DEV Community

Dheeraj Ramasahayam
Dheeraj Ramasahayam

Posted on

Fiber end-face inspection: what the scope shows before you clean

Fiber end-face inspection: what the scope shows before you clean

Most fiber faults start at a face nobody looked at. The connector gets mated, the link comes up or it doesn't, and the troubleshooting starts at the optic, the patch, the port — everywhere except the two square millimetres that actually carry the light. A scope takes ten seconds. It tells you whether the face is clean, what is on it if it isn't, and whether cleaning will fix it at all. The habit that saves the most time on the floor is not cleaning more — it is scoping first, every time, before anything gets mated or cleaned.

Inspect-clean-inspect, and why the order matters

The rule is inspect, clean, inspect. Scope the face first, clean once based on what you saw, then scope again and only mate the connector when the second scope passes. The order matters because cleaning blind can make a dirty face worse: a dry wipe drags a grit particle across the core and turns a dust speck into a scratch, and a wet wipe on a face with a pit leaves solvent residue baked into the defect. Scoping first tells you what you are dealing with, so you pick the right clean instead of guessing — dry clean for loose dust, wet-dry for oil and film, and no clean at all for a scratched core, which is a replace, not a clean. If you only remember one thing from this piece, make it this: never mate a connector you have not scoped, and never clean a face you have not looked at.

The four zones, and why the zone decides the call

The IEC 61300-3-35 standard divides the end-face into four concentric zones, and a defect that fails in one zone can pass in another — the zone decides the call, not the size alone.

  • Zone A — the core (0–25 microns single-mode, 0–65 microns multimode). The light path. The limit here is zero: no scratches, no pits, no dust, no oil. Any defect in the core is a fail.
  • Zone B — the cladding (25–120 microns single-mode, 65–120 microns multimode). Glass around the core. Scratches over 3 microns or pits over 2 microns fail; smaller ones pass.
  • Zone C — the adhesive ring (120–130 microns). Scratches over 5 microns or pits over 5 microns fail.
  • Zone D — the contact zone and ferrule face (130–250 microns). No light travels here, so the limits are loosest: scratches over 10 microns or pits over 10 microns fail.

Zone A is the whole game. A scope with pass/fail grading does this math for you, but you should know the zones anyway: when a scope flags a "fail" you need to know whether the defect is in the core (replace the connector) or out on the ferrule edge (clean it and move on). The practical test I use: if the defect touches the bright centre circle, it is a Zone A problem. If it is out in the rings, check the size against the table.

Reading the picture: what each defect looks like

Dust shows as sharp dark specks or short fibres lying across the face, sometimes with a hard shadow. Loose dust cleans off with one dry pass — it is the most common thing you will see and the easiest fix.

Oil, usually from a fingertip or a contaminated cap, shows as a faint halo or smear that softens the whole image, like looking through a greasy lens. Oil needs a wet-dry clean: solvent first, then a dry pass, because a dry wipe alone just smears it wider.

A scratch is a straight line, often bright against the dark field, crossing the core or cladding. A scratch in Zone A is a dead connector — clean once to confirm it is really a scratch and not a fibre of dust, then replace it.

A pit is a small dark crater, usually round, where glass has chipped out. A pit in the core fails; a pit in Zone D usually passes.

A chip is a missing bite out of the ferrule edge, and a crack is a line running from the edge inward. Either one means the connector is done — chips and cracks spread under mating pressure, and a cracked ferrule will damage the mating face too.

The mistake to avoid is calling everything "dirty" and cleaning it the same way. Dust, oil, scratches, pits, chips and cracks are six different problems with three different answers: clean it (dust, oil), clean it once to confirm then replace it (scratch or pit in the core), or replace it outright (chip, crack). The scope image is what tells you which answer applies, and it takes ten seconds to get.

The defects that fool the light test

A dirty face can pass a quick light-level check and fail weeks later, and that is the expensive kind of fault — the one that looks fixed. Dust on the core adds insertion loss, but a small amount may still leave the received power inside the acceptable range, so the link comes up and the ticket closes. The contamination is still there, though, reflecting light back into the laser and raising return loss, and it shifts with temperature and vibration. The link flaps under load or after a temperature swing, and the next tech on the ticket starts over from the optic. That is why the scope image matters more than a single power reading: the reading says the link works now, the scope says whether it will keep working. If you inherit a flapping link and the light levels look fine, scope both ends before you swap anything — the fault is usually sitting on a face, waiting to be seen.

The pass/fail call, in order

Scope the face. First look at Zone A, the core: any defect there — dust, oil film, scratch, pit — is a fail, clean it and re-scope. If the core is clean, check Zone B: scratches over 3 microns or pits over 2 microns fail, smaller ones pass. Then Zones C and D: a scratch under 5 microns in C or under 10 in D passes, and a chip only matters if it reaches the contact zone. Clean once — dry for dust, wet-dry for oil — and scope again. If the second scope passes, mate the connector and record the reading. If the second scope still fails in Zone A, stop cleaning: a third clean will not fix a scratch, it will only polish the damage. Replace the connector or the patch cord, scope the new one, and mate that. The whole sequence takes under a minute, and it is the difference between a link that holds and a ticket that comes back.

Recording the inspection in the ticket

The scope result belongs in the ticket, because the next person on the link needs to know the faces were checked and what was found. The minimum useful line:

Panel A port 12 ↔ Panel B port 12,
end-face scoped pre-mate: Zone A clean both ends,
Zone B scratch 2 µm on A (pass), oil film on B —
wet-dry clean, re-scope pass.
Rx -3.1 dBm vs -6 dBm budget — pass.
Enter fullscreen mode Exit fullscreen mode

The shape to copy: endpoints, what the scope showed per zone, what clean was used, the re-scope result, and the received power against the budget. Five fields, and the next tech knows the faces were inspected rather than assumed clean, what defect was found, and that the link was verified after the fix — which is exactly the information that stops the same ticket from being opened twice.

A question for you

What is the one defect you see most often on the scope — dust, oil, or scratches — and has a "clean" face ever fooled you into closing a ticket that came back?

Top comments (0)