[j-nsp] 1000B-LH PIC lamda 1550 works with RS gige card at 1310 nm

Richard A Steenbergen ras at e-gerbil.net
Sat Nov 27 01:15:59 EST 2004

On Fri, Nov 26, 2004 at 11:30:26PM -0500, Matt Yaklin wrote:
> Hi,
> I have a juniper with a gige pic. (LH model).
> The laser inside it was matched up to the manufacturers
> site to have a lamda of 1550 nm. (It appears Juniper makes
> two version of this card? 1550 and 1310 nm?)

Juniper LH is 1550nm, and LX is 1310nm. This is different from Cisco, 
where ZX is 1550, and LX and LH are used interchangably to refer to 1310. 
The actual transceiver is the OCP 70km 1x9, you can find the detailed 
specs on their website.


> I used a 10dbm attenuator on the TX which went to a Riverstone
> gige card (LX mid-range) which uses lamda 1310. The Riverstone
> TX also has a 10dbm attentuator attached.

Lose the 10dBm attenuator on the 1310 TX, it is unnecessary.

> I THOUGHT the RS card was 1550 nm during this whole time.
> Well, the darn thing works. I was using a test set to make
> sure my signal was up to par when I discovered my error.
> These devices that produce the laser must have quite the
> wide range from its center.

It's not that the TX is wide, it's that the RX is wide. These transceivers 
all use the exact same components for RX, which are usually wide band and 
will receive anything in the 1200 - 1600ish range. You need to attentuate 
the 1550nm at short distances so that your launch (between -3 to +3dBm, 
usually around 0-1dBm) isn't higher than your saturation (-3dBm), but that 
is it. You really need a minimum of 6dB attenutation to guarantee success 
assuming you don't have any other loss, so 10dBm is pretty much perfect.

> Well, my questions are, has anyone tried this before?
> If you did, what were the results?

Sure, works fine. It is a perfectly valid config too, and self explanatory 
if you understand what is going on under the hood. :)

Richard A Steenbergen <ras at e-gerbil.net>       http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)

More information about the juniper-nsp mailing list