Optical Attenuation

Unanswered Question
Jun 14th, 2010
User Badges:


I've got a quick question that I really cannot find much information about.  I'm currently moving into a facility that's going to be using single-mode fiber to the desktop, and folks have voiced concerns that there could potentially be problems between the SFP's, and the end devices if there was no optical attenuation. 

So far what i've found online was pretty sparse, not many folks are using single-mode fiber to the desktop, but from what I can find, it seems like it's a valid concern that needs to be addressed.  What types of problems could occur?  Burnout of recievers at the end devices, or the SFP's?

Any ideas, or explinations are greatly apprecated.


  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (2 ratings)
Giuseppe Larosa Mon, 06/14/2010 - 11:45
User Badges:
  • Super Silver, 17500 points or more
  • Hall of Fame,

    Founding Member

Hello Jkloza,

the SFPs are not likely to be damaged.

Actually attenuation may be needed to avoid to blind the  receivers:

if detection of 0 or 1 bit  is based on power level on a time interval if the power is too high there is no difference between high level and low level (saturation).

So bit error rate can become high if the signal is too strong.

Said this, to use or not use a 5 dB attenuator depends on transmitter power, receivier sensibility and fiber length so it has to be seen from case to case.

However, it is really uncommon to have single mode fiber to the desktop I guess there are security concerns in this environment and the use of fibers has been chosen to avoid interception.

Hope to help


viyuan700 Mon, 06/14/2010 - 14:31
User Badges:
  • Silver, 250 points or more

What types of problems could occur?  Burnout of recievers at the end devices, or the SFP's

Since you will be using short distance SFP and most of the short distance SFP dont have problem with no attenauation (or more power). You dont need an attaenuator for short distance. See the link below )

For fast ethernet


For gigabit ethernet


The max  Trans optical power is equal to the Max Receive sensitivity and even if a just take one connector (most likely you will have more when it is going till desktop) then you can get 1dB loss so for neglible distance also you there is no possibilty of Burnout.

Burnout is a problem with long haul SFP where Max Trans power is much more than the receiver sensitvity.

Problem can be with bending loss can be more loss or fiber can break. Not as flexible as copper. Since your disatnce is too less so bending los will be taken care by the more power you have.

I dont see any other issue ONLY if the fiber is broken then you have no link.

Jkloza_2 Mon, 06/14/2010 - 18:35
User Badges:

When you say i'll be using a "short distance SFP", which are you referring to?  Any of the gigabit ethernet SFP's?  As far as I know we'll be using the LX/LH SFP's, which, in my mind aren't short distance SFP's.

Thanks -


viyuan700 Mon, 06/14/2010 - 23:12
User Badges:
  • Silver, 250 points or more

                                             TX                    Rx

1000BASE-LX/LH                -3 to -9.5       -3 to -20

This is the spec of Gigabit module (LX/LH )you are considering (you can check the link in my earlier message) and you can see

Max Trans is equal to Max Receive.For this module even if you just loop the Trans to receive of same module nothing will happen. As rarely you will get -3dBm from the Tx. that value will be somewhere -5-6 dBm typical.

Sometime back a guy asked the range is too much for latest technologies but we are not latest in terms of Optics. There are so many problems while maintaining the power which left from the Transmitter till the power you get from the cable connecetd to Tx.

may be 10Km is a big disance for copper or Multimode but for SMF this distance is quite small. EX & ZX are long haul modules where you talk about 40-70Km. 


This Discussion