700Ft cat5 run - no problems & running 1Gbps!

tankman1989

Active Member
Reaction score
5
I've always thought that 100m was the maximum length for cat5. I needed to run 700ft and was planning on an extender in the middle but thought I would try it out without the extender and found that I'm not having any problems and the connection speed is rating at 1Gbps. It is all standard networking equipment (Dell Latitude notebook) and a Netgear 1Gbps switch.

Has anyone else ever tried this out and found it to work also? Am I going to find some issues down the road with dropped/collided packets or something? I thought one of the reasons for the length limit is because of the twisted pairs creates a difference in transmission speeds between the pairs.

Anyone experience this or similar? Any thoughts?
 
I've come across plenty of setups that exceeded specs...and still "worked".
BUT....worked how well? If you push things...does it start to crumble? Sure...you can plug a computer in at the other end...and it may say you're connected at 1x gig. And perhaps you can surf the 'net fine, or browse shares on the server. But what if you start hammering that connection? Crank up a database, do large file tranfers, I'm sure you'll start seeing the wheels fall off if you do any serious analyzing of that traffic (in other words..more than just a link light and the Windows connection status). How many times per year might that quickbooks database become corrupted now?
 
I've come across plenty of setups that exceeded specs...and still "worked".
BUT....worked how well? If you push things...does it start to crumble? Sure...you can plug a computer in at the other end...and it may say you're connected at 1x gig. And perhaps you can surf the 'net fine, or browse shares on the server. But what if you start hammering that connection? Crank up a database, do large file tranfers, I'm sure you'll start seeing the wheels fall off if you do any serious analyzing of that traffic (in other words..more than just a link light and the Windows connection status). How many times per year might that quickbooks database become corrupted now?

It was streaming video nicely but I doubt it was pushing more than 10Mbps though. This is plenty for the application so that is all that matters for me. If I needed sustained 1Gbps, I would be more uptight about it.
 
Just because windows shows it connected at 1Gbps doesn't mean it is actually running at that. In fact at 700ft I can guarantee you it isn't running at that speed. The 100m limit doesn't mean it won't work over that distance it means the connection begins to become unstable and the throughput speed goes down the more you go past that distance.
 
I would have either thrown a few repeaters in or done it with fiber. Have you run a speedtest on it yet? That should give you a rough idea as to the stability.
 
Is this for a client, or for your own use? If it's for your own use, experiment away. I wouldn't use it for any production use, but hey, it's your call. If it's for a client, this is the very definition of substandard work.

Personal use. It's just to feed data from a monitoring post, so very low data flow - probably 5MB/day.

I'll do some real testing on this with a variety of files - doing data transfer tests. I'll see how large files vary vs smaller files and smaller amounts of data.
 
Last edited:
Personal use. It's just to feed data from a monitoring post, so very low data flow - probably 5MB/day.

I'll do some real testing on this with a variety of files - doing data transfer tests. I'll see how large files vary vs smaller files and smaller amounts of data.

Was this run underground?
 
I thought one of the reasons for the length limit is because of the twisted pairs creates a difference in transmission speeds between the pairs.

Anyone experience this or similar? Any thoughts?

The 328 foot, 100 meter for my overseas and Northern peers, length restriction has to do with the timing and reliability of sending and receiving data frames over copper wire.

You are correct about UTP cabling. Each pair has a different amount of twist per foot to help with NEXT and other "noise." So the 4 pairs are all different lengths.

Ethernet utilizes a contention protocol where stations wanting to transmit don't have to wait their turn like a Token Ring network. They "listen" and if they "hear" no activity the station transmits. Now if two stations listen, hear nothing, then transmit a frame at about the same time there is a collision. This happens because of propogation delay...the time it takes the signal to move down the cable between stations. The data is corrupted and the stations jam, use an algorithm to determine how long to back off and then resend. This is normal behavior and a correctly designed, heavily used cabling plant should have no more than 5% collisions and a 40% utilization rate.

Since you only have one station 700 feet away and there are no other nodes competing for the Ethernet's bandwidth there isn't a lot data loss due to collisions. With twisted pair voltages averaging from -1 to +2.5 Volts you will have measurable signal attenuation and impedance losses over a 700' span. You are also throwing Ethernets timing way off which results in excessive data loss and retransmissions. 700' of UTP cable will also pickup more "noise" than its designed to handle.

There are very specific rules for segment length, number of nodes per segment etc. For any professional work follow these to the letter for a happy and harmonious cabling plant. For yourself you can push the envelope.
 
Was this run underground?


Yes in conduit. Just so everyone knows, the reason I did this was to test how functional a setup like this would be if I needed to use it in a pinch sometime - so I would know if I had to run a cable 500-1,000 ft, what my chances are of having successful transmission. I would never do this professionally unless the owner specifically asked for this and knew the risks.
 
Last edited:
My longest run was 1200', above ground, on a 10Gbps net. Command said no on repeaters, but they at least let me run it separately from the High Voltage lines. It was a run that connected 2 separate networks. Actual throughput speed was <500Mbps, constant transfer and connection errors.

I learned, what I already knew, that a single run over the 100m limit is not worth it. I let the run speak for itself when people were wondering why the connection kept dropping or transfers just suddenly stopped because the net kept getting jumbled information. After a while, they let me do a Site to Site RF connection, as I got tired of having to splice fiber and Ethernet that got cut to let vehicles go through.
 
Back
Top