Fiber Optic vs Cat6

[Ricoch3T]

I went with Cat6 shielded for some hdmi streaming. The problem for me is trying to figure out how to terminate without problems.

 

[Vinci]

We're running 10G over old single mode fiber too, which kinda surprised us first time we tried turning up a link. Our 12 year old million bucks worth of multi-mode is no good for 10G. So if you buy fiber, better bone up on what kind for the distance and speed. It's not obsolescence proof.

You can run 10G over multi-mode, but not very far. 50 micron works better than 62.5, FWIW. I have a couple short links at work that we are able to do it with, but everything beyond a couple hundred feet has to be single-mode.

 

[mopho]

I went with Cat6 shielded for some hdmi streaming. The problem for me is trying to figure out how to terminate without problems.

Keep it as tight as you can. It's not as big of a deal as people make it out to be. Plus, you will never know unless you certify it. ( about 5k worth of test gear) Cat 5E speeds (gigabit) are easily attainable and more than sufficient.



--FIRST POST

 

[ShawnJ]

We burn them in that way, but we wont run them long term. You can spend whatever you want on networks. Here's our last burn in a few months ago. I wish I could spend that on tools!

How'd you get a pic of my office?

 

[Ryf]

I would say if your just running internet/etc for parts lists and pricing etc, streaming music.. cat 5e or cat 6 is rated to 295 feet. your limit in most residential situations is your internet connection (10-50meg for most), so why spend $$ on hardware to break? run two lengths of cat 5e or 6..

IF I was going to buy hardware, I would buy a wifi router for the shop, then my phone would have internet as well as the computer.

my shop is just past the limit of the houses wifi, my options were external antenna to house or wire I am running wire and adding a wifi router in the shop. not saying you should just saying.

 

[Stuntmonkey]

You connection is only as fast as the slowest point, and I'm guessing you have coax/copper coming into the house, therefore you're never going to go faster than that.

 

[tjcanno]

I am reviving this old thread because it is relevant to the planning I am dong for my new garage in Hiawassee, GA.

I built a new detached garage/shop late last year that is about 350 feet from the house. I trenched and ran power out there to a 100 amp subpanel through conduit. I also laid a second conduit for ethernet, cable TV, phone, anything else I want that is not power. Right now the second conduit has no wires in it, just a pull string. I did not know to keep the two conduits spaced apart from each other, so that might be a problem...

I was planning to run a couple of Cat5 or Cat6 out through the conduit, plus a few other wires, like cable TV. I would like to have a high speed link out there to my shop to access big files stored on the computer in the house. I brought the conduit right into the room in the house where the big computer is.

Can Cat6 handle this distance?

I was thinking that it might need to be fiber, which is how I ended up in this year-old thread.

Even if I ran fiber optic, I'd probably run a Cat6 anyway.

Has anything changed in the last year that would make fiber more viable now?

Thanks for your input.

Picture of front of new building below.

 

[mooseracing]

350ft is pushing it for CAT spec. You would need a switch at both ends and you might get lucky.

Fiber ***** unless it is long haul, indoor Copper is still king and is backward compatible 10Gb-10Mb. It makes emergency replacement so much easier in the business side.

 

[MFortie]

Cat x copper standards are established for 100 meters -- 90 meters (295') for the permanent link segment and the remainder for patch cords (330' in total).

You will get a signal at 350' but it will not pass any standards testing - but you won't be testing anyway.

I would buy the best Cat 6a cable you can afford (6E is not a standard BTW) and live with a dropped packet or two - you'll probably never even notice an issue.

TE Connectivity (AMP for us older folks) just released a new gel filled indoor/outdoor Cat 6 (can't recall if it's 6a) cable - if you want, I can get you the part number when I get back in the office next week.

And, just as an FYI, structured cabling design is what I do for a living...

 

[big.jim]

cat 6 comes in 1000 foot boxes because that is the limit you can run without another switch to boost signal, i have wired factories with multiple runs for cat 6 back to patch cabinets ,cat 6 will run gigabit with no problem , patch cabinets only used to have 1 cat 6 cable feeding from main coms room with a redundant spare ran alongside

 

[jeffmoss26]

"cat 6 comes in 1000 foot boxes because that is the limit you can run without another switch to boost signal"

Not true. The spec is 100 meters/328 feet.

 

[CNGsaves]

cat 6 comes in 1000 foot boxes because that is the limit you can run without another switch to boost signal, i have wired factories with multiple runs for cat 6 back to patch cabinets ,cat 6 will run gigabit with no problem , patch cabinets only used to have 1 cat 6 cable feeding from main coms room with a redundant spare ran alongside

Really ???? Surely you don't believe that Cat 6 is capable of 1000 ft ???

Cat 6a is ethernet with it's limitations of approx 100 meters or 328 ft.
http://en.wikipedia.org/wiki/Category_6_cable

OP . . . Go with MFortie or JeffMoss advice . . . . they know what they're talking about !!

 

[happy2rv]

cat 6 comes in 1000 foot boxes because that is the limit you can run without another switch to boost signal, i have wired factories with multiple runs for cat 6 back to patch cabinets ,cat 6 will run gigabit with no problem , patch cabinets only used to have 1 cat 6 cable feeding from main coms room with a redundant spare ran alongside

Dental floss comes in 100 ft rolls, I hope I never meet the person that needs 100' for one flossing. I found 2000' rolls of cat5e direct burial, but I wouldn't expect to get anywhere near that far. The spec is 350 feet, and with any quality cable, you should be guaranteed to get that. In reality, you will be able to go further with no noticeable degradation in signal quality. How much further you can go will depend on a number of factors, but I really wouldn't expect 1000' or even 500' to be reliable.

There are benefits to fiber that have barely been touched on or ignored all together. The biggest benefit to fiber is not the increase in length possible, but the electrical isolation provided. I believe someone suggested earlier one benefit of this is the prevention of possible grounding issues if the building's grounds aren't tied together as they should be. But another issue is the potential for lightning damage. Being run in underground conduit minimizes the chance for lightning but any surges or lightning on either end is likely to destroy all equipment at both ends of a copper cable.

Copper is definitely cheaper and easier. I ran copper for mine, but make an informed decision...

 

[rlitman]

Really ???? Surely you don't believe that Cat 6 is capable of 1000 ft ???

Just to clarify, Cat 6 is capable of as long a spool as you can buy. If you could get a spool miles long, it would be fine.

ETHERNET is not, over copper Cat 6. This has nothing to do with signal degradation over distance, so the distance limitation is not based on the quality of the wire or signal. It has to do with signal timing specific to ethernet, and the speed that a signal travels over a copper wire.

But Cat 6 cable will happily carry signal longer distances using other protocols, such as RS-485, DSL, ISDN, etc. With longer distances however, come steeper bandwidth limitations.

As an example of this, where I work, we have an underground conduit that goes across the parking lot to a utility pole with IP cameras mounted on it. With a run of right around 400 feet, 100baseT was not establishing link. 10baseT worked sporadically. Eventually we pulled out the copper and replaced it with fiber, and now we get gigabit to the pole flawlessly (which helps when you have an HD PTZ, plus two more HD virtual PTZ's) on it.

 

[CNGsaves]

^ ^ ^ True, copper is capable of over 330 ft, but for normal "Joe homeowner" the realistic distance limitation is around 350 ft. Joe homeowner would not be putting in the remote nodes or other equipment needed for other protocols to extend distance limitations. Old school DSL (digital subscriber line) has an approximate distance limitation of 18,000 ft over good wire !!

 

[Sludge Puppy]

THANK YOU mrjaw, from a fellow network admin.

Concur as well from another fellow network engineer.

Run the Cat 6 in conduit, save the time, and money, and energy from the fiber.

 

[2ManyProjects]

I am reviving this old thread because it is relevant to the planning I am dong for my new garage in Hiawassee, GA.

It really ought to have been posted as a new stand-alone thread; but that's water under the bridge at this point.

I built a new detached garage/shop late last year that is about 350 feet from the house. I trenched and ran power out there to a 100 amp subpanel through conduit. I also laid a second conduit for ethernet, cable TV, phone, anything else I want that is not power. Right now the second conduit has no wires in it, just a pull string. I did not know to keep the two conduits spaced apart from each other, so that might be a problem...

It won't help; but it's also not very likely to be a huge problem, as long as there was SOME separation (even if only an inch or two) between the two conduits when they were laid in the trench. The point is to avoid induced signals in the low-voltage lines, due to their close proximity to "heavy" current-carrying conductors. You can easily enough test for this by simply running some (2 or more conductor) wire in that second conduit, tie the two conductors together at one end with a fairly high-value (100k Ohm to maybe 1M Ohm, perhaps) resistor, and measuring for AC voltage at the other end using a high-impedance VTVM, or an oscilloscope. If you get no more than a few microvolts, you're good to go. BTW... For the test to be valid, you need to turn on EVERYTHING in the shop, so as to produce the maximum possible current flow through the feeder cable during the test.

I was planning to run a couple of Cat5 or Cat6 out through the conduit, plus a few other wires, like cable TV. I would like to have a high speed link out there to my shop to access big files stored on the computer in the house. I brought the conduit right into the room in the house where the big computer is.

Can Cat6 handle this distance?

As noted by others, "CAT-6", per se is not the issue; "Ethernet" is the issue.

Officially, the maximum TOTAL length for any one Ethernet segment is 100 Meters (~328 feet), INCLUDING the patch cables at each end. If you keep the patch cables as short as possible, there is a fair chance that you could "get away with" 350 feet; but your effective throughput, collision rate, dropped frames, etc. will likely all be somewhat short of ideal.

I was thinking that it might need to be fiber, which is how I ended up in this year-old thread.

You don't want to mess with fiber. It would be an expensive PITA for this application.

If simply "stretching the spec" does not produce the results you're looking for, one possible approach would be to re-open the trench somewhere near the middle of the run, then dig a hole nearby big enough to accommodate a moderate-size all plastic and completely sealable NEMA enclosure. (You'll want at least NEMA Type 4; preferably Type 6 or 6P) Re-arrange the conduit to run into, and back out of that enclosure. Make certain that ALL of the connections are thoroughly sealed. In the enclosure, put a small inexpensive Ethernet switch, such as (just for example):

http://www.newegg.com/Product/Product.aspx?Item=N82E16833704042

Ideally, you want one which does the AC/DC rectification (and at least some of the smoothing/regulation) within the wall wart itself, as opposed to simply feeding reduced-voltage AC into the unit; but it's not a big deal.

Now you're going to need to do some "surgery" to the switch's "wall wart" power supply: Making VERY sure to keep polarity straight, cut the cord between the wall wart and the plug which connects to the switch. Run some appropriate wire (need not be anything "heavy"; one AWG grade up from whatever the stock wall wart used will be fine) from whichever end of the conduit is closer (I'm guessing the shop/garage) to your new NEMA box. Splice the wires from the wall wart and the plug to each end (again, polarity is CRITICAL, if it is a DC feed), as appropriate. Plug in the wall wart and test your Ethernet connection(s). Assuming that goes well, seal up the box, bury the whole thing, and don't look back. (You MIGHT want to buy a second switch at the same time, so as to have a ready spare if/when the seal on your underground NEMA enclosure fails, and the switch gets trashed by water or similar. They're certainly cheap enough.)

 

[ishiboo]

It really ought to have been posted as a new stand-alone thread; but that's water under the bridge at this point.



It won't help; but it's also not very likely to be a huge problem, as long as there was SOME separation (even if only an inch or two) between the two conduits when they were laid in the trench. The point is to avoid induced signals in the low-voltage lines, due to their close proximity to "heavy" current-carrying conductors. You can easily enough test for this by simply running some (2 or more conductor) wire in that second conduit, tie the two conductors together at one end with a fairly high-value (100k Ohm to maybe 1M Ohm, perhaps) resistor, and measuring for AC voltage at the other end using a high-impedance VTVM, or an oscilloscope. If you get no more than a few microvolts, you're good to go. BTW... For the test to be valid, you need to turn on EVERYTHING in the shop, so as to produce the maximum possible current flow through the feeder cable during the test.



As noted by others, "CAT-6", per se is not the issue; "Ethernet" is the issue.

Officially, the maximum TOTAL length for any one Ethernet segment is 100 Meters (~328 feet), INCLUDING the patch cables at each end. If you keep the patch cables as short as possible, there is a fair chance that you could "get away with" 350 feet; but your effective throughput, collision rate, dropped frames, etc. will likely all be somewhat short of ideal.



You don't want to mess with fiber. It would be an expensive PITA for this application.

If simply "stretching the spec" does not produce the results you're looking for, one possible approach would be to re-open the trench somewhere near the middle of the run, then dig a hole nearby big enough to accommodate a moderate-size all plastic and completely sealable NEMA enclosure. (You'll want at least NEMA Type 4; preferably Type 6 or 6P) Re-arrange the conduit to run into, and back out of that enclosure. Make certain that ALL of the connections are thoroughly sealed. In the enclosure, put a small inexpensive Ethernet switch, such as (just for example):

http://www.newegg.com/Product/Product.aspx?Item=N82E16833704042

Ideally, you want one which does the AC/DC rectification (and at least some of the smoothing/regulation) within the wall wart itself, as opposed to simply feeding reduced-voltage AC into the unit; but it's not a big deal.

Now you're going to need to do some "surgery" to the switch's "wall wart" power supply: Making VERY sure to keep polarity straight, cut the cord between the wall wart and the plug which connects to the switch. Run some appropriate wire (need not be anything "heavy"; one AWG grade up from whatever the stock wall wart used will be fine) from whichever end of the conduit is closer (I'm guessing the shop/garage) to your new NEMA box. Splice the wires from the wall wart and the plug to each end (again, polarity is CRITICAL, if it is a DC feed), as appropriate. Plug in the wall wart and test your Ethernet connection(s). Assuming that goes well, seal up the box, bury the whole thing, and don't look back. (You MIGHT want to buy a second switch at the same time, so as to have a ready spare if/when the seal on your underground NEMA enclosure fails, and the switch gets trashed by water or similar. They're certainly cheap enough.)

If you go through this whole process of digging a hole, it'd be easier to just run POE that distance to a small switch versus hacking up a wall wart/etc. I have a couple outdoor access points that I use the D-Link POE injectors/splitters ($30) with and they work great… provides 1A @ 12v. This ups the voltage to send it over the cable and then lowers it at the destination so current is not a big issue.

Instead of burying the whole thing deep, T into the conduit and bring it up near or to the surface.

Note for GigE/etc. you will need a more expensive unit as they use those pins, but you could run a basic Cat5 or Cat3 just for power.

My guess is there will be no issues with that distance though.

 

[98ramtough]

Keep it as tight as you can. It's not as big of a deal as people make it out to be. Plus, you will never know unless you certify it. ( about 5k worth of test gear) Cat 5E speeds (gigabit) are easily attainable and more than sufficient.



--FIRST POST

When using shielded cable if you don't use the proper terminations and ground the plugs you are making matters worse. Your basically creating a huge antenna to bring in interference. Why in the world would you want shielded cable for home use? Are you running an old school PRI or T1 from an NIU???

 

[rlitman]

If you go through this whole process of digging a hole, it'd be easier to just run POE that distance to a small switch versus hacking up a wall wart/etc. I have a couple outdoor access points that I use the D-Link POE injectors/splitters ($30) with and they work great… provides 1A @ 12v. This ups the voltage to send it over the cable and then lowers it at the destination so current is not a big issue.

Yes, the POE voltage is higher, so it will work better over a distance than a 12V adapter, but the voltage drop over that distance may be so much that it will not power up the switch at the other end.

In my case (see my post above), we ran 12gauge copper in the conduit with the fiber, and powered a 120V outlet in the box on the pole. That was enough to run several cameras. The PTZ is a power hog, as it has a heater in the dome to prevent fogging.

 

[2ManyProjects]

If you go through this whole process of digging a hole, it'd be easier to just run POE that distance to a small switch versus hacking up a wall wart/etc.

IF you can find an appropriate switch which is actually powered via PoE, sure. But that may not be as easy as it sounds. I've never had need for such, so I did not do an extensive search; but pretty much all of the ones I did find via quick Google & NewEgg searches, while providing PoE on at least some of their ports, still require external power.

Yes, the POE voltage is higher, so it will work better over a distance than a 12V adapter, but the voltage drop over that distance may be so much that it will not power up the switch at the other end.

I kind'a doubt that. These small switches don't draw very much power at all; typically WELL under an amp @ 12V. Just to throw a dart:

CM = K x I x L / E
CM = 12.9 x 0.5 x 150 / 0.36
CM = 2,687.5

AWG 16 = 2,583 CM

So in theory, you'd want AWG 16 to keep voltage drop right at 3%. But at these voltages, 3% is SO little (only about 1/3 volt), and the devices themselves are generally so tolerant of varying input voltage, that we can confidently exceed that by what would normally seem "excessive" margins. So if one were to drop back to AWG 18 or perhaps even AWG 20, I'd be very surprised if the resulting voltage drop (which would still be well under a volt) would be enough to threaten the reliable operation of the switch. If push comes to shove, move the box/switch location somewhat closer to whichever end it is being powered from (say, at the 250-300 foot mark, leaving a 50-100 foot run to the power supply).