Low output from an ultrafast system says little about which part failed, and a closed enclosure says nothing about whether it can be repaired. What helps is identifying the installed architecture.
PhaseX Laser Services · 2026-09-28
There is a belief in a lot of laboratories that a sealed ultrafast laser is a closed book: if it is one box and the manufacturer has moved on, the instrument is finished. It is worth saying plainly that this is not true. A closed enclosure is a service access question, not a verdict on whether an instrument can be repaired, and the instruments behind that assumption are frequently the ones most worth recovering.
What a sealed enclosure does change is how the work is scoped, and how much the identity of the installed system matters before anyone quotes anything.
Modular and integrated are descriptions of packaging, not of physics. An open installation keeps the oscillator, the pump and the amplifier as separately identified products, each with its own label and its own failure modes. An integrated platform brings those functions into one product: Coherent’s Astrella is a one-box Ti:Sapphire amplifier that includes an integrated oscillator and pump laser inside a single footprint.
So an integrated box can be a Ti:Sapphire system, and an open bench can be a fiber or ytterbium system. Assuming either way from the outside of a housing is how a job starts with the wrong scope. The model and configuration carry far more information than the words "open" or "sealed".
Reduced power, unstable behaviour, an operating state the system will no longer reach, or a fault message: all of these can arise on any of these arrangements. They belong in a service request, and none of them identifies the oscillator, pump, amplifier or conversion stage as the culprit.
One distinction is worth making because it is free. A normal front-panel display and a usable result at the experiment are different observations. If the display looks right and the work no longer behaves, say exactly that. If the system includes a frequency-conversion stage, that stage and the chain feeding it are separate parts of the installation and both sit inside the service scope; the harmonic generation page sets out what that scope covers.
Three things, none of them mysterious. Whether the configuration can be established from the labels, which on long-lived and relocated instruments is not a given. Whether the subsystems inside the enclosure can be reached and characterised as separate items. And whether the parts that have failed can be repaired, substituted with a current equivalent, or engineered to perform the original function inside the original envelope.
That third route is what keeps unsupported instruments working. Manufacturer support status and the outcome of an independent repair are two separate questions: the first describes what the original supplier offers today, and the second depends on the configuration, the accessible subsystems and whether a failed part can be repaired, substituted or engineered. We answer the second one, in writing, for the instrument in front of us.
Give the model and serial number of the main laser, then list the controller, power supply and any separately identified pump, amplifier or conversion equipment that is part of normal operation. Configuration sheets and previous service reports are worth more than they look, particularly on legacy platforms.
Say how the system is used, because a shared microscopy facility, a spectroscopy bench and a materials-processing installation have different definitions of working. Mention a recent relocation or a changed connected instrument as history, without implying that it caused anything.
One practical warning before anything ships: confirm which components and interconnections we need. A head arriving without the controller or the cabling its configuration depends on is an incomplete instrument on the receiving bench, and the evaluation waits until the missing items arrive.
A defined evaluation scope before you ship, then written findings that separate the condition as received, the work agreed, and the condition documented after service. Where original acceptance data exists it stays with the instrument records; where it does not, we say so rather than substituting an assumption drawn from a model name.
That applies to current platforms and to instruments whose manufacturer stopped supporting them years ago.
Five fields to start, and no obligation either way. We reply with an assessment and, where work makes sense, a written scope and a fixed fee.