User avatar
jbeale
Posts: 3476
Joined: Tue Nov 22, 2011 11:51 pm
Contact: Website

maximum practical length of ethernet (UTP) ?

Tue Mar 07, 2017 11:40 pm

I know the standard says 100 meters is the maximum length over the standard UTP cable, for both 10BaseT and 100BaseT.
http://www.cisco.com/c/en/us/support/do ... hbase.html

I've also heard anecdotal evidence that people have successfully run IP cameras using much longer cables than that.

I have a project in mind needing a really long run, and it would be better to avoid using a repeater or hub. RF wouldn't work at all.
The right way is probably fiber optics but I'm trying to be cheap.

I note that you can get 1000' and even 2000' spools of Cat5e and Cat6 but before I buy anything, anyone have experience to report along those lines?

321
Posts: 94
Joined: Mon Feb 13, 2017 7:26 pm

Re: maximum practical length of ethernet (UTP) ?

Tue Mar 07, 2017 11:56 pm

How long a run and what sort of environment (temperature, electromagnetic interference) and what data through put do you need, eg hooking up a camera at the end of garden say or something else?

Whilst you can get less dropped packets & retrys with good quality shielded cable, depending on you needs particularly the run length, might dictate what you have to do.

W. H. Heydt
Posts: 10755
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: maximum practical length of ethernet (UTP) ?

Wed Mar 08, 2017 1:42 am

The maximum specified Ethernet *backbone* run is (IIRC), 2Km, but that requires using 1" coax and echo cancellation on the ends.

mfa298
Posts: 1387
Joined: Tue Apr 22, 2014 11:18 am

Re: maximum practical length of ethernet (UTP) ?

Wed Mar 08, 2017 9:20 am

jbeale wrote: I've also heard anecdotal evidence that people have successfully run IP cameras using much longer cables than that.

I have a project in mind needing a really long run, and it would be better to avoid using a repeater or hub. RF wouldn't work at all.
The right way is probably fiber optics but I'm trying to be cheap.
I've seen a few runs that were slightly over 100m that seemed to work, and shorter runs that have issues - there's too many variables to say something would definitely work.

It would help if you defined what you mean by a really long run (if you mean 110m then try it, if you mean 200m then I would doubt it'll work in any useful way).

As you've already said for a longer run (and I'd include anything going between buildings in that) the correct way is to use Fibre - If you only need 100mbps then you can probably get most of the required parts fairly cheaply on your favoured auction site.

geoffr
Posts: 85
Joined: Wed Aug 22, 2012 11:25 am
Location: Melbourne, VIC

Re: maximum practical length of ethernet (UTP) ?

Wed Mar 08, 2017 11:01 am

jbeale wrote:I know the standard says 100 meters is the maximum length over the standard UTP cable, for both 10BaseT and 100BaseT.
http://www.cisco.com/c/en/us/support/do ... hbase.html

I've also heard anecdotal evidence that people have successfully run IP cameras using much longer cables than that.

I have a project in mind needing a really long run, and it would be better to avoid using a repeater or hub. RF wouldn't work at all.
The right way is probably fiber optics but I'm trying to be cheap.

I note that you can get 1000' and even 2000' spools of Cat5e and Cat6 but before I buy anything, anyone have experience to report along those lines?
To comply with standards (CAT5, CAT5e, CAT6 etc.) you need to be under 100m, and more specifically 90m for your trunk (solid core) and 10m for the sum of the lengths of the flyleads either end (stranded cable).
You may be able to run more, but then you won't be able to certify the installation for compliance, and there is also no guarantee that it will work.
In some countries you will find that there are legal requirements that data cabling must be done by a qualified installer, but sometimes nobody checks.
In a commercial environment, I would recommend using a qualified installer, and having the installation certified. At home - up to you, I guess :lol:
For all the value of work done by qualified installers, I have seen some horrendous things done by commercial cabling companies.

tvjon
Posts: 710
Joined: Mon Jan 07, 2013 9:11 am

Re: maximum practical length of ethernet (UTP) ?

Wed Mar 08, 2017 11:04 am

@jbeale,

Is this long run line-of-sight?

User avatar
aTao
Posts: 1087
Joined: Wed Dec 12, 2012 10:41 am
Location: Howlin Eigg

Re: maximum practical length of ethernet (UTP) ?

Wed Mar 08, 2017 11:45 am

What data rate do you need, would other wired transmission systems work, audio modem for one?
>)))'><'(((<

User avatar
jbeale
Posts: 3476
Joined: Tue Nov 22, 2011 11:51 pm
Contact: Website

Re: maximum practical length of ethernet (UTP) ?

Wed Mar 08, 2017 7:52 pm

This is a speculative project and likely won't happen due to practical difficulties and expense of what is just an unfunded hobby anyhow.

The run would be entirely underground and underwater- straight down a vertical borehole. It is an old oil well now filled with water near the top and who knows what else lower down; it is I assume subject to collapse at any time in the future (or in the past?) County records show it was originally drilled about 500m deep. but it was closed and abandoned over 50 years ago. Records also claim the well was capped, but an open 60 cm diameter bottomless hole sitting there now says otherwise. The site is owned by a friend of mine.

I have a Raspberry Shake (RPi-powered geophone seismometer) at home, which is part of a global network (see: http://raspberryshake.net/stationview/ ) and it's working nicely. Like all such devices on the surface in an urban area, sensitivity is limited by "cultural noise" eg. washing machine, ventilation fans, and road traffic. You can mostly avoid that if you could bury your sensor deep enough down.

So, the old well seemed like an opportunity to try a fairly deep borehole sensor. The geophone sensor is designed for air ambient, I can't run it saturated with silicone oil like some underwater electronics do. The pressure housing on the sensor package would be enough of a pain to construct, so with 100baseT I would probably just stop around 100 m to avoid needing waterproof repeaters. It seems that plain multimode fiber is actually cheaper per meter than direct-burial Cat5 in longer lengths, but I would still have to deal with delivering a few watts of power also to run the Pi and the front end electronics, and I assume the fiber is not so mechanically rugged; the cheap stuff anyhow.

Like I said... the whole thing is just idle speculation. Fun to think about, though.

W. H. Heydt
Posts: 10755
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: maximum practical length of ethernet (UTP) ?

Wed Mar 08, 2017 8:53 pm

The other problem that I think you'd run into is that the geophone probably needs to be in solid contact with the ground...or the side of the borehole in this case. That would be...awkward...to arrange.

User avatar
jbeale
Posts: 3476
Joined: Tue Nov 22, 2011 11:51 pm
Contact: Website

Re: maximum practical length of ethernet (UTP) ?

Wed Mar 08, 2017 9:17 pm

Apparently the usual practice in downhole sensing is a long tubelike pressure vessel slightly smaller than the ID of the bore, dropped down with one or more external spring-loaded arms that start out folded in against the vessel. The arm extension is triggered by a signal or a cable pull once the correct depth is reached, whereupon the arm swings out and sideways, to brace the sensor package and wedge it firmly against the sidewall of the bore. Maybe the arm has some toothy bits at the end for a better grip. I can imagine that working. If that was the only problem with the plan, I'd think it was doable. :-)

User avatar
aTao
Posts: 1087
Joined: Wed Dec 12, 2012 10:41 am
Location: Howlin Eigg

Re: maximum practical length of ethernet (UTP) ?

Thu Mar 09, 2017 12:12 am

That far down a water filled hole... I would put as little as possible there, certainly not a network connected computer. Rather just the geophone and a line amplifier, with all the rest above ground. If you did hit 500m and its all flooded then thats 48 atmospheres.
>)))'><'(((<

wayne.dolesman
Posts: 64
Joined: Sat Feb 25, 2017 8:10 am

Re: maximum practical length of ethernet (UTP) ?

Thu Mar 09, 2017 1:29 am

Having some experience with this there is the standard and then there is what works.

Some cheaper devices have shorter ranges on the same wire than others. The faster it goes the shorter the run as a general rule as well. So 10Mbps can do what 1000Mbps cant - especially with consumer grade equipment.

Remember long wires are giant antennas and near lightning strikes can cause sufficient current to flow on the wire that stuff gets damaged. So it may work today but after the next storm it may not. YMMV.

I have successfully done 250 foot runs on cheap cheap cheap consumer grade gear at 10Mbps. I have done 200 foot runs with 100Mbps.

A POE range extender (some are weatherproofed) can dramatically help get the signal where you need it. They are not horribly expensive.

mfa298
Posts: 1387
Joined: Tue Apr 22, 2014 11:18 am

Re: maximum practical length of ethernet (UTP) ?

Thu Mar 09, 2017 7:57 am

jbeale wrote: So, the old well seemed like an opportunity to try a fairly deep borehole sensor. The geophone sensor is designed for air ambient, I can't run it saturated with silicone oil like some underwater electronics do. The pressure housing on the sensor package would be enough of a pain to construct, so with 100baseT I would probably just stop around 100 m to avoid needing waterproof repeaters. It seems that plain multimode fiber is actually cheaper per meter than direct-burial Cat5 in longer lengths, but I would still have to deal with delivering a few watts of power also to run the Pi and the front end electronics, and I assume the fiber is not so mechanically rugged; the cheap stuff anyhow.
PoE will add an extra hurdle to the setup, doing some quick sums for a device using if you're pulling 0.2A through the cable there's over 3v of loss over a 100m run. If you ran PoE to a proper specification (usually around 48v over the cable) then you would likely be fine, If you just tried feeding 5v to the Pi over that spare pairs then you'll have issues (remember the voltage drop will vary with current so if the current varied between 0.2 and 0.4A the voltage drop would be between 3 and 7 volts).
wayne.dolesman wrote:Having some experience with this there is the standard and then there is what works.

Some cheaper devices have shorter ranges on the same wire than others. The faster it goes the shorter the run as a general rule as well. So 10Mbps can do what 1000Mbps cant - especially with consumer grade equipment.
...
I have successfully done 250 foot runs on cheap cheap cheap consumer grade gear at 10Mbps. I have done 200 foot runs with 100Mbps.
...
The standard says that with the right cable (at least cat5 for 100BaseT or cat5e for 1000BaseT) you can do 100m (328ft). If you can't manage those distances then something isn't to the standard (bad or too many joints, wrong cable, wrong connectors etc.). A particular thing to watch out for is cheap network cable which is CCA (Copper Clad Aluminium).

User avatar
aTao
Posts: 1087
Joined: Wed Dec 12, 2012 10:41 am
Location: Howlin Eigg

Re: maximum practical length of ethernet (UTP) ?

Thu Mar 09, 2017 9:23 am

aTao wrote:That far down a water filled hole... I would put as little as possible there, certainly not a network connected computer. Rather just the geophone and a line amplifier, with all the rest above ground. If you did hit 500m and its all flooded then thats 48 atmospheres.

also, while some cable is specified as underground, this does not mean they will remain in spec at the depths you may encounter. With UTP spacing between core in a pair is vital to its performance, squashing a cable will ruin this.
>)))'><'(((<

texy
Forum Moderator
Forum Moderator
Posts: 5155
Joined: Sat Mar 03, 2012 10:59 am
Location: Berkshire, England

Re: maximum practical length of ethernet (UTP) ?

Thu Mar 09, 2017 9:53 am

jbeale wrote: I note that you can get 1000' and even 2000' spools of Cat5e and Cat6 but before I buy anything, anyone have experience to report along those lines?
Those spools would be for the cable manufacturer, so they can cut to the required length...
Various male/female 40- and 26-way GPIO header for sale here ( IDEAL FOR YOUR PiZero ):
https://www.raspberrypi.org/forums/viewtopic.php?f=93&t=147682#p971555

User avatar
RaTTuS
Posts: 10409
Joined: Tue Nov 29, 2011 11:12 am
Location: North West UK

Re: maximum practical length of ethernet (UTP) ?

Thu Mar 09, 2017 10:22 am

texy wrote:
jbeale wrote: I note that you can get 1000' and even 2000' spools of Cat5e and Cat6 but before I buy anything, anyone have experience to report along those lines?
Those spools would be for the cable manufacturer, so they can cut to the required length...
get a Crimper like
https://www.amazon.co.uk/Ethernet-Netwo ... B005FE84H2 [others will do]
do a bit of practicing
stick one on each end - and see if it works - then binary chop the real until it does ,....
How To ask Questions :- http://www.catb.org/esr/faqs/smart-questions.html
WARNING - some parts of this post may be erroneous YMMV

1QC43qbL5FySu2Pi51vGqKqxy3UiJgukSX
Covfefe

Return to “General discussion”