Three people ask us the same question about core diameter and mean three different things by it.
The urologist asks whether the fiber will still let him reach a lower-pole stone. The engineer asks whether a bigger core will stop the fiber burning back. The product manager asks which three sizes to stock so his distributor stops returning orders. All three end up in the same sentence — “should we go bigger?” — and all three get the wrong answer if it is given as one number.
That is because core diameter is not a durability specification, a spot-size specification or an access specification. It is all three at once, and which one dominates depends on two things most datasheets never state: the working distance to the tissue, and the wavelength of the laser behind it. Get those two facts straight and the choice usually makes itself.
The available cores on the bare fiber range run from 200 to 1000 µm, and on the holmium fiber range from 272 to 550 µm. That is a factor of nearly four in diameter, or fifteen in area. It is worth knowing exactly what you are buying with that factor before you spend it.
Contact or non-contact decides whether core diameter is even the right knob
Start with the geometry, because it settles most of the arguments. When a fiber touches tissue, the spot is the core, full stop. Move the tip a distance d away and the spot grows as the cone of light opens:
spot ≈ core + 2 · d · tan θ, where θ = arcsin(NA)
With the common NA of 0.22, θ is 12.7° and tan θ is about 0.23 — the beam adds roughly 0.45 mm of spot for every millimetre of standoff. Run that for a two-millimetre working distance, which is a realistic non-contact setting:
| Core (NA 0.22) | Spot at 2 mm | Tissue irradiance at 30 W |
| 200 µm | 1.10 mm | 3.2 kW/cm² |
| 400 µm | 1.30 mm | 2.3 kW/cm² |
| 600 µm | 1.50 mm | 1.7 kW/cm² |
Read that against the same three fibers in contact, where a 200 µm fiber at 30 W delivers about 95 kW/cm² at its face and a 600 µm about 10.6. Tripling the core costs you a factor of nine in power density when the tip is on the tissue, and a factor of only 1.9 when it is two millimetres away. The spot, meanwhile, grows only 36 % over the same change.
The practical reading is blunt: in non-contact work, core diameter is a weak lever on spot size and a diluted lever on irradiance. If a case needs a wider spot at a fixed standoff, changing the core from 200 to 600 µm buys you 0.4 mm and costs you half your power density — a bad trade. What actually widens the cone is the numerical aperture: at the same 2 mm standoff, moving from NA 0.22 to NA 0.37 on a 200 µm core takes the spot from 1.10 mm to 1.79 mm, a far larger effect than any core change on the shelf.
The reverse is also true, and it is the part surgeons notice. On a contact tip the core is the spot, so the core is the only dial you have and every change is paid for in full. That is why a 200 µm fiber at 10 W genuinely does cut more aggressively than a 600 µm at 10 W — but it is not a universal property of small fibers. It is a property of small fibers in contact.
The lower limit on core diameter is optical, not thermal
Here is the counter-intuitive one, and it is worth more to a buyer than any efficiency table.
You cannot simply keep shrinking the core to concentrate power. Ho:YAG energy couples best into cores above 200 µm and ideally above about 240 µm; below that, a meaningful fraction of the beam launches into the cladding rather than the core. The consequences are not subtle. Light in the cladding is light outside the designed path — it scatters, it heats the coating, and it can destroy the fiber from the inside. Published bench work on commercial lithotripsy fibers has found measurable differences in durability within the same size class, with some small fibers markedly more prone to fracture and to damaging the scope’s working channel than others.
So the instinct “smaller core, finer control, less collateral damage” runs straight into a floor. On a 2100 nm Ho:YAG platform the floor is real, and it sits around 240 µm. Below it you are not buying precision, you are buying a fiber that fails, and possibly a scope repair.
Small cores are also not free in the channel. A 272 µm fiber occupies less of a flexible ureteroscope’s working channel than a 365 µm one, which leaves more room for irrigation — and irrigation flow is what keeps the field visible while you are fragmenting. That is the argument nobody puts in a datasheet: the fiber’s diameter is competing with the fluid, not just carrying light.
The upper limit is the anatomy, and it charges interest on bends
If the lower limit is optical, the upper limit is mechanical and anatomical, and it shows up as deflection.
Passing a stiff, large-core fiber down a flexible ureteroscope costs you steering angle. Reported losses in flexible scope deflection run roughly 7–16 % of maximum (about −9° to −19°) with a 200 µm fiber, against roughly 18–37 % (about −24° to −45°) with a 365 µm fiber. Deflection is not a comfort feature. It is how the tip reaches a lower-pole calyx at all. A fiber that fragments a stone beautifully but cannot be pointed at it is not a better fiber.
That is where the published selection guidance lands, and it is refreshingly specific:
| Access | Preferred core | Why |
| Kidney / flexible ureteroscope | 240–272 µm, ball tip | Preserves deflection and the scope’s inner lining |
| Ureter / flexible ureteroscope | 240–272 µm, ball tip | Same, plus room for irrigation |
| Semi-rigid scope, 4.5–6 F | 240–272 µm, flat tip | Channel is the constraint |
| Semi-rigid scope, > 6 F | 365 µm, flat tip | Deflection less critical, fragmentation more efficient |
Two things stand out in that table. First, 365 µm fibers are consistently reported as more durable and more efficient at fragmenting than 200–272 µm — and are still not the default, because the anatomy in front of them vetoes it. Second, the recommended range for flexible work is 240–272 µm, not “as small as possible”. The clinical consensus and the optical floor agree, arrived at from different directions.
Why 2100 nm takes the choice away from you
There is a reason the urology numbers look so settled while other specialties argue endlessly about core size. It is the wavelength.
Ho:YAG at 2100 nm sits on a strong water absorption band, with an absorption coefficient around 30 cm⁻¹ and a tissue penetration depth of roughly 0.4 mm; 95 % of the energy is absorbed within about half a millimetre of water. For comparison, a 1940 nm thulium fiber laser is absorbed even harder, at around 40 cm⁻¹ and roughly 0.2 mm of penetration.
Shallow penetration is not a limitation, it is the safety mechanism — and it also removes an option. A Ho:YAG fiber has to be in contact with its target to do anything; firing it from a distance just heats irrigation fluid. And once contact is mandatory, the earlier arithmetic collapses to its simplest form: the core diameter is the spot size, the spot size is the irradiance, and there is no standoff available to dilute the choice. Urology gets a hard specification because physics removed the free parameter.
Now look at what happens on a 1470 nm diode platform, where penetration is also shallow but standoff is clinically usable — the fiber can be held off the tissue and still do work. Aesthetics, ENT and proctology all live in that space. There, the same core change that costs 9× in contact costs under 2× at a couple of millimetres, so core diameter becomes a secondary decision and the argument shifts to tip geometry and wavelength instead. Same arithmetic, different answer, because one number changed.
This is also why core diameter is close to irrelevant for some products in the range. A diffusing fiber’s treatment length is set by the length of its diffusing segment, not by the core that feeds it — a 400 µm and a 600 µm diffuser fiber with the same 30 mm active length differ in how much power they can carry, not in how much tissue length they cover.
One power rating, three different fibers
There is a last problem with how these fibers are sold, and it is worth raising because it is invisible until something fails.
A product line is usually published with a single maximum power — our holmium fiber range is rated at 120 W. That figure is honest for the largest core in the family and progressively less honest for the smaller ones, because irradiance goes as the inverse square of diameter. At 120 W, the face of a 272 µm fiber is loaded to roughly 206 kW/cm²; the face of a 550 µm fiber to about 50 kW/cm². Against the practical CW end-face working level of about 250 kW/cm², the small fiber is running at 82 % of its damage budget and the large one at 20 %.
The same “120 W laser” is a comfortable pairing with one fiber and a marginal one with another, and the datasheets on both sides can be individually correct while the combination is not. If you are specifying a system rather than buying a spare part, ask for the rating at your core diameter, not the rating for the family. That is the question the bare fiber page is built to answer for the diode range, and it is the one worth asking of any supplier.
Choosing in the right order
The order matters, because three of these constraints are hard and only two are soft.
- Ask the wavelength first. If the platform must work in contact — 2100 nm, and to a lesser degree 1940 nm — core diameter will dominate everything. If the platform works at a standoff, demote core diameter and spend your attention on NA and tip geometry.
- Find the channel. Working-channel diameter, minimum bend radius, and acceptable deflection loss set the ceiling. For flexible scopes, expect to be pushed to 240–272 µm regardless of what the stone needs.
- Find the optical floor. On Ho:YAG, do not go below about 240 µm. Below it you are risking cladding light and fiber failure, not gaining precision.
- Then size for the dose. With the channel and floor known, pick the smallest core that delivers the irradiance your tissue needs at your intended standoff — and remember that at a standoff, the difference between cores is much smaller than the table on the wall suggests.
- Check the power pairing, not the family rating. Compute irradiance at your core and your watts before you approve the combination.
Notice that this is a procedure, not a table, and that is deliberate. A selection matrix without the wavelength and the standoff built into it is a matrix that answers a question nobody asked. The physics does not change between procedures; only the distance to the target does.