Hello all, I`m trying to come up with a couple of equations to determine the characteristics of residential networks.

Basically, what I`ve got so far is that the bandwidth is determined by the interval in which PSACR (power sum attenuation to cross talk ratio) has a positive value.

I do realize that bends in the cable will modify the impedance and thus modify the twisted pair characteristics.So, what I am looking for are the equations (or at least some help in coming up with them) that would give me length and bend (corner) dependencies (for example, if I have a 130m link that has 4 corners, my link will have a bandwidth of X, attenuation of Y and max data rate of Z).

Thanks!

I don't know the equation you ask about or even if one exits but what I do know is that you would not use Cat6 over 100 meters, and in some circumstance as lot less. The attenuation would be so great as to make communications unreliable. Bear in mind that stranded cable has a greater attenuation than solid core cable.

You wont be able to calculate and precise data rate, only a theoretical maximum which is the bandwidth, because data rate is dependant on factors outside of the cable.

Stuart

Thanks Stuart for the reply.

You CAN go over 100m, but it depends on a number of factors (for example if the cable has no bends and no outside interference). The 100m limit is for the worst case scenarios.

Obviously the data rate depends on the terminals capabilities, but I`m looking for the theoretical rate the the cable can hold.Thanks again.

Virtually everybody installing cable just follows industry standards without doing these calculations. I doubt that any regular contributor to this forum has ever done such calculations. If you need formula you could try an Internet search or contact a cable manufacturer.

Hi, I have been researching this on the internet for a couple of weeks now.

I`ve found several equations for capacitance, inductance, impedance, NEXT, ATTENUATION and other characteristics for twisted pair cables, but they are to complex or do not depend on the length of the cable.

Basically, the impedance varies when the cable is bent thus modifying the attenuation.Thanks

Unless you are working on designing a new cable spec (or if this is for a college course assignment), you're just wasting your time. I would need to pull out my old books, but if I remember correctly, the sharpest bend radius is 4 times the diameter (or maybe it's circumference) of the cable. If you go sharper than that, you start running into trouble. While it is possible to go well beyond the maximum 100 meter distance, it really shouldn't be done. To our surprise, this last year at a PGA tournament our company hosted, we had a 5,000 ft run of cat 6 that "worked" for a 1.5Mb dsl connection for a limited number of users, but I certainly wouldn't recommend it.

Just stay within the industry standard specs and if you need to go over the 100 meters for a single run, then use fiber.

I see what you`re getting at :) I thought that the 4 times radius formula was a mechanical consideration...

I would like to find out at least what happens if I go over 100m.

For example if I have a link where I need just a couple more meters and end up with a 110m long channel, if that section of my network will still be within spec.Thanks again!

Obviously if you go 110 meters it's not with the specs, but it should work, although I would not expect it to work at the upper end of the bandwidth i.e., gigabyte speeds. 100Mb is iffy at that distance but 10Mb is almost guaranteed to be ok assuming you don't have any kinks in the cable. As a real life example, one of our stores had a few stations (on a mini 10/100 switch) with intermittent connectivity issues and only had a stable connection after forcing 10Mb half duplex. After a little troubleshooting. I discovered that the contractor had pulled a 365ft run for the switch uplink. We pulled a new run (about 200ft) from a closer wiring closet and was then able to have good connection at 100Mb full duplex.

Ask Your Question

Weekly Poll

Would you ride in a self-driving car from Tesla?

Discuss in The Lounge

Poll History