Fiber Loss Budget Calculator

How it works

It adds up the loss in the link, in three parts:

• Fiber: the attenuation for the fiber type at the wavelength you picked, over the length.
• Mated pairs: a set loss for every connector pair.
• Splices: a set loss for every splice.

Loss values sets what each mated pair and splice costs. TIA maximums uses the most loss the standard allows per connection: the worst link that still conforms. Typical field uses good-practice field values, which are lower and aren't the standard's limits.

The wavelengths follow the fiber: multimode offers 850 and 1300 nm, singlemode 1310 and 1550 nm. Change the fiber type and the wavelength goes back to the first one on its list. OS2 is listed twice, inside plant (ISP) and outside plant (OSP), because the two carry different attenuation.

Pick an application and the total is checked against that application's channel insertion loss budget and its reach, the longest run it supports. Pick No target and you get the total on its own.

The calculator opens on 100 m of OM3 multimode at 850 nm, 2 mated pairs and 0 splices on TIA maximums, against 10GBASE-SR (OM3): 1.85 dB of 2.60 dB budget, +0.75 dB of margin, PASS.

The coefficients follow ANSI/TIA-568.3-D and the budgets are IEEE 802.3 channel insertion loss, cited by designation under every result.

Reading the result

The big number is the total loss in dB. With an application picked, it reads against that application's budget, the tick on the bar marks the budget, and the margin (budget minus total) sits under the bar. The lines below show each part of the total and the loss it was charged at.

• PASS: half a dB of margin or more, and the run is within the application's reach.
• THIN: within the budget, by less than half a dB.
• OVER BUDGET: the total is over the budget, or the run is longer than the application's reach. Past the reach it fails whatever the loss says.
• MEASURED: the application is No target. There's nothing to pass or fail against, so you get the total.

Anything that needs attention goes in one of three buckets under the stamp:

• WON'T PASS: the total is over the budget, shown as measured against budget with the margin. Or the run is past the reach: loss can pass and the link still won't, because dispersion is the limit there, not attenuation.
• WON'T LAST: less than half a dB of margin. One dirty endface or one re-termination puts the link over.
• JUST UGLY: four or more mated pairs on TIA maximums, on a link that's a clean PASS or a MEASURED total. That's a lot of connections for one channel, and every one is another chance at a bad endface.

When the total is over the budget, the fix line says how many mated pairs would fit with the fiber and splices you entered, or to shorten the run. It answers the loss budget only, not the reach.

What it doesn't check

• Whether the fiber matches the application. It won't stop you checking an OM3 link against an OS2 application. Pick the application that matches the glass.
• What's actually in the link. It budgets from the length and the counts. A dirty endface or a bad splice shows up on a test set, not here.

Questions

My loss passes. Why does it say OVER BUDGET?

The run is past the application's reach. Reach is its own limit: a link can have loss to spare and still fail at that length. Take the opening link to 301 m on Typical field values. It has +0.95 dB of margin and reads OVER BUDGET, with a WON'T PASS line: 301 m installed / 300 m maximum.

Should I use TIA maximums or Typical field?

TIA maximums charges every mated pair and splice the most loss the standard allows, so the total is the worst link that still conforms. If the link has to pass with any conforming hardware, budget it on TIA maximums. Typical field uses good-practice field values, which are lower and aren't the standard's limits. The opening link reads 1.85 dB on TIA maximums and 0.95 dB on Typical field. The citation under the result says which set it used.

What counts as a mated pair?

Each place two connectors meet in an adapter. A link with a patch panel at each end has two, one at each end, and that's what the calculator opens with. Count every one between the two ends of the link. Splices count the same way: every splice in the link, pigtail splices included.

Why does a link right on the budget read OVER BUDGET?

The calculator judges the sum at full precision before it rounds for display, so a link that lands exactly on the budget can come out a hair over. 400 m of OM4 multimode at 850 nm with 2 mated pairs on TIA maximums, against 10GBASE-SR (OM4), reads 2.90 dB of 2.90 dB budget, a margin of -0.00 dB, and OVER BUDGET. A link on the budget has nothing left for a dirty endface anyway. The same thing can make exactly half a dB of margin read THIN.