Enter for the full results page · Esc to close
Testing, Coding & Operations
A passive plug that returns each transmit lane to its paired receive lane, so a port can be tested without a link partner.
A loopback is a plug that returns each transmit lane to its paired receive lane, so a port transmits to itself. Insert one, run a bit-error test, and you have exercised the switch SerDes and the module's transmit and receive chains with nothing else in the path.
The value is subtraction. An optical link has three plausible culprits — the host, the module and the plant — and a loopback removes two of them in about ten seconds. A port that passes clean under loopback and fails on the fibre has told you where to walk. It is a passive part with no electronics and no power, which is why it belongs permanently in a rack toolkit rather than in a stores cupboard.
A duplex loopback is trivial: transmit to receive. A multi-fibre one is not, because the mapping must match the application. For eight-fibre optics in an MPO-12 — 100GBASE-SR4, 400GBASE-DR4 and their relatives — transmit occupies positions 1 to 4 and receive positions 9 to 12, so the loopback connects 1↔12, 2↔11, 3↔10 and 4↔9. Positions 5 to 8 stay dark.
A sixteen-fibre application in an MPO-16 uses every position and folds about the centre: 1↔16 through 8↔9. Fitting the wrong one produces a port that comes up on some lanes and not others — among the most time-wasting symptoms there is. See MPO/MTP and MPO polarity for the convention.
A long-reach module looped back through zero loss delivers its full launch power to its own receiver, and receivers have a maximum as well as a minimum. A 10GBASE-ER transmitter is specified to launch as high as +4 dBm while its own receiver is rated to accept no more than about −1 dBm; an 80 km ZR module is worse, launching 0 to +5 dBm into a receiver that saturates near −7 dBm.
The result is errors that look exactly like a fault and are an artefact of the test. Fit an attenuated loopback — our MPO loopback modules are stocked at 0, 5, 10 and 15 dB, and flange attenuators cover the duplex case. Short-reach multimode optics almost never need one.
Modern CMIS modules can do much of this in firmware. Page 13h advertises host-side and media-side loopback capability, in both directions, along with pattern generation and checking. A host-side loopback proves the electrical path to the switch; a media-side one proves the optical front end. Neither needs anyone in the rack.
The physical plug still wins for burn-in, for modules whose firmware does not advertise the capability, and for proving a patch panel port rather than a transceiver.
Manufacturing test, switch burn-in before a cluster carries traffic, and the opening move of an optical investigation once DDM has been read. It is also the honest way to answer a supplier who blames your fibre — the link-down diagnostic guide places it in sequence with the other checks. Keep one per form factor and attenuation value in the spares kit.
| Types | LC duplex, MPO-12, MPO-16, and electrical DAC-style plugs |
|---|---|
| MPO-12 for 8-fibre optics | 1↔12, 2↔11, 3↔10, 4↔9 — positions 5–8 stay dark |
| MPO-16 for 16-fibre optics | 1↔16, 2↔15 and so on through 8↔9 |
| SR, SR4, SR8 multimode | 0 dB — loop straight through |
| LR, LR4, FR4, DR4 | 0 to 5 dB, depending on launch power |
| ER and ER4 | 5 to 10 dB — 10GBASE-ER launches to +4 dBm against a −1 dBm receiver limit |
| ZR, 80 km direct detect | 15 dB — 0 to +5 dBm launch against a receiver saturating near −7 dBm |
| Software alternative | CMIS Page 13h host-side and media-side loopbacks, with pattern generation |
| What it cannot test | the fibre plant, the far-end module, or end-to-end polarity |
DDM, BER, insertion loss, MPO/MTP, CMIS.
Send the specification and an engineer answers it — no obligation, and no sales sequence. If the answer is that you do not need what we sell, that is the answer you get.
Usually answered within one working day.