Choosing a Laser Micro Welder Starts With the Part, Not the Machine
Three answers exist for any joint: we can set it up now, we weld one of them first, or we turn the work down. What nobody has written down is how you reach one of those three.
Nobody Can Make This Call From a Description of Your Part
The machine is the second question. The first one is whether the joint in front of you belongs to this process at all.
You came here to decide about equipment, and the honest first thing this page can tell you is that the equipment is not where the decision lives. It lives in the part. Two pieces that look alike on a drawing can land in different groups because of what one of them was dipped in, or what is fixed next to the joint and cannot come off.
So this page does not sell you anything in its first half. It hands you the same reasoning we run when a part arrives here, in the order the information actually reaches you.
- What you can settle by looking at the piece.
- What you have to ask someone else.
- What to do when nobody can tell you.
What you should leave with is one sentence — which of the three groups your part is in, and what that makes the next move. One sentence, because that is the shape the answer has to travel in: the person who signs off on this will never see this page, and a page of reasoning does not survive being repeated. A sentence does.
Next: the group that ends the question.
Two Kinds of Work We Turn Down
Read this first. If your part is one of these, nothing further down the page changes the answer.
- The question is not how thick it is. It is whether you need to join it or drive through it. A pulsed benchtop machine fuses at the surface and just under it. Wire, thin wall, sheet stock, tacking, putting material back where it was lost — all of that sits inside the band. If your joint has to fuse right through a solid section in one pass and then carry load on that weld, this is the wrong process for it. More power does not move that line; it only makes a wider, hotter weld that is still shallow. We give that answer early rather than late, because it does not change with budget.
- Two metals from different families, with load on the weld. Iron-based, titanium-based, copper-based, aluminium-based, precious-metal-based: if your two halves come from different families and the weld itself has to hold, this is that group. The seam can come out looking perfect and still let go the first time the assembly is used. The direction we would point you in is a mechanical joint, or a transition layer between the two — not a better setting.
- This is the one group where welding a piece of it first is not the safe move. A test weld on a pair like this comes back looking like a pass, and the failure arrives later, in service, with the part already in someone's hands. So when you are unsure whether your pair belongs here, the thing to settle is the question itself — before anything gets welded, not after.
If both of your halves are in that description, this is where the page ends for you, and we would rather you found that out here than after a machine arrived.
If your part is not in that first section, the rest of this page turns on a single question — whether we need to weld a piece of it before anybody commits to anything.
Two ways of answering it, in the order the answers reach you.
Three Things About the Part You Have to Ask Someone Else
Each one moves the part between groups, and none of them can be settled by looking at the piece in front of you.
- "It's silver" and "it's aluminium" are not answers yet. Two follow-ups. Is it pure? Sterling, aluminium alloys and copper alloys are easier to work with than the pure metals, but they still go through a weld on a real piece first. And is that bright face the body of the part, or a layer on top of it? If it was plated on, the metal underneath is something else, and the part belongs with the next item rather than this one.
- Zinc is the one you cannot see, so this is a question for your supplier, not for your eye. Three ways in. Is the piece brass? Brass is a copper-zinc alloy, so the zinc is in the metal itself, not sitting on it. What process put the bright layer on — galvanised, Dacromet, nickel, PVD? That answer belongs to whoever supplied it. And what surface work did the part go through before it was assembled?
- Set stones: the thing you need is the history, and the history does not show. Ask the chain three things. What the stone is. Whether it has been treated, and how — filled, irradiated, heat-treated, dyed. Who supplied it, and whether any of that can be documented. Treated material routinely looks untreated, which is why this stays a piece-by-piece call and not a rule anybody can publish.
Next: the half of it you can settle without asking anyone.
Two Things You Can Settle at Your Own Bench
The only part of this page you can act on right now, without asking anyone.
- Can you put it down twice the same way? Not whether the clamp is tight — whether the same piece, set in the fixture twice, puts the joint in the same place both times. Set it, take it out, set it again, and look at where the joint landed. If the second placement sits somewhere else, the weld is not repeatable, and no setting recovers that. What we need is a part that can be located the same way twice, which is a different requirement from being held firmly.
- Is there a route out for the gas? Find the path shielding gas would take to reach the joint and leave it again. Potted assemblies, sealed cavities, and joints sitting behind a component that is already fitted usually have no such path, and a joint the gas cannot reach oxidises whatever else is right.
Fail either one and the part is in the group we weld once before anyone commits — not because the metal is difficult, but because the situation is.
Next: what to do when an answer never comes back.
When the Answer Does Not Come Back, Take the Harder Reading
Two of the questions above have a default, and both defaults run the same direction.
- Bought-in fasteners, hardware and standard parts, where nobody upstream can say what the coating is: treat them as zinc-bearing. It is the commonest gap in the chain, and it is the one people are most tempted to wave through, because the piece looks like plain steel.
- A stone whose history you cannot get: treat it as treated. It stays a piece-by-piece call. A missing document is not permission to release it as untreated.
This is the section people skip, because a blank feels like a reason to wait for someone to get back to you. It is not. A missing answer is itself an answer — it puts the part in the group we weld once first, and you can carry that conclusion out of here today rather than parking the decision until an email arrives.
One group has not been dealt with yet.
Everything above sorts a part into work we turn down, or work we weld once before anyone commits. There is a third position. This page has been holding it back — not because it is good news, but because reading it first would let you skip the two that come before it.
It is narrower than it sounds.
The Work We Can Set Up Without Welding One First
The group that has been on our home page all along, without the reasoning that gets you into it.
- Cobalt-chrome, titanium, stainless and precious metals — same metal on both sides, or the same family. A fine section: wire, thin wall, sheet. And nothing heat-sensitive fixed beside the joint that cannot be taken off first. That combination is routine here, which means we can go straight to settings rather than asking you to put a piece in the post.
- And it still is not a promise. Two metals from that list are not automatically this group: the pairs we turn down are drawn from these same four. What puts a joint here is the pair being same-metal or same-family, not the metals appearing on a list.
- What this group means is narrower than it sounds, and it is worth being exact. It means we do not need to weld one of them first before we can give you settings. It does not mean the job cannot go wrong. Parts still arrive carrying a history nobody wrote down. The difference is where the uncertainty sits — not whether it exists.
All three groups now have an answer. One thing left to do with it.
Say It in One Line
The person who signs off on this will not read this page.
Whatever you concluded, it leaves here as one sentence with two halves: which group the part is in, and what that makes the next move. Two halves, because a group on its own invites the question so what do we do, and an action on its own invites why.
This part is in the group we , so the next step is .
That is the whole deliverable. Not a form to fill in, not a folder to assemble — a line short enough to be repeated accurately by somebody who was not standing at the bench when you worked it out.
Three ways this ends, one for each group.
Three Ways This Page Ends
One for each group, and they are deliberately not the same size.
- If you stopped at the first section, you already have your answer, and there is nothing here for you to click. The direction is a mechanical joint or a transition layer, and that conversation belongs with whoever designs the part.
- If it landed in the last group, the open question is no longer whether, but which of the four suits the work — that is the next page.
- If your part needs one weld before anyone commits:
One of those three has no next step, and that is the honest answer for it.