Is testing Cat5 cable using a testing device necessary if I have verified connectivity with a computer?

Question

I'm running some CAT5 cable around the house and this is the first time I've needed to build/terminate my own cables. I hooked up the cables to a computer and router and was able to reach the router configuration page with no trouble.

If it matters, the cable run is fairly long and I will be running Power-Over-Ethernet on that cable. I also tested with the POE device and everything seems to be in working order.

So all seems good, but I'm wondering if I also need to test with a handheld cable tester device. Is this device simply a way to test a cable without going to the trouble of hooking it up to an actual network, or does it test something that I really should be checking?

I'd hate to have to buy one of those devices if I don't need it since I don't really foresee the need to be running a lot of cable in the future. This is a one-off project.

So is it completely necessary in this situation to test using one of those devices?

Solution 1

These handheld devices are useful, since they test the cabling on the physical level and can help discover problems (cross-talk, wrong impedance, etc.) which are difficult to diagnose otherwise.

If you pulled some cable too much, or some turn inside your wall is too sharp, your network will "kinda work", but you won't get maximum throughput, or you'll get weird connectivity issues, etc.

It will be a pain to re-do stuff once you make everything neat and declare the job is done. Maybe you can lease such a device or hire someone who have it to test your network? These testers are prohibitively expensive to buy if you need it for just one job...

So, it is not strictly necessary to test with such tester, but it gives an extra assurance you did things right, especially if you're not a pro in network cabling.

Solution 2

Yes, you should test it!

The computer may not use all of the pairs... if you're testing with a 100Mbit rather than a Gigabit connection, for example, you're only using two pairs rather than all four. If you don't test this cable properly, it might check out fine on your computer but then fail when you try to use it with a feature like power-over-ethernet, or may only be able to work at 100mbit rather than gigabit speeds.

That said, I often neglect to test my cables. I don't always have a tester to hand, and the cables don't move around that much.

Solution 3

If the cable works, you're probably good.

The cable itself (we're talking copper, not glass here) is pretty foolproof -- hard to damage and unlikely to pick up interference from poor placement unless you do something really stupid.

The hard part is the connections. You do, of course, have to get the wire colors right, and you need to get all the wires connected (sometimes you may not have a wire pushed all the way into a RJ45 plug, eg, prior to crimping). Having the cable "work" gives you a good bit of confidence, but sometimes you can have wires crossed and some adapters will, well, "adapt", whereas others will choke on the bad wiring. Plus you may have currently unused wires improperly connected and they will fail if you ever use the unused wires (though ever using more than 4 of the 8 wires is unlikely outside of PoE, and many PoE schemes don't use more than 4).

But if you leave enough slack on both ends then you can always re-terminate a poorly terminated cable.

But note that you can purchase a simple RJ45 cable continuity tester fairly cheaply. These don't test for frequency response and cross-talk, but they check that the connections are correct end-to-end.

[I'll add that one likely novice mistake is to quickly glance at the color code chart and make the improper assumption that the RJ45 is wired pair/pair/pair/pair. In fact, there are pairs on each end, a pair in the middle, and then the last pair is split and "straddles" the middle pair.

Another novice mistake is to get confused by the several different color code "standards". There are three or four different ones, that put different colors on different pins. It really doesn't matter which standard you use, so long as you use the SAME STANDARD ON BOTH ENDS.

And it DOES matter that pairs are kept together -- where there is supposed to be a "pair", you should not use one wire from one pair and another wire from a different pair. While the cable may "check out" on the cheapie cable tester, there will be crosstalk and poor performance/high error rates in practice.]

Comments

[afrazier]

I'd consider testing particularly important since PoE is involved.

[JohnFx]

I did test it with a PoE device, so that should be good. Another factor is that one end of the cable is on top of a 50' pole and is very hard to access (without a bucket truck).

[henry700]

On the flip side: my employer (a very small small business) made some cables without following the spec (aA bB cC dD instead of aA bC cB dD). This worked OK for 6 ft cables, but anything longer and it was a crap shoot. The tester he has said they were OK, but in actual use, they failed. make sure you follow the standard pinout.

[haimg]

What tester was it? Any tester that certifies for CAT6e should have flagged these cables. But these testers cost $5K+, and cheap "continuity testers" mentioned in other answer cannot detect such cables of course.

[kinokijuf]

GbE uses all four wire pairs