Your Cutting Head Model Is Discontinued — What Now?

A diagnosis-first guide for the workshops that repair these heads: how to tell a repairable head from an uneconomic one, and which assemblies to order when you go back in.

Short answer: almost any cutting head can be repaired, and the parts that fail are usually still available. The real question is whether repairing it costs less than replacing it — a number, not an opinion.

This guide is written for the service engineers who do the work: what to inspect, what to test, and which assembly to order.

Discontinued is the wrong first question

Nobody finds out a head is discontinued on the day it fails. You find out later, when the quote comes back, or when a part number comes back with a note attached.

The failure comes first. The kerf drifts, the head stops following the sheet, the edge goes rough on 12 mm stainless. Obsolescence is a constraint you meet on the way to fixing a fault; it is not the fault.

So the useful question is not what to replace the head with. It is which assembly has failed, whether that assembly can still be bought, and what the customer is prepared to spend on it.

What actually fails

In our own repair records the split is lopsided. Almost every head that arrives is an optical fault, and most of those leave with the same pair replaced: a protective window and a lens group. A little under half also need the spring-loaded seals that sit behind them.

We have not written off a head as beyond repair. Not one.

That is a narrower picture than most writing on discontinued heads implies. The other failure modes are real — about one head in eight shows some electrical fault, and a mechanical one turns up roughly once a year — but they are the exception, not half the story. If a head is cutting badly, the problem is almost certainly in the optical path, and the answer is a set of windows, lens groups and seals.

This is close to a census rather than a sample. A cutting head is not repaired in the customer’s own workshop — the bench conditions are not there, and neither are the spare parts — so nearly every head that fails comes to a shop like ours. Returns are split roughly evenly across machines below 3 kW, 4–8 kW and above 10 kW. Old low-power machines are not a niche; they are a third of the work.

Diagnose by symptom, but start in the optical path

Not because it is cheap to check, but because in our records it is right nine times out of ten.

● Reproduce it on known-good material and a known-good program. A share of “the head is failing” calls are a parameter or a gas problem, and that takes ten minutes to exclude.

●  Swap one variable at a time — head, fiber, source, gas line. Change two and you learn nothing about either.

●  Open the head and look at the optics first: protective window, collimating lens module, focusing lens module. That is where the fault was, almost every time.

●  Electrical and mechanical checks last. Together they are a small minority.

If the fault follows the head — move it to another machine and it cuts badly there — the head is your problem. If it stays with the machine, it is not the head, and no amount of parts will fix it.

The height sensor: the most common failure the return statistics miss

Ask the after-sales desk which part actually fails most often, and the answer is wider than the repair bench suggests. The height sensor is one of the most common failure points on a cutting head — it is simply not what gets packed up. A sensor travels on its own: returned to the factory for repair as a part, or replaced with a new one, because it is a component you can swap on site rather than a head you have to crate and ship. That is why it barely registers in return statistics and turns up constantly in support calls.

A dirty window and a drifting sensor produce the same complaint: a bad edge on thick plate. What separates them is the pattern. Optics fail consistently, the sensor fails intermittently and follows the sheet.

The symptoms that point at the sensor rather than the optics:

●  The head stops following the sheet, or dives into it at the start of a cut.

●  Height readings jump, or drift over a shift.

●  Intermittent height alarms mid-program, with no repeatable trigger.

●  A cut that is fine on flat plate and bad on wavy plate — the fault follows the sheet, not the head.

Three checks settle it:

●  Jog the head over a flat plate and read the height signal at the controller. It should be stable and linear, not stepped or noisy.

●  Swap in a known-good sensor. It is the fastest test in this article, and the one most often skipped.

●  Inspect the sensor face for spatter, the sensor's lower assembly for cracks, and the connector and cable at the head end.

Then recalibrate. In most cases the head itself is fine — which is why replacing a head over a sensor fault is the most expensive mistake on this list.

We cover sensor diagnostics, calibration and the differences between models in a separate guide. It is here because a fault list that leaves out the failure you meet most often is not a fault list.

Replacement is not drop-in

When the repair stops making sense, five interfaces have to agree before a new head will run on an existing machine:

●  The mounting interface: bolt pattern and flange geometry.

●  The beam connector: QBH, QD, LOE3.2 and Q+ all appear among the options on the heads we supply, and they are not interchangeable.

●  The focal length range: the head has to reach the thickness and the standoff your process needs.

●  The height sensor connection: signal type and connector.

●  Gas and cooling ports: flow, pressure, and the fittings on the machine side.

Anyone who tells you a model is a drop-in replacement without asking for the nameplate and a photo of the flange is guessing. It is a guess that costs a machine a week.

Four control questions

On top of the interfaces, four questions decide whether a new head drags a control retrofit behind it:

●  Can the controller drive this head’s focus axis?

●  Does it accept this height sensor’s signal?

●  Does it have a focus profile for this model?

●  Does the safety interlock still work with it?

If the first answer is no, you are not buying a head, you are buying a project. If the second and third are no, you are buying a control retrofit as well. That is the point at which a cheap replacement stops being cheap.

There is no unrepairable head, only an uneconomic repair

There is no unrepairable cutting head. There is only a repair that costs more than a new one — and that is a number, not an opinion.

In our own returns we have written off none. Not because every head was worth repairing; plenty were not. But “unrepairable” and “not worth repairing” are two different verdicts, and only the second one is real. A head that has lost its optics, its seals, its sensor or its motor can be brought back. What ends the conversation is arithmetic, and it has three lines.

●  The part costs too much. Sometimes the assembly is out of production and the way back is a substitute part or a machined one. That is still a repair — until the substitute costs about what a new head costs, at which point the arithmetic has answered for you.

●  The wait costs too much. The part exists, but it lands after the customer’s deadlines. For planning purposes a part that arrives too late is not available, and the decision stops being whether to replace and becomes when.

●  The certainty costs too much. The head has already absorbed a certain amount in repairs, and the next failure is a when, not an if. Past that point the money stops buying reliability.

Notice that none of those three is a technical verdict. That is why the honest answer to “can this be repaired” is almost always yes, followed by a price and a date. A quote that opens with “this model is obsolete, you need a new head” is not a diagnosis; it is a price with the arithmetic hidden.

What to order, and what to keep on the shelf

The parts that end an old head’s life are not the expensive ones. They are the specific ones — a window, a lens group, a seal kit for one model — that nobody stocks because they cost little and fail rarely. Until they fail, on a job with a deadline.

If you service several machines, the sensible holding for each model family is one optical set and one seal kit. For service providers, a first order that covers windows, one lens assembly and a seal kit for the model families you see most often is what keeps the second visit from happening.

To quote any of it we need four things: the head model and serial number, a photo of the nameplate, a photo of the failed assembly with its part number visible, and the quantity you want to hold. With those, the answer is normally the same working day — part number, price and lead time.

If you are choosing a replacement head

Five things decide the spec, and the quote is usually same-day:

●  The head model and serial number.

●  A photo of the nameplate.

●  A photo of the mounting flange, or the Fiber Interface Connector.

●  The controller brand and model. A photo of the controller screen is usually faster than hunting for the model number — send it and we will identify the controller for you.

●  Laser power, and the material and thickness you cut.

FAQ

Can a discontinued cutting head still be repaired?

Usually yes. Discontinued means the model is no longer built; it does not mean the assemblies inside it have vanished. Check the window, the lens group, the sensor and the seals first — most of what fails is still available.

Can you match a part if I only have the head model?

Yes. Send the head model and serial number, a photo of the nameplate and a photo of the failed assembly. Most assemblies are identified by model, brand and connector, so the part number usually comes back the same working day.

Why does the head cut badly when the lens is clean?

The height sensor is the next suspect. A drifting sensor changes the standoff, and standoff changes the kerf — the same symptom a dirty lens gives, but intermittent and tied to the sheet rather than to the optics.

Is a newer head from the same brand a drop-in replacement?

Not automatically. Five interfaces have to match, and the sensor signal and focus profile have to be supported by your controller. Same brand shortens the list; it does not empty it.

How long can I keep running an out-of-production head?

As long as you can still get the assemblies that fail. The useful test is not the age of the model but the availability of one window, one lens group and one seal kit. When those get hard to buy, plan the replacement — do not wait for the failure.

Do I need a new controller for a new head?

Only if your controller cannot drive the head’s focus axis, read its sensor signal, or hold a focus profile for that model. Plenty of retrofits are a head plus a parameter file. Some are a head plus a control system.

What is cheaper — repair or replace?

It depends on one number: the repair bill against the price of an equivalent new head. Get both quotes before deciding, and be suspicious of anyone who gives you only one.

Raysoar supplies cutting heads, optical assemblies and seal kits. If you are rebuilding a head its maker no longer supports, send us the model — we will tell you which assemblies are still available, what they cost, and how long they take to reach you.