crystal out, beyond keeping heat away from them. With them still in, there is nothing to hold the case rigid against and nowhere to put shielding gas where the weld pool will be. Taking them out is what makes the joint something we can actually control rather than something we hope about.
What we will tell you at the end of that is one sentence, and it is deliberately a short one: the sealing parts are back in their original seats, and the groove faces are undamaged. Anything past that is a test on the assembled watch, and it is not something a welding operation can sign for.
How deep the damage sits is how far we go
One case we can keep to a single face. Five ways it stops being that case.
Whether this case gets welded is decided by how deep the damage sits — not by how large it is, and not by what the watch is worth.
Where the answer is yes, it is yes because the job can be held to one face: gasket out, the finished surfaces covered, one facet refinished rather than the whole case. That is the promise, and the rest of this section is where it stops being true.
-
The case we take.
The damage sits inside one continuous face; that face is a brushed or polished finish we can reproduce; the gasket will come out; the movement and the crystal will come out.
Where it stops: the first layer, and it stays there. -
Damage lying on the line between a brushed flank and a polished face.
Refinishing it means taking that line somewhere it was not, and once the line moves, the case reads as a different shape than the one it left the factory as. The work cannot be held inside one face, because the damage was never inside one. We hand this one back.
Where it stops: still the first layer — the position, not the depth, is what makes it uncontainable. -
A PVD, IP, gold-plated, enamel or painted surface.
A local repair does not put that surface back the way it was. This one does not go to a test weld first; the answer is that we don't weld this case. These finishes arrive on cases and bracelets that are already complete, with nothing behind them to fall back on if the answer turns out to have been wrong.
Where it stops: at the first layer too — because the first layer is the part that cannot be put back. -
Damage in the gasket groove, or on the sealing face itself.
The seal does not go back. This is the one failure on the whole list that nobody can see afterwards: the case closes, the faces line up, and the result is settled somewhere else, on someone's wrist, weeks later. We don't weld this case.
Where it stops: the second layer — the one that cannot be checked from the outside. -
A movement or a crystal that will not come out.
Then the weld would be made over a closed cavity, with nothing between the heat and the parts living in it, and the range of what can go wrong is no longer ours to hold. We don't weld this case.
Where it stops: the third layer, and there is nothing below it to stop at. -
An antique or discontinued piece carrying an original factory finish.
Refinishing it is not reversible, and there is no replacement to order if it turns out badly — the piece is the only one there is. We don't weld this case, and this is the one we decline earliest of all.
Where it stops: nowhere on the depth scale. It stops before the first layer, because what is at risk is not a surface but the piece.
Two metals in one piece: what joins, what doesn't, and what gets tested first
Three answers, and only one of them is yes.
Up to here the question has been where the damage sits. This one is a different question altogether: what the two sides of the joint are made of.
Like to like — 316L to 316L, 904L to 904L, 18K to 18K of the same colour and the same composition, titanium to titanium. Same metal on both sides of the joint behaves as one piece of metal once it has cooled, which is the whole reason this column is short and the other two are not.
The two-colour pieces: a precious metal case with a steel back, a precious metal springbar sitting in a steel lug. A joint between those can finish beautifully and still let go along the interface the first time the piece is worn — and it lets go at the join, which on a watch means it lets go while it is on someone. What we recommend instead is a mechanical joint, or a transition layer between the two. What we don't do is the version where it simply gets welded anyway and handed back looking correct.
Gold-filled, gold-plated and brass-cored parts. These go to a test weld before anything is quoted, and the joint form itself often has to be changed so that what comes out of the material during the weld has somewhere to escape to. We don't promise this one comes right on the first attempt, and a supplier who does has not welded many of them.
The one piece: we don't go looking for the parameters on it
The first discontinued piece to land on your bench is the day "order a new one" stops being an available answer.
The third thing that decides this job is neither the damage nor the metal. It is a rule we keep about the piece itself: the parameters never get found on it.
They get settled first on stand-in material — the same alloy, the same section as the piece that is actually going to be welded. That is not caution for its own sake. A parameter set arrived at by trying something slightly hotter and then slightly hotter again has to be paid for somewhere, and on a watch bench there is nowhere for that cost to land except the customer's only watch.
Before we accept the piece at all, the irreversible parts of the job get confirmed in writing: which face will be refinished, whether the gasket is being replaced, and whether the piece carries an original factory finish. Those three are the ones that cannot be walked back afterwards, which is precisely why they get agreed while they are still decisions rather than described once they are facts. The person who has to live with all three is often not the person who sent the piece in — it is whoever owns the watch.
And we don't tell you it can be redone if it goes wrong. A bench that runs one piece at a time, with no replacement to order for it, gets one attempt at every job it takes. That is the reason we would rather say no early than take a case we are unsure of: a supplier who says yes to everything is only useful to you until the first case where the honest answer was no. It also settles what a sample weld can be on this kind of work: what we can put heat into is an offcut or a scrap piece in the same material and the same section as your case — or material we hold to the same spec here — and never the watch itself.
Which machine this kind of watch work lands on
Two conditions, two different machines — and the choice between them is not settled on a page.
What decides the machine here has very little to do with how much power it has.
If you can't see it, you can't place it — and where a bench has multiple operators and long sessions, that condition points at the High Value Type. On watch work the reason is a specific one, and it is not the reason a dental lab or a frame workshop would give. One piece at a time and no replacement to order means that not seeing the joint properly is a gamble, and this is the piece you can't afford to gamble with.
Fine tack welding, small spot — that condition points at the EU Style, and on a watch it comes up on springbar tube seats and on lugs. Those joints sit down inside a recess in the case body, where the thing making them hard is not their size but that you are placing them into a spot you can barely reach and barely see from the only angle the case gives you.
Which of those two is right for your bench is not a question a page can answer honestly, because it depends on the pieces that actually come through your door. That is what the sample weld is for.
If you already know which machine you want, ask for a quote. If the question is neither a quote nor a sample weld, contact us.
Send something we can put heat into, and keep the watch on your bench
One condition, one action.
The condition first, because it is the part most enquiries get wrong: what you send is not the watch.
It is an offcut, or a scrap piece, in the same material and the same section as the case or the bracelet that needs the joint. If there is nothing on your bench to spare, tell us the material and the section and we will reproduce it here from stock held to that spec, and weld that instead.
With that in hand, the sample weld answers the things this page has been circling: whether the damage on your case stays inside one face, whether the gasket in front of you is one that will come out, and what your material actually does at the joint. The page that takes it from here sets out what to send and what comes back.