Question Ethernet to Garage

Basinguy

Prominent Member
Joined
Oct 8, 2006
Messages
740
Reaction score
137
Points
258
Age
35
Location
Basingstoke
Currently i've got a standard ethernet cable running from my house to my garage.
I need networking in the garage as it runs 2 CCTV cameras as well as my weather station and door alarms through a gigabit switch

When i plug my laptop directly into the cable (garage end) windows tells me the connection speed is 1.0Gbps but the connection speed is actually pretty poor, no issue anywhere else in the house.
I figured this is probably because the ethernet cable is run along a catenary wire for a couple of meters between the house and the garage along with the power cable, and is probably picking up interference.

I figured my options are:

  1. use Wi-Fi bridge. The Wi-Fi access point is at the other end of the house with no real easy method to move it, Wi-Fi signal in the garage is almost 0 so that option is out.
  2. Use shielded ethernet cable and ground the shield to remove interference.
  3. Get a pair of media converters and use fiber between the house and the garage.

Anyone with experience please comment on which solution is best

Thanks :)

p.s I cant run the ethernet cable separately from the power cable without digging up the garden which is not preferred
 
If you have power in the garage fed from the same supply as the house (and its single phase) you can use powerline adaptors - I feed a 8 port POE switch in my outbuilding using these and have no issues
 
Personally, I would get some fibre. No chance of interference. I definitely wouldn't downgrade to power line adapters.
 
Are you sure you have the cable terminated properly using the TIA-568B wiring scheme (although 568A is also ok as long as both ends match)? Very often when I get reports of poor performance or no connection over a “tested” cat cable I often find it’s been incorrectly wired using the same randomn wire sequence, which passes the simple continuity tests of a basic cable tester, but would show up as failed with “crossed pairs” with a proper test kit that actually certified the performance of the link. Just having 8 wires in the same order is not good enough, and the correct cable sequence ensure the signals are passed on twisted pairs that help reject interference etc.

If you are sure you have correct terminations then by all means consider a new cable, but don’t go down the shielded cable route unless you can properly deal with the cable shielding earth requirements otherwise you may end up with more problems, not less!

If you want to use optical cable as suggested earlier then mulrimide OM4 would be a good place to start - LC duplex OM4 with media converters at either end would work, or if you switches have SFP ports you could add LC SFP modules to allow you to connect the fibre directly.

Do make sure that any external runs are either in a conduit for protection or an outdoor grade ruggedised jacket.
 
To be honest the best solution should be the cable. As another has said it needs proper connection and I have a decent switch garage end.

it was I use for my garage and it works perfectly.
 
It’s unlikely that your cable is picking up interference. More likely you have a problem elsewhere. I’d run iPerf to start with and see what results it gives you. It will help get to the root cause of the problem. If the cable is the problem then replacing it with fibre maybe a better option.
 
Thanks for all the suggestions.

Are you sure you have the cable terminated properly using the TIA-568B wiring scheme (although 568A is also ok as long as both ends match)?

Yep, the cable is terminated to the 568B standard. I've checked both ends.

Of course, I cant check the cable along its length but I'm assuming (possibly incorrectly) that if one of the 4 gigabit cores were broken then it would drop back to 100M, and if one of those were broken, it wouldn't work at all?

The cable from the garage connects directly into the router in the living room, which everything else also connects to, so I'm not sure where else the problem could lie.

I'd be tempted to try out a fibre link, but hesitant to go through all the cost and effort without really knowing where the issue lies.

Is it possible to do a crude speed test of that link?
I could do a broadband test but I'm fairly sure that wouldn't tell me much as my internet is only c.30Mb so i suspect the cable could handle that.
I could move a file from my laptop to my Synology through the network, but even moving from my PC gives varied transfer speeds so I don't think that would work either.
 
Use iPerf3

there is a Docker image for it you can install on your Synology NAS. Then with your laptop at the other end of the connection run the client and connect back you your NAS.
 
Thanks for the suggestion. I'll give that a try :thumbsup:
 
Right, the results are in of the test using iperf.
I used a 60s test for each

Connected via gigabit ethernet switch = 851 Mbit/sec
Connected to garage via gigabit switch x2 = 853 Mbit/sec

so looks like there is no issue with that cable at all :confused:
 
ive got cables running down catenary wires for about 8m one run and prob about the same on the 2nd.
Not got any bottlenecks because of wiring and been up approx 5 years or so. I only ran bog standard cat5e cable as it was all i had to hand at the time and was supposed to be a temp run anyway in standards with the intention to replace with outdoor spec. Allas, wont happen as things changed and now the cables are actually inclosed where they run other than outside! :D
Also got powers running alongside with some fairly large draws running through them too.... not a problem noticed.

I'd say it could either be a break in the cable rather than interference if its not enclosed in some protection or outdoor grade cabling. Or some peice of equipment has failed...

Have you got a gigabit switch or is it 10/100 one?

Whats the spec on the cameras and weather station/ equipment you've got? Some might only be 10/100 speeds anyway...
 
People get confused about running power and network cabling in close proximity. The recommendation for separation is not because the power cable may cause interference in the network cable, but to prevent induced power in the network cable from the power cable. It’s a safety issue not a data issue.
 
People get confused about running power and network cabling in close proximity. The recommendation for separation is not because the power cable may cause interference in the network cable, but to prevent induced power in the network cable from the power cable. It’s a safety issue not a data issue.

Isn't it both?
 
Right, the results are in of the test using iperf.
I used a 60s test for each

Connected via gigabit ethernet switch = 851 Mbit/sec
Connected to garage via gigabit switch x2 = 853 Mbit/sec

so looks like there is no issue with that cable at all :confused:

That's a bit lower than I'd expect but not far off.
 
Isn't it both?
Not as far as I am aware, its purely to prevent inductance. That is why there is a specified separation distance. There are thousands of meters of network cabling installed in close proximity to LV cabling and I have never heard of it causing interference. You are more likely to interference from NEXT or Shielding being connected at both ends.
 
Consider the respective rates involved - mains (in the UK) runs at 50Hz and 100/1000mbps ethernet (both) pulse the line a 125,000,000Hz (or more correctly "symbols per second" AKA "baud.") Symbol to symbol, any mains induced voltage changes are going to look like DC.

Over copper cable, ethernet encodes the data as the "change" in voltage, symbol to symbol, rather than the actual value of the voltage itself (as that varies with cable length.) As long as the transceiver can "lock on" to and track the voltage changes, it doesn't actually matter what the voltages themsleves are (or whether the duration of the symbols drifts over time - ethernet is called "self clocking" in that there's no separate timing signal required.)

By way of example, imagine it works like this (roughly based on the 100mbps "fast ethernet" signalling scheme.) The line is constantly cycling at +3v, +1v, -1v, -3v, -1v, +1v, and back to +3v,. We encode the data such that a change in voltage encodes a "1" and two successive symbol periods without a change encodes a "0." We also then use a "line code" that means there are never two successive "0's" and most of the symbols are "1's" which helps us lock on and track the symbol period (as there's lots of changes happening and a rate we can predict from the standards.) At the receiver, we don't need to "measure" the actual voltages, we just need to track the changes in order to recover the data - change == "1," no change == "0." Thusly if the actual value of the voltages drifts a bit over the course of hundreds of symbols, or is shunted a bit by some DC offset, a decent transceiver should be able to track it and still grab the data. It also uses "balanced" transmission (over each wire pair) which also helps cancel out any induced voltage changes.

I should add the caveat that whilst I'm interested enough in electronics to have gone and found out about all this, I'm not an electrical or electronic engineer by either qualification or inclination, so the bona-fides might be having a good laugh at my expense! :)

So my understanding has always been what @mushii says - the "not running cabled parallel" thing is about mitigating a shock hazzard from induced currents/voltage rather than mitigating low frequency interference in the data transmission. I'd think that "radio" like frequencies in the millions of Hz (Wi-Fi anyone..?) are likely to be a much bigger problem.

And I'll add my usual anecdote that I've had (over my carreer) literally thousands of UTP cables run in through things like plant rooms, and I've never had any problems due to interference.
 
Last edited:
It's the induced current from the power that causing the problems/interference (EMI). Ethernet only uses a p.d. of a few volts on the wire.

A bit like mickevh, it's very rare that I've had issues with interference, but then we run a lot of STP/FTP especially on routes near the HV. Or use fibre on the longer runs.
 
Last edited:
Likewise - if I'm at all "worried" about interference and/or have long cable runs to do, I prefer to use optical fiber. I also prefer fiber for 10G. In conversations with my "sparky" colleagues over the years, I've found they also tend to be happier with fiber between buildings as, despite the transformer coupling in the ethernet transceivers, it removes an electrical pathway and make them less worried about "ground loops" and "earth lifts" (whatever they are!)
 
Last edited:
When i plug my laptop directly into the cable (garage end) windows tells me the connection speed is 1.0Gbps but the connection speed is actually pretty poor, no issue anywhere else in the house.

Right, the results are in of the test using iperf.
I used a 60s test for each

Connected via gigabit ethernet switch = 851 Mbit/sec
Connected to garage via gigabit switch x2 = 853 Mbit/sec

so looks like there is no issue with that cable at all :confused:

So back to the original problem :)

What is telling you\why do you think the connection speed is poor?

How are you testing etc?

What symptoms are you seeing?
 
So back to the original problem :)

What is telling you\why do you think the connection speed is poor?

How are you testing etc?

What symptoms are you seeing?
Thanks to all for your contributions, very interesting discussion has been made.

The original problem is that when I am in the garage I tend to bring my laptop and plug in to the network so that I can play either Internet radio or music from my nas.
Doing either of these results in lots of buffering and random 'no Internet' messages from Windows for short periods. It's very random often works fine for half an hour or and hour before having a wobble.
It does the same whether I'm playing Internet radio or music from my network so I can kind of isolate these as being an individual problem.
I can do the same inside the house for hours on end without even a tiny problem, either wired or on WiFi (just tested speed over the WiFi connection using iperf and managed a whopping 3mbps and yet the radio is playing seamlessly)
 
3Mbps is fine for Internet radio.

Did you disable your laptop WiFi when you plug into the cable. Depending on the configuration some won't so may continue to use the WiFi rather than cable.


use Wi-Fi bridge. The Wi-Fi access point is at the other end of the house with no real easy method to move it, Wi-Fi signal in the garage is almost 0 so that option is out.

Do you have an access point in the garage, sounds like you don't from above so that's why the connection is so bad. It is persisting with the WiFi connection over the plugged in one.
 
No access point in the garage. I will have a try later and see if the Wi-Fi is still active
 
Generally Windows will favour the fastest connection which will often wired. But if you have something "in session" over Wi-Fi when you plug in the network cable, it will continue (at least that session) over the incumbent. So if, for example, you are copying a file or steam video or whatever over Wi-Fi, then plug in a network cable, it'll continue using the Wi-Fi link for the stuff already "in flight." Just click off the Wi-Fi (which will kill anything in progress) and that'll force it to use the wired link.
 
Generally Windows will favour the fastest connection which will often wired. But if you have something "in session" over Wi-Fi when you plug in the network cable, it will continue (at least that session) over the incumbent. So if, for example, you are copying a file or steam video or whatever over Wi-Fi, then plug in a network cable, it'll continue using the Wi-Fi link for the stuff already "in flight." Just click off the Wi-Fi (which will kill anything in progress) and that'll force it to use the wired link.

I have seen quite a few cases of Windows being thick for lack of a better word and it still using the WiFi connection. It depends sometimes on the metric it has been given.

MacBook, however, will remain on the WiFi connection even when there is a 10Gbps adapter available, unless you tell it to prioritise the other one. Even then it forgets sometimes.
 
Thanks for all the suggestions.



Yep, the cable is terminated to the 568B standard. I've checked both ends.

Of course, I cant check the cable along its length but I'm assuming (possibly incorrectly) that if one of the 4 gigabit cores were broken then it would drop back to 100M, and if one of those were broken, it wouldn't work at all?

The cable from the garage connects directly into the router in the living room, which everything else also connects to, so I'm not sure where else the problem could lie.

I'd be tempted to try out a fibre link, but hesitant to go through all the cost and effort without really knowing where the issue lies.

Is it possible to do a crude speed test of that link?
I could do a broadband test but I'm fairly sure that wouldn't tell me much as my internet is only c.30Mb so i suspect the cable could handle that.
I could move a file from my laptop to my Synology through the network, but even moving from my PC gives varied transfer speeds so I don't think that would work either.
I use LAN Speed, simple and free..........

 

The latest video from AVForums

Is 4K Blu-ray Worth It?
Subscribe to our YouTube channel
Back
Top Bottom