FOA Guide

Power Budget And Loss Budget

The terms "power budget" and "loss budget" are often confused.

The power budget refers to the amount of loss that a datalink (transmitter to receiver) can tolerate in order to operate properly. Sometimes the power budget has both a minimum and maximum value, which means it needs at least a minimum value of loss so that it does not overload the receiver and a maximum value of loss to ensure the receiver has sufficient signal to operate properly.

The loss budget is the amount of loss that a cable plant should have. It is calculated by adding the average losses of all the components used in the cable plant to get the total estimated end-to-end loss. The loss budget has two uses, 1) during the design stage to ensure the cabling being designed will work with the links intended to be used over it and 2) after installation, comparing the calculated loss to test results to ensure the cable plant is installed properly.

Some standards refer to the loss budget as the "attenuation allowance" but there seems to be very limited use of that term.

Obviously, the power budget and loss budget are related. A data link will only operate if the cable plant loss is within the power budget of the link.

Remember the calculated loss budget is an estimate that assumes the values of component losses and does not take into account the uncertainty of the measurement. Be aware of this because if measurements are close to the loss budget estimates, some judgement is needed to not fail good fibers and pass bad ones!

Power Budget

All datalinks are limited by the power budget of the link. The power budget is the difference between the output power of the transmitter and the input power requirements of the receiver, both of which are defined as power coupled into or out of optical fiber of a type specified by the link. The receiver has an operating range determined by the signal-to-noise ratio (S/N) in the receiver. The S/N ratio is generally quoted for analog links while the bit-error-rate (BER) is used for digital links. BER is practically an inverse function of S/N. Transceivers may also be affected by the distortion of the transmitted signal as it goes down the fiber, a big problem with multimode links at high speeds or very long OSP singlemode links.

fiber optic power budget
When testing a fiber in a cable plant to determine if the cable plant will allow a specific link to operate over it, the test should be made from transceiver to transceiver, e.g. the cable plant with patchcords installed on either end that would be used to connect the transceivers to the cable plant. When doing a link loss budget (below) for the cabling to be used with a given link to determine if the link will operate over that link, the loss of the patchcords should also be included.

Cable Plant Link Loss Budget Analysis

Loss budget analysis is the calculation of a fiber optic cabling system's estimated loss performance characteristics. This is sometimes confused with the communication system "power budget" which is a specification of the dynamic range of the electronics, the difference between the output power of the transmitter coupled into the fiber and the minimum received power required at the receiver for proper data transmission. The communications system power budget will set a limit for the loss of the cable plant.

The cable plant loss budget needs to consider transceiver wavelength, fiber type, and link length plus the losses incurred in splices, connections and other passive devices like FTTH or OLAN PON splitters. Attenuation and bandwidth/dispersion are the key parameters for the cable plant loss budget analysis.

FOA Loss Budget App
FOA has a free app for smartphones and tablets that will calculate loss budgets for the cable plant you are designing or testing. See the app store for your device for details.

Analyze Link Loss In The Design Stage

Prior to designing or installing a fiber optic cabling system, a loss budget analysis is recommended to make certain the system will work over the proposed link. That same loss budget will be used as to compare test results after installation of the cabling to ensure that the components were installed correctly. Both the passive and active components of the circuit have to be included in the loss budget calculation. Passive loss is made up of fiber loss, connector loss, and splice loss. Don't forget any couplers or splitters in the link. Active components are system gain, wavelength, transmitter power, receiver sensitivity, and dynamic range. Prior to system turn up, test the circuit with a source and FO power meter to ensure that it is within the loss budget.

The idea of a loss budget is to insure the network equipment will work over the installed fiber optic link. It is normal to be conservative over the specifications! Don't use the best possible specs for fiber attenuation or connector loss - give yourself some margin!

The best way to illustrate calculating a loss budget is to show how it's done for a typical 2 km multimode link with 5 connections (2 connectors at each end and 3 connections at patch panels in the link) and one splice in the middle. See the drawings below of the link layout and the instantaneous power in the link at any point along it's length, scaled exactly to the link drawing above it.


This concept gets many questions - but two are most common. Why do you include the loss of the connectors on the ends if they are connected to a transmitter and receiver. And what about testing a permanently installed cable plant from patch-panel (or wall outlet) to another patch panel, not including the final patchcords used to connect equipment.

Why do you include the connectors on each end? Depending on the design of the transceivers (and especially if they have pigtailed lasers or detectors), practically every factor in connector loss affects coupling to a transmitter or receiver as well. Whether these connections are included in the loss budget should depend on whether the margin for the link to be use on the cable plant was specified to include these connectors. A
s far as we know almost all system specifications are considering connection losses at both ends. Unless you know the system was not specified for loss including the end connectors, include them in calculations of the loss budget.

Testing is another issue.
When the cable plant is tested, the reference cables will mate with those end connectors and their loss will be included in the measurements but the results depends on the method used to set the "0dB" reference.

If the "0dB" reference for the insertion loss test was done with only one reference test cable attached between the light source and power meter which is the most common way, the connectors on the end of the cable will be included in the loss so the loss budget should include both connectors.

Most tests are specified and done with the one cable reference when the test equipment is compatible with the connectors.

If the "0dB" reference for the insertion loss test was done with three cables, the launch reference cable, a receive reference cable and a third reference cable between them, a method used for many plug and jack (male/female) connectors such as MPOs, the loss budget should not incude the connectors on the end. When making the "0dB" reference with three cables, two connections are included in setting the reference so the measured value will be reduced by the value of those two connections. If the loss budget is calculated without the connectors on the ends, the value will more closely approximate the test results with a 3-cable reference. The three cable reference is generally done with plug/jack or male/female connectors like the MPO or when doing a "channel" test specified in some standards that includes the permanently installed cable plant with patchcords attached but excludes the connectors on each end that attach to transceivers.

While the two-cable reference method is rarely used, it includes only one connector. Thus you could use the same approach when calculating loss budgets for this test method.

Whatever test method is presumed, it must be documented when the loss budget is calculated.

Cable Plant Passive Component Loss

Step 1. Fiber loss at the operating wavelength

Cable Length 2.0 2.0

Fiber Type Multimode
Wavelength (nm) 850 1300 1310 1550
Fiber Atten. dB/km 3 [3.5] 1 [1.5] 0.4 [1/0.5] 0.3 [1/0.5]
Total Fiber Loss 6.0 [7.0] 2.0 [3.0]

(All specs in brackets are maximum values per EIA/TIA 568 standard. For singlemode fiber, a higher loss is allowed for premises applications. )


Step 2. Connector Loss

Multimode connectors will have losses of 0.2-0.5 dB typically (see note about "connector" vs. "connection" loss). Singlemode connectors, which are factory made and fusion spliced on will have losses of 0.1-0.2 dB. Field terminated singlemode connectors may have losses as high as 0.5-1.0 dB. Let's calculate it at both typical and worst case values.
Remember that we include all the components in the complete link, including the connectors on each end.

Connector Loss 0.3 dB (typical adhesive/polish conn)  0.75 dB (TIA-568 max acceptable)
Total # of Connectors 5  5
Total Connector Loss 1.5 dB  3.75 dB

Note: When we say connector loss, we really mean "connection" loss - the loss of a mated pair of connectors, expressed in "dB." Thus, testing connectors requires mating them to reference connectors which must be high quality connectors themselves to not adversely affect the measured loss when mated to an unknown connector. This is an important point often not fully explained.  In order to measure the loss of the connectors you must mate them to a similar, known good, connector. When a connector being tested is mated to several different connectors, it may have different losses, because those losses are dependent on the reference connector it is mated to.

(All connectors are allowed 0.75 max per EIA/TIA 568 standard)

 Remember that we include all the components in the complete link, including the connectors on each end. In our example above, the link includes patchcords on each end to connect to the electronics. We need to assess the quality of these connectors, so we include them in the link loss budget and if we test the link end to end, including the patchcords, these connectors will be included in the test results when connected to launch and receive reference cables. On some links, only the permanently installed link, not including the patchcords, will be tested. Again, we still need to include the connectors on the end as they will be included when we test insertion loss with reference test cables on each end.

Step 3. Splice Loss

Multimode splices are usually made with mechanical splices, although some fusion splicing is used. The larger core and multiple layers make fusion splicing abut the same loss as mechanical splicing, but fusion is more reliable in adverse environments. Figure 0.1-0.5 dB for multimode splices, 0.3 being a good average for an experienced installer. Fusion splicing of singlemode fiber will typically have less than 0.05 dB (that's right, less than a tenth of a dB!)

Typical Splice Loss 0.3 dB
Total # splices 1
Total Splice Loss 0.3 dB

(All splices are allowed 0.3 max per EIA/TIA 568 standard)


Step 4. Total Passive System Attenuation

Add the fiber loss, connector and splice losses to get the link loss.

Typical  TIA 568 Max
   850 nm  1300 nm  850 nm 1300 nm
Total Fiber Loss (dB) 6.0 2.0  7.0 3.0
Total Connector Loss (dB) 1.5 1.5  3.75  3.75
Total Splice Loss (dB) 0.3 0.3  0.3  0.3
Other (dB) 0 0  0 0
Total Link Loss (dB) 7.8 3.8  11.05  7.05


Remember these should be the criteria for testing. Allow +/- 0.2 -0.5 dB for measurement uncertainty and that becomes your pass/fail criterion.


Equipment Link Power Budget Calculation: Link loss budget for network hardware depends on the dynamic range of the electronics, the difference between the sensitivity of the receiver and the output of the transmitter into the fiber. You need some margin for system degradation over time or environment, so subtract that margin (as much as 3dB) to get the loss budget for the link.

Step 5. Data From Manufacturer's Specification for Active Components (Typical 100 Mb/s link)

Operating Wavelength (nm) 1300
Fiber Type MM
Receiver Sens. (dBm@ required BER) -31
Average Transmitter Output (dBm) -16
Dynamic Range (dB) 15
Recommended Excess Margin (dB) 3


Step 6. Power Margin Calculation

Dynamic Range (dB) (above) 15  15
Cable Plant Link Loss (dB) 3.8 (Typ)  7.05 (TIA)
Link Loss Margin (dB) 11.2  7.95

As a general rule, the Link Loss Margin should be greater than approximately 3 dB to allow for link degradation over time. LEDs in the transmitter may age and lose power, connectors or splices may degrade or connectors may get dirty if opened for rerouting or testing. If cables are accidentally cut, excess margin will be needed to accommodate splices for restoration. The 3dB rule, of course, is irrelevant if the power budget is ~2dB like some of the 10G multimode links. Then the need for the best quality installation is critical!


NOTE: Many techs forget when doing a loss budget that the connectors on the end of the cable plant must be included in the loss budget. When the cable plant is tested, the reference cables will mate with those connectors and their loss will be included in the measurements.

Related Topics:

Guidelines On What Loss To Expect When Testing Fiber Optic Cables For Insertion Loss With A Meter and Source or OLTS

Table of the cable plant length and loss margins for most LANs and Links

More detailed information can be found on the FOA Online Reference Guide.

Return To The FOA Home Page

Return To FOA Tech Topics

(C) 2004-13 The Fiber Optic Association, Inc.