OM1 and OM2 legacy LC patch cords

When you still need an OM1 or OM2 LC cord

OM1 and OM2 are the older multimode fibre grades. OM1 has a 62.5 micron core and OM2 a 50 micron core, and both were installed widely in buildings and campuses from the 1990s and early 2000s. New installations use OM3, OM4 or single-mode, but the installed fibre stays in the walls. When a patch cord on that fibre breaks, or a switch is replaced with one that has LC ports, the buyer needs a cord that matches the legacy fibre, not a modern one. That is the reason this product exists.

Buy an OM1 or OM2 cord only when it replaces a link on installed OM1 or OM2 cabling, and keep the fibre type of the cord the same as the fibre type of the installed cable. A mismatch costs light, as the worked example later in this page shows. For new links, the choice is covered in OM3 vs OM4 vs OM5 and OM4 vs OS2: which to buy.

Radix Fibrotech makes fibre patch cords and assemblies to the customer's drawing in Pune. Legacy-fibre builds are specified by drawing, so confirm in the quote what is available for the fibre, connector and length you need. Send us the drawing and we reply with what we can build. The wider LC range is described under LC to LC patch cords.

OM1 and OM2 side by side

The bandwidth and attenuation values come from the ISO/IEC 11801 and TIA-568 cabling grades, and the numbers in a particular cable's data sheet may be better than the minimum. Both grades share the orange jacket, so colour does not tell them apart. The core size is printed on the jacket legend, and for a bare cord the only reliable check is the label or a core measurement. See multimode fibre colours for the colour scheme.

The LED-era design is why these grades have low modal bandwidth. An LED fills many modes of the fibre, so the pulse spreads as it travels, and bandwidth drops quickly with length. Vertical-cavity lasers (VCSELs) launch fewer modes into the fibre, and the newer grades OM3 to OM5 were optimised for them. The older grades work with VCSELs, but their reach is short at higher speeds. See chromatic and modal dispersion for the physics.

Reach on legacy fibre: what the standards give you

Reach depends on speed and on the transceiver. The IEEE 802.3 optical interface types fix the distance for each fibre grade. The figures below are the usual published values, and the transceiver data sheet governs in practice.

1000BASE-SX on OM1 is stated as 275 m for fibre with 200 MHz.km and 220 m for fibre with 160 MHz.km, so older installs on lower-bandwidth 62.5 micron fibre can be limited to 220 m. 1000BASE-LX on multimode fibre requires a mode-conditioning patch cord at the transmitter when used on 62.5 micron fibre, and that cord is a special part, not an ordinary one. The Fibre Channel standards have similar tables, summarised in Fibre Channel speeds and reach, with 1, 2, 4 and 8 Gbit/s on legacy fibre reaching roughly 300 m, 150 m, 70 m and 21 m on OM1 and 500 m, 300 m, 150 m and 50 m on OM2, per the published tables.

The common thread is that OM1 and OM2 cords suit one gigabit links and low-speed Fibre Channel. They are poor choices for 10G and impossible for 25G and above except over a few metres. For the full list of Ethernet reach, see the IEEE 802.3 fibre Ethernet reach table.

The 62.5 to 50 micron mismatch

The commonest mistake with legacy fibre is mixing the two core sizes. Light going from a larger core to a smaller one loses power, because part of it falls outside the smaller core and outside the smaller acceptance angle. Going the other way, from small to large, loses little.

The loss can be estimated as the sum of two effects, expressed in dB as 10 times the log of a ratio.

• Core area effect: the ratio of areas is (50 / 62.5) squared = 0.64, which is 10 x log10(0.64) = about -1.9 dB.
• Numerical aperture effect: the ratio is (0.200 / 0.275) squared = 0.529, which is 10 x log10(0.529) = about -2.8 dB.
• Both together: 0.64 x 0.529 = 0.339, which is about -4.7 dB.

The 4.7 dB figure is the worst case, for a source that fills the whole core and the whole numerical aperture. Real loss depends on the mode fill of the source and is often lower. It is large enough to bring a working link down, and a typical gigabit transceiver loses a few dB of its power budget to it. A 62.5 micron cord on a 50 micron cable can be a hidden cause of a link that is marginal or intermittent.

Because the loss is one-way, a mismatched cord can work in one direction and fail in the other, which makes the fault hard to trace. Match the cord to the cable at every position.

Worked example: replacing cords on an old floor

All numbers are assumed. A small office has gigabit links on installed OM1 cabling from a wiring closet to each room. The switch is being replaced with one that has SFP ports taking 1000BASE-SX modules. The installer needs LC patch cords at the switch and at the wall panel for each run.

Now the loss check. Assume 3.5 dB/km at 850 nm, which is the maximum for these grades, and four mated pairs in the channel (switch, panel on each side, and wall outlet) at 0.5 dB each.

• Fibre: 0.196 km x 3.5 dB/km = 0.69 dB
• Connectors: 4 x 0.5 dB = 2.0 dB
• Total: about 2.7 dB

The IEEE 802.3 channel insertion loss figure for 1000BASE-SX on 160 MHz.km, 62.5 micron fibre at 220 m is about 2.4 dB, and about 2.6 dB for 200 MHz.km fibre at 275 m. Check the current text of the standard. The 2.7 dB estimate is over both, and the margin disappears. The fix is better connectors (0.3 dB each gives 1.2 dB and a total near 1.9 dB, inside both figures), fewer mated pairs, or direct cords. This shows why connector loss on legacy links needs attention even over short distances, and why an installer who buys the cheapest cords may find a link that does not come up. The method is in the fibre loss budget guide.

Had the installer ordered OM2 cords for this OM1 cable, the transmit side could lose up to 4.7 dB in the cord-to-cable step, and the link would fail.

Common mistakes with legacy multimode

• Ordering by colour. OM1 and OM2 are both orange. The core size must be on the order.
• Mixing 62.5 and 50 micron in one channel. One cord of the wrong type gives a one-way loss that may be hard to find.
• Expecting 10G over OM1 at 100 m. The standard gives 33 m on OM1.
• Using single-mode SFPs on multimode fibre, or the reverse. Match the module label to the fibre.
• Using a mode-conditioning cord by accident or leaving it out. Gigabit LX on legacy multimode needs one, and it is not an ordinary cord.
• Forgetting dust caps and cleaning. Old patch panels collect dust, and a dirty connector loses more than the fibre does. See how to store and handle patch cords.
• Assuming new cords make an old link faster. The installed cable sets the limit.

When you should not buy an OM1 or OM2 cord

Do not buy one for a new link, for any link at 10G or above beyond a few tens of metres, or for a data-centre refit. The cost saving is small and the cabling will not last. Choose OM3 LC duplex, OM4 LC duplex or OS2 LC duplex cords, depending on speed and reach.

If you are planning to upgrade a site, installed OM1 and OM2 cable is usually the first thing to replace, since the cords are the cheaper part of the change. A short survey of lengths and speeds will show whether the existing fibre can carry the new switch. The installed-cable decision is covered in multimode fibre optic cable.

Request a quote

Send the fibre type, connector at each end, length, cable diameter and jacket, and quantity through the request a quote page, along with a photo of the installed cable legend if you have one. Radix makes fibre patch cords to the customer's drawing in Pune, with quotes by form, phone or WhatsApp. Legacy builds are confirmed one by one, so tell us what you need and we reply with what we can build.