[c-nsp] No Link between SFP-10G-LRM and X2-10GB-LX4?

[Mack McBride]

The LRM looks like the way to go for you if you are limited by SFP+ as the Andrew pointed out.
The Cisco spec is for 220M at 500MHz fiber on LRM.
Basically with dispersion you will get shorter error free distance so
although the LX4 is only listed for 500MHz it will work at lower modal BW.
It just shortens the range.  Same is true for the LRM.
It will work at lower modal BW but you get shorter ranges.
You are probably close to the limit with 160M and 200MHz modal BW.
And I would say you are over the 'rated' limit.
For reference the 50 micro with 400MHz modal BW the rated limit is 100M.

Mack

-----Original Message-----
From: Jeroen van Ingen [mailto:jeroen at zijndomein.nl] 
Sent: Monday, October 10, 2011 3:49 PM
To: Mack McBride
Cc: cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] No Link between SFP-10G-LRM and X2-10GB-LX4?

On 10/10/2011 06:00 PM, Mack McBride wrote:
> That older fiber is probably not rated at 500 Mhz Modal BW suggested by Cisco.
> Older fiber is usually 200 Mhz.

True, and I expect the fibers between our dorm to be no better than 200 
MHz*km.

> The LX4 would be a better choice (more dispersion resistant) and max distance of 300M.

We're mainly testing with LRM because I don't like the concept of WDM 
plus serial->parallel->serial conversion in a single transceiver. And 
LRM is a newer standard. No, you're right, not the most technical 
arguments ;)

Oh, if you look at 
http://www.cisco.com/en/US/prod/collateral/modules/ps5455/prod_white_paper0900aecd806b8bcb.html, 
that doc doesn't mention LX4 being supported on MMF with modal bandwidth 
< 500 MHZ*km...

> Your distance limitation is probably going to be much less than the rated value
> if the dispersion characteristics of the fiber don't meet the spec.

Other vendors also spec LRM to work on MMF with 160 or 200 MHz*km modal 
bandwidth...

> Of course you will still get connectivity but you will increment errors.
> For dorms it probably doesn't matter much.  If the streaming music and video is effected
> by errors it will probably save you money since the students won't stream as much.

Their streaming doesn't cost us anything and we're closely monitoring 
the links to check if they remain error-free. We'll only run these links 
if they remain reliable.

We have two links under test right now; a relatively short one 
(somewhere between 30-50 meters) and a longer one, around 160 meters. 
Both perform fine and have been error free for a couple of weeks. The 
shorter without any mode conditioning; the longer with a mode 
conditioning patch cord on one side (it'd flap & run with too many 
errors without the MCP).
More testing to follow, but the LRM gear looks like a good solution for us.

Regards,

Jeroen van Ingen
ICT Service Centre
University of Twente, P.O.Box 217, 7500 AE Enschede, The Netherlands