Eathernet cabel quality tester.

[ZaXaZ]

Hi, I'm curious about the quality of some cables I have.
and a Fluke Networks CableIQ is WAY out of what I got to spend, is there any other tools that a normal person can afford to test network cables other than those cheap eBay tools that just show is you got a connection or not?

[cvanc]

I have one of these Klein cable testers and can highly recommend it.  It was under $100USD if I remember correctly (I got it a few years ago).

https://www.kleintools.com/catalog/cable-testers-accessories/vdv-scout-pro-2-tester-kit

[ZaXaZ]

I have never seen that brand in EU

probaly will end up beeing expensive to ship here with told and all that

but thx i will have a look at it

[LaurentR]

I have one of these Klein cable testers and can highly recommend it.  It was under $100USD if I remember correctly (I got it a few years ago).

https://www.kleintools.com/catalog/cable-testers-accessories/vdv-scout-pro-2-tester-kit

I can confirm that this model is nice, easy to use although it is expensive for what it does and there are much cheaper Chinese knock-offs.

This kind of tester (<$100) will not help you check the actual quality of the cable or its fitness for a particular application (e.g. 1000BaseTX).
For that, you need to go into the $1k-2k testers like the CableIQ you mention. Note that if you're ok to drop $500 on a tester, you can get CableIQ for about that much on EBay as they have been around for a long time. Resell value is also good.
I own a CableIQ and was surprised to fail some cables due low-quality connector assembly or even a female-female RJ-45 adapter that had a cross-talk issue...

There are a few devices trying to provide CableIQ-type services at a more reasonable price like this Byte Brothers device claiming to be a Certifier (same as CableIQ) for less than $300. I have no experience with it.
https://smile.amazon.com/Triplett-Brothers-RWC1000K2-Verifier-Printable/dp/B000J157WQ/

[Ranayna]

Yes, properly checking an Ethernet cable is expensive. The expensive testers will inject test signals at various frequencies and will then measure attenuation, crosstalk, resistance, capacitance and various other parameters.
That is not something a cheap tester can do.

A poor-mans test might simply be to test if the cable works.

Take two computers with gigabit ethernet. For this I would recommend Linux as operating system, booting a live system is enough. Preferably use machines that use good network adapters. Intel adapters are generally considered best.
Install iperf or net-io.
Connect the two machines with a known good cable and set up proper networking.
Run tests in both directions to get a baseline.

Then you can do the same tests on your suspect cables.

Generally, in my experience, ethernet cables are very robust. The main weakness lies in the connectors. The latch is generally the thing that breaks and makes the cables useless. Sliding across a sharp edge also causes obvious damage. Sometimes we have cables with visible corrosion on the contacts. A lot of cables that look ugly, with visible kinks and sometimes damage to the outer sheath still work fine. All that is for S/FTP cables. UTP Cables are a lot less resistant against damage, and that is the main reason we avoid using them, even though they are fine for gigabit Ethernet.

For permanent cabling, the connections at the patch panel or the jack are what fails, but that is very rare. Out of several thousand RJ45 ports at my workplace, we had to repair maybe 5 in the last several years. Other than that, all repairs we had to do were due to someone cutting the cable.

[rjp]

A poor-mans test might simply be to test if the cable works.
 

2  machines with 2 good working ports  and timing a large transfer will give the most accurate results of all, for any given setup

[ogden]

2  machines with 2 good working ports  and timing a large transfer will give the most accurate results of all, for any given setup

Right. It's like measuring mains AC voltage using bulb. If it glows - you quite accurately detected that voltage is there 

[rjp]

2  machines with 2 good working ports  and timing a large transfer will give the most accurate results of all, for any given setup

Right. It's like measuring mains AC voltage using bulb. If it glows - you quite accurately detected that voltage is there 

aww thats not fair, you do get the bits per second thing on the dialog or the time taken via the console

my point was that for that network, the cable either worked at the expected transfer rate or it didnt.

[Jeroen3]

Most network cards can do some tests.

Note: cable is 30 cm.

[ogden]

my point was that for that network, the cable either worked at the expected transfer rate or it didnt.

Such test using two Ethernet devices/computers is only marginally better compared to cheap eBay wiring checker. Your test computer pair can "pass all the tests" on non-compliant cable while other devices fail to operate (get frame/FCS errors). Ethernet does not send digital "1" and "0" bits over cable. There's different transformers, PHY's, varying transmit signal strength, receiver sensitivity, crosstalk, interference and not only that - lot of things that can be off or close to margin treshold so one device works but another does not.

That's why cable certifiers are needed - so you measure cable and pass/fail compare it against standard ensuring that any compliant Ethernet device will operate over such cable.

[Ranayna]

Ogden has a good point. Different devices can cause different results. That is why I labeled this a "poor mans test".

I was not aware that the Intel driver has those capabilities. I wonder if it is possible to get the raw data that the test produces. But that huge discrepancy in the length test does not really inspire a lot of confidence in those tests...
Was your computer attached to a switch at time of testing? Maybe a point to point connection between two Intel cards gives better results.

@OP: If this is only for a few cables, have you considered renting a proper cable tester?

[mikeselectricstuff]

If you're wanting to weed out poor manufacture as opposed to failures over time, then as the main way to cheap out on cables is to reduce copper content, a simple DC resistance test should do most of the job.
Pair-to-pair capacitance should be able to detect cat5 that's masquerading as cat6, though you can probably just feel it to detect the lack of the seperator core.

[mikeselectricstuff]

Yes, properly checking an Ethernet cable is expensive. The expensive testers will inject test signals at various frequencies and will then measure attenuation, crosstalk, resistance, capacitance and various other parameters.
That is not something a cheap tester can do.

A poor-mans test might simply be to test if the cable works.

Take two computers with gigabit ethernet. For this I would recommend Linux as operating system, booting a live system is enough. Preferably use machines that use good network adapters. Intel adapters are generally considered best.
Install iperf or net-io.
Connect the two machines with a known good cable and set up proper networking.
Run tests in both directions to get a baseline.

Then you can do the same tests on your suspect cables.

That will only give a yes/no answer. A more qualitative approach may be to add some deliberate degredation - e.g. add 50m of known cable, to see if you still have enough margin when the test cable is added.
You could probably make a simple analogue network with pots etc. to control various common degredation modes, so you can measure how much degredation is tolerated with a particular cable.