Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: 10GBase-LR/LW Supporting 1550 nm




Hi Ali, 

The numbers are indeed quite close.  But there's still a difference of 0.5
to 1 dB.  For a 1550 Tx talking to a 1300 Rx, in-line attenuators would
solve any overload problem.  But what happens when you have a 1300 Tx
talking to a 1550 Rx?  I assume BER suffers - this needs to be quantified.  

Regards, 
Jugnu 

-----Original Message-----
From: Ali Ghiasi [mailto:aghiasi@broadcom.com]
Sent: Thursday, April 12, 2001 10:10 AM
To: Jugnu Ojha
Cc: 'Ali Ghiasi'; Serial
Subject: Re: 10GBase-LR/LW Supporting 1550 nm


Hi Jugnu

I looked at three manufactures of InGaAs Photodiode which are 
suitable for 10 Gig Operation and the responsitivity are as the 
following for 1300nm and 1550 nm:

	1300 nm		1550 nm
mfg 1	0.9		0.8	A/W
mfg 2	0.8		0.85	A/W
mfg 3	0.95		1.0	A/W

Responsitivity between 1300 nm and 1550 is nearly identical.
Any overload need to be addressed similar to 1550 nm 
case by adding in line attenuator, unless one build a 
lower power version for SR.

Thanks,

Ali

Jugnu Ojha wrote:
> 
> Ali,
> 
> I think interop between 1310 and 1550 would be extremely useful.  Does
> anybody have any precise numbers for responsivity in the two ranges?  It
> would be advisable to quantify this and determine any impact on overload,
> BER, etc.
> 
> Regards,
> Jugnu Ojha
> Caspian Networks
> (408)382-5213
> 
> -----Original Message-----
> From: Ali Ghiasi [mailto:aghiasi@earthlink.net]
> Sent: Wednesday, April 11, 2001 7:01 PM
> To: Serial; stds-802-3-hssg-serialpmd@ieee.org
> Cc: Ali Ghiasi
> Subject: 10GBase-LR/LW Supporting 1550 nm
> 
> Hi
> 
> Last week during the serial PMD call I suggested possible need for
> Interoperability between 1550 and 1310 nm.  I was asked to write a
> short description.
> 
> MY intention was if you have a 1550 nm source assuming you met overload
> for the 1300 nm receiver you are allowed to operate with 10GBase-LR/LW
> over 10 Km.  If we look at Table 52-13 the wavelength range currently
> specified from 1265-1355 nm.  Without adding any burden we can
> define a second window around 1550 nm to allow use of 1550 nm lasers
> with 10GBase-LR/LW.  InGaAs/InP detector sensitivity is nearly flat
> in this region and it may only require to specify dual wavelength AR
> coating, which is already very common.
> 
> Thanks,
> 
> Ali Ghiasi << File: Card for Ali Ghiasi >>

-- 
Ali Ghiasi
Broadcom Corp.