3

I'm fairly new to fiber cabling. I have an OM1 cable (running 750 metres) already installed on a site that currently works fine on an old system that is being upgraded. I'm trying to find out if these Cisco compatible SFP transceivers (http://www.farnell.com/datasheets/2037500.pdf) will be able to run at 10 and hopefully 100 Mbps. Fiber will be spliced for duplex LC connectors.

2950s being specified.

SFPs run at 1310nm which according to this table: OM Cables

Is this better than running a SFP at 850nm. Greater performance/distance for OM1.

Can anyone provide me their thoughts on if it is compatible? I read that OM1 and 2 is LED and OM3 and OM4 is laser? I am slightly confused by this.

Improve this question

3 Answers 3

Reset to default
4

Not likely.

The transceiver you've linked to is for the 1000BASE-LX10 standard. It uses 1310 nm wavelength and has a reach of 550 m on 500 MHz·km fiber like your OM1. You'd exceed the reach by 30+% which is not likely to work (reliably). OM1 has limited use, the higher the bandwidth the shorter the reach. A PHY only runs at the indicated speed(s). There's no throttling when the link quality is poor, it simply won't work at all or cause errors.

Since you're only looking for 10 or 100 Mbit/s - the according PHY is 100BASE-FX which has a reach of 2000 m on MMF (pretty much any grade). Transceivers are still available, check for compatibility with your switches.

If you require more bandwidth you can try working around with link aggregation, WDM and such - but most often running higher-grade fiber is the better option in the long run (I'd go for single mode for this distance).

PHYs up to 100 MBit/s often use LED sources which are limited by frequency and are rather unfocused for fiber coupling. Gigabit+ PHYs require lasers that are very focused and benefit from OM3+ "laser-optimized" fiber. You can use any combination but the lower the fiber quality, the shorter the reach for higher speeds.

Improve this answer
0
2

Actually if your fiber is good and your splices / terminations are good, you very well may make 750 meters with 1000baseLX transceivers.

With mode conditioning patch cords, 1000BaseLX has a specified distance limit of 550m. This distance is based on loss, not timing, with worst-case fiber quality / loss. If your fiber is better than worst-case, you can go longer distances. In practical use it's not common to go much, much further than the 550m distance minimum. Cisco will even support distances beyond 550m under certain conditions with testing.

You'll want to make sure your connections are all clean, terminations are all order, patch cords are of the best quality and in the best possible condition to have the best possible chance of success.

If you have a meter that can test at 1300nm or 1310nm, use the best strands you have available rather than the next two unused strands in the patch panel.

Improve this answer
4
  • The problem isn't only attenuation but especially modal dispersion with 'smears' the signal. It may work but hopes shouldn' t get high.
    – Zac67 ♦
    Commented Aug 21, 2018 at 6:29
  • Hi, We got link up yesterday to asess the cable. Will be conducting throughput/error checking tests at various rates. Batsplatsterson - We have some conditioning patch leads to run single mode over multi mode. Will report back on results. They are alos a number fibres free and will consider load balancing/bonding etc if needed.
    – Aidan Venn
    Commented Aug 21, 2018 at 9:11
  • @AidanVenn - glad to hear it, and thanks for the follow up. I have had similar results many times in a number of facilities I installed multimode in in the 90's, saved a bundle compared to installing new single mode to get gigabit backbone.
    – batsplatsterson
    Commented Aug 21, 2018 at 9:14
  • @Zac67 - well aware of how signal degrades in fiber, just explaining that the 550m is a worst case number, typical fiber will reach further, and it's widely known and accepted that that is the case.
    – batsplatsterson
    Commented Aug 21, 2018 at 9:17
0

I ran tests and got serious fluctuations. Speed ranging from 200 to 400 Mbps. Not consistant at all. Ran single mode over MM using patch leads. This is greater than is required for now so hopefully will. Had a chat with client that future budgets need to consider an upgrade to the link.

Thanks

Improve this answer
1
  • I was afraid of that... I'd prefer a stable 100 Mbit/s link (or two aggregated ones) to a assumedly frequent-error "200 to 400 Mbit/s" gigabit link though.
    – Zac67 ♦
    Commented Aug 24, 2018 at 20:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.