Ethernet Cable Speeds varies with a longer cable

[leonmoneill]

Hello,

I'm new to this forum but have an issues that i would love your opinion on. So the issue is i'm trying to increase speeds of my wired internet.
I have perfect 500Mb per second when plugged in in my sitting room (approx 40 meters away), Yesterday i ran an external eithernet cable outside along the top of the house and back into the house on the other side. I crimped both ends, checked with a cable tester that the wires are Ok and all seems fine.

When i plug in my home office i get between 80-90mb but using the same adapters but just with a shorter cable i get 500mb. What am i missing? Surely a 40m cable shouldn't reduce the speeds that much.

Any help on this would be appreciated.

Setup sky fiber broadband in Ireland
Hard wired with 1m cable i get 500mb
Hard wired with 40m cable i get 80mb
I'm using a macbook pro using fast.com to check the speeds.

Thanks,

 

[Sharpend]

Which Cat cable are you using?
Are the crimps a good quality variety?

 

[leonmoneill]

Which Cat cable are you using?
Are the crimps a good quality variety?

Thanks for the reply

Cable is external Cat5e
The crimps are brand new, Probably not the highest quality, i bought them in a set from a local electrical store. they seem solid though. When i plug both ends of the cable tester (the tester which shows if you mixed your cables over) all seems OK

 

[Richard-the-Ninth]

Have you any more of the same cable left?
If so make another lead and try that.
You may also want to try re crimping the leads. (Don't cut the plug off, just put it back in the crimper and squeeze again)
Try another brand of plug

 

[leonmoneill]

Thank you Richard the ninth and Sharpend for the excellent suggestions.

Ok i tried recrimping that didn't make any difference.
I made a new cable and get exactly the same speeds so thats really interesting.

Looking the cable up Cat5e, i shouldn't be the cable because we should be able to get 1000mb fro it, so maybe the crimps or plugs. The plugs fit fine and work fine would you think the crimps or plugs would have an impact on the speed?

 

[Sharpend]

What are you plugging into? It may be that one of the sockets is not so good?

 

[leonmoneill]

the sockets on both ends of the cable are fine, i checked a bringing my laptop cable with a new certified cable and i get full 500mb. To confirm one end of the cable is plugged into sky broadband box the other end in a USBC Cable Matters 1Gb adapter, the adapter and everything works fine if i bring the laptop beside the broadband box.

I have some old Cat6 lying around, i'm going to try to make a lead out of that to test if its the cable

 

[leonmoneill]

Ok so confirmed its not the cable, i found some old Cat6, made a lead and confirmed the speed is the same (91mb), So something is up with either the plugs or the crimps. I need to go back to the store and get new brands of plugs and crimps

 

[binky]

as a rough guide cat5 is suppossed to be good for around 100m.

 

[steptoe]

It's generally easier to have both ends of the cable into sockets rather than plugs,
Plugs can be notorious to crimp properly without an expensive crimp tool, the cheap ones don't seem to cut it too well.
Cat5e is good for gigabit speeds well over 200m

 

[SPECIAL LOCATION]

As another point to note...

A single cable on its own will generally be good to run at much faster speeds than multiple versions of the same spec cable all bunched together..

i.e.
The higher the frequency transmitted, (which are needed for higher speeds), the greater the electrical interference transmitted from the cable..
(e.g. it starts acting more like an aerial transmitter)
So any cables in close proximity to that cable are more likely to have issues of electrical interference..
reducing the signal quality and slowing the reliable data speeds between all the bunched cables due to re-transmission requests.
That is one of the differences of Cat5 / Cat6.. Cat6 is much better bunched up than Cat5...
But Cat5e is quoted as good for 1gig speeds.


Although your description suggests otherwise......
it sounds to me as though you have a component somewhere in the connection that is limited to 100Meg?

(We have fibre internet greater than 100Meg coming into our house...
But I do still have an old Toshiba Laptop that cannot cope with greater than 100Meg..
And I am also using some old routers around the house to boost the wifi signal...
one is good for 100meg+ another cannot cope above 50Meg!!)

 

[King Arthur]

'Cat' cables are defined within the ISO/TIA-568 standards and are basically higher performance versions of basic twisted pair telephony cables, which led to so-called 'structured cabling systems', designed to provide a highly flexible wiring systems for commercial environments. As home networks became more prevalent, 'Cat' cables became increasinglu used in the home. Over the years there have been lots of clever signal processing techniques developed and employed to increase the speed at which electrical signals can be transmitted over twisted-pair cables. When I started designing high-speed modems in the early 80s, 9600 bps over a phone line was cutting edge performance! But I digress.

Firstly, it might be helpful to read up about the various 'Cat' cables and their specified performance. There is loads of info online but this might help for a start: https://www.nmcabling.co.uk/data-cabling/data-cabling-faqs/ One thing to note is that the standards specify solid copper cables but there are inferior copper/aluminium cables onthe market which will have poorer performance. However, in most domestic applications cable runs are so short (compared to the 100m specification) that this is rarely an issue as it would likely be in a commercial installation. My home network cabling runs happily at gigabit speeds, though not all the connected devices can fully utilise this speed as it depends on things like the PC application, IP-camera resolutions or NAS performance. Some can, some can't, so it's important to be sure what is actually being measured.

The basic performance benchmark for a 'Cat' cable is based on a 100m link consisting of a 5m stranded patch lead + 90m solid cable + 5m stranded patch lead. In this 'maximum' configuration a Cat5e cable should provide a 100MHz link and a Cat6 cable a 250MHz link. Note that these are electrical signal characteristics and don't relate directly to data transmission speeds because that depends on other factors defined within the Ethernet standards. Thus '100BaseT' is an ethernet standard for a 100Mbps LAN over twisted-pair cabling, '1000BaseT for Gigabit LANs etc etc.

Also, as mentioned above, data throughput speed is not the same as cable transnission speed. Thus a link could be transmitting at gigabit speeds but interference and/or bad connections could introduce data errors that require re-transmission to correct. Without specialist test equipment it can be difficult to fully diagnose such effects.

 

[user 30718]

Be aware of the twisted pairs in the cable. Your tester will confirm you have connected the pins correctly (1 to1..... 8 to 8) but it won't confirm you have the twisted pairs correctly distributed. Please check diagrams, from memory, it Goes something like;
1 and 2 are a pair
3 and 6
4 and 5
7 and 8

If you get the pairing wrong - I would expect the results you are seeing.

Might be better asking about this on a computing forum.

 

[user 30718]

he

Quite right for cable standards but be careful about sockets vs plugs. You're probably talking about cheap common sockets that require punch down tools. Many 'nicer' sockets incorporate their own 'tool' which is also the socket back. Some of these are much worse than plugs to install.

In the latest patch panels I specify, plugs are required on the ends of the cables. The patch panels are more like an rj45 joiner. (In my opinion) these are far better because a single cable can easily be rewired without disturbing much else. I also like the cable coming straight from the connector, less bending required.

And on the copper coated aluminium cables, I only have one self righteous opinion - copper coated aluminium will never be allowed in any of my data or signal specifications (it is the work of Satan himself)

 

[UNG]

Using an external speedchecker IMO can be hit and miss when it comes to consistancy of the results within your internal / local network. I would suggest you do any testing within your internal / local network first to prove your cabling before running checks using external server sites that you cannot verify how much traffic that it is carrying at the time of the test

 

[leonmoneill]

Hello, thank you all for your feedback and help. I resolved the problem by terminating the cable into sockets as opposed to plugs. My understanding was that the external Cat 5 wasn't designed to be terminated into a plug, it worked fine but only got 90mb/s. My speeds at a full 500mb/s when terminated into sockets. I wonder if i had a better crimps would I of gotten the 500mb/s on a plug, i'm not sure. Anywhere its a better and cleaner job using the sockets so i'm happy with the result. Thanks again.

 

[Striker]

out of curiosity have you tried the industrial eth plug.

 

[leonmoneill]

out of curiosity have you tried the industrial eth plug.

Hrmm...I'm really not sure. The store i got them from only had one type. The plug doesn't look or feel very "Industrial", its a light clear plastic plug.

 

[Barx]

Are you connecting the cores in correct order? Following either T-568A or T-568B on both ends? Not following this layout can cause bad cross-talk.
Are you using pass through RJ45 connectors? I heard many times of these causing issues. Try closed end.
Are there any LEDs on whatever you are connecting cable to. These will usually indicate connection link speed. What do they seem to indicate? Also in the mac / pc operating system you should be able to see the link speed of the connection (not just a net speed test). It looks to me like it is only connecting at 100mb/s rather than 1000. This again could be related to the connectors used. All pins may be showing OK on your tester, but 1 or 2 may not be making good connection within the device port. I've had this before and a little wiggle side to side sometimes helps the springy fingers fall into place.