Efficiently Test FTP Cat 5 Cable: Tips & Tools for Accurate Results

  • Thread starter Thread starter soltice_9
  • Start date Start date
  • Tags Tags
    Cable Testing
AI Thread Summary
Testing the 15,000 meters of Foil Twisted Pair cable for Cat 5 compatibility involves using equipment like a 20/20 TDR and an HP/Agilent 4191A RF Impedance Analyzer. The cable's construction includes two insulated conductors and a drain wire, with the user noting difficulties in crimping due to its 20-gauge size. Initial capacitance measurements showed values significantly higher than Cat 5 specifications, raising concerns about its suitability for data communications. The discussion also highlighted the importance of proper testing methods, including measuring capacitance between the insulated conductors while isolating the foil shield, to ensure accurate results. Overall, the cable's characteristics and testing methods are critical for determining its viability for the intended use.
soltice_9
Messages
16
Reaction score
0
Hi,

I've just bought 15,000 meters of Foil Twisted Pair cable (2 PVC insulated conductors and a Drain wire with Aluminum Foil around them) and I'm trying figure out if it will be good enough for Cat 5.

How should I test it without having to chop it down to smaller pieces?

The test equipments that I have are: a 20/20 TDR (Time Domain Reflectometers), an HP/Agilent 4191A RF Impedance Analyzer, and an Agilent WIRESCOPE 350 Cable Certifier.

Thanks
 
Engineering news on Phys.org
Welcome to PhysicsForums!

I was going to remark that Cat-5 (and Cat-5e) contain 4 sets of twisted-pair cabling, but upon re-reading, it appears as if you actually have foil-sheathed pairs, and not a plastic/rubber jacketed pair. If it's 22-24 gauge copper, and has the right capacitance, it'd probably be okay:
http://en.wikipedia.org/wiki/Category_5_cable#Characteristics

I'd suspect that the bigger problem with having 15km of this stuff is the likelihood that none of your pairs will be coloured coded (nevermind properly colour coded). That might be fine if you're using this internally, and can do connectivity testing, and maybe if you're making patch cables, but that's probably going to be a no-go for anything else.

Note that Cat-5, 5e and 6 don't require foil shielding, and rely on being twisted pair for their noise immunity. However, it is an option for higher-noise environments:
http://en.wikipedia.org/wiki/Foiled_twisted_pair#Cable_shielding
 
Thank you MATLABdude!

I bought only 1 set of twisted-pair. It has 2 insulated conductor and 1 drain (no insulation) with a foil around the three.

It appears to be 20 gauge! This thing is bigger than the average Cat-5. So I've been unsuccessful in crimping it into the RJ45 connectors I currently have. I'm going to shop around to see if there are bigger RJ45 connectors out there. I may have to resort to soldering ...

It is color coded and I'm going to do a capacitance per unit length like you suggest, but I expect it will be higher than that of 22 AWG type. Also, I'm not sure I can do a connectivity testing right now. Further, I'm going to do a SWR test using my 20/20 TDR to make sure there's no "kink".

I'm not sure what else I can do to test it further without having to chop it down (because I may have to send it back).

Thanks
 
Most likely the capacitance per unit length will be above spec, so the Zo will be below spec. It takes special construction to make STP that meats Cat-5 specs for capacitance (thicker insulation, for example).

Why did you get STP? It's not needed in most cases.
 
If you are planning on using that for (reliable) data communications you will have to meet these specifications:

http://en.wikipedia.org/wiki/Category_5_cable

Also keep in mind for gigabit communications you will require to meet a category 5e spec and the cable must contain 4 pairs. If you have the agilent wirescope 350 that is all you need to certify that cable. If you have to "resort to soldering" you will probably throw the cable out of specification because of the twist close to the connector will not be tight enough and you will have excessive crosstalk.

good luck.
 
I finally get to measure the capacitance between the insulated conductor and the drain (no insulation). The capacitance decreases with increasing frequency (1 Mhz increasing to 250 Mhz). I'm using an Impedance Analyzer.

Is this inline with expectations?

Oh, we're using STP because it's what's the customer requested.

Thanks
 
soltice_9 said:
I finally get to measure the capacitance between the insulated conductor and the drain (no insulation). The capacitance decreases with increasing frequency (1 Mhz increasing to 250 Mhz). I'm using an Impedance Analyzer.

Is this inline with expectations?

Oh, we're using STP because it's what's the customer requested.

Thanks

The capacitance is not changing with frequency, the losses are. What do you get for the capacitance per foot at 1MHz?
 
I have to confess that the test was done on another cable that is not Cat 5. It was just a trial test. However, it consists of Foil Twisted Pairs like the Cat 5 I bought.

I got like 430 pF over 1.2 meter at 1 MHz.

So, what exactly are these losses? Why do I get like 45 pF over 1.2 meter at 250 MHz on the same cable?

Sorry it's been over 6 years since I graduated from EEng ... my memory is kinda hazy ... Please help me remember!

Thanks
 
Also, when someone ask for the Capacitance/meter ... does he/she have to specify the frequency too?
 
  • #10
soltice_9 said:
I have to confess that the test was done on another cable that is not Cat 5. It was just a trial test. However, it consists of Foil Twisted Pairs like the Cat 5 I bought.

I got like 430 pF over 1.2 meter at 1 MHz.

Holy smokes! Cat-5 spec is like 16pF/foot. Your number is like 132pF/ft. That foil must be wound good and tight!

That's way out of spec for a 100 Ohm data grade cable. What Zo does that give you at 1MHz on your analyzer? Probably tiny!

soltice_9 said:
So, what exactly are these losses? Why do I get like 45 pF over 1.2 meter at 250 MHz on the same cable?

Sorry it's been over 6 years since I graduated from EEng ... my memory is kinda hazy ... Please help me remember!

Thanks

The high frequency losses are in the dielectric, and those losses are frequency dependent. If you look at the loss specs for Cat-5 (or any data grade cable), it will be given over several frequency bands.
 
  • #11
soltice_9 said:
Also, when someone ask for the Capacitance/meter ... does he/she have to specify the frequency too?

I think that it's better to think of it in terms of the full equation for the Characteristic Impedance:

http://en.wikipedia.org/wiki/Characteristic_impedance

The loss terms R and G are frequency dependent. The L and C should not be, since they are determined by the geometry. If the dielectric constant of the insulation changes with frequency, that would change the C value over frequency.
 
  • #12
Thank you Berkeman.

I going to take a look at my setup again. I must be doing something wrong for the machine to tell me that C varies with frequency.
 
  • #13
soltice_9 said:
Thank you Berkeman.

I going to take a look at my setup again. I must be doing something wrong for the machine to tell me that C varies with frequency.

Does it give you all 4 numbers for the Zo equation? If not, it may be assuming a lossless model.
 
  • #14
Also, how long of a piece of cable are you testing, and how is it connected to the machine?
 
  • #15
Hi Berkeman,

It is a very old Impedance Analyzer.

Although I didn't try it out, but the machine can tell me |Z|, |G|, L, and C.

And it is a 1.2 meter cable.

The test rig consists of 2 metal posts ... so I connected the Insulated Conductor to the 1st metal post, and the Drain (Conductor without insulation) to the 2nd ... [making sure on the other end of my cable ... the same Insulated Conductor and the same Drain are not touching] ... Then I asked the machine to tell me C ... at 1 MHz ... then 50 MHz ... up to 250 MHz ... Again, it keeps on decreasing!

Thanks again for your help. So what is the best way to figure out C/meter ... just in case I cannot get this machine to work properly?
 
  • #16
soltice_9 said:
Hi Berkeman,

It is a very old Impedance Analyzer.

Although I didn't try it out, but the machine can tell me |Z|, |G|, L, and C.

And it is a 1.2 meter cable.

The test rig consists of 2 metal posts ... so I connected the Insulated Conductor to the 1st metal post, and the Drain (Conductor without insulation) to the 2nd ... [making sure on the other end of my cable ... the same Insulated Conductor and the same Drain are not touching] ... Then I asked the machine to tell me C ... at 1 MHz ... then 50 MHz ... up to 250 MHz ... Again, it keeps on decreasing!

Thanks again for your help. So what is the best way to figure out C/meter ... just in case I cannot get this machine to work properly?

Wait, I thought it was foil insulated twisted pair. But you're saying it's a single conductor with a foil "ground" conductor?

You should be measuring the capacitance between the two conductors in the twisted pair, with the foil shield floating.
 
  • #17
I twisted the 2 Insulated Conductors together ... and hook them up to the 1st metal post ... and the foil to the 2nd metal post ...

Ok, tomorrow ... I'll hook the 1st Insulated Conductor to the 1st metal post... and the 2nd Insulated Conductor to the 2nd metal post ... and don't do anything with the foil/drain ( keep it isolated from the test)
 
  • #18
But my guess is that ... the machine will still tell me that C decreases with increasing frequency ... which is not correct.
 
  • #19
soltice_9 said:
But my guess is that ... the machine will still tell me that C decreases with increasing frequency ... which is not correct.

Agreed on both counts. Do you have some Cat-5 cable that you can test too? It would be good to see the comparison on that machine.
 
  • #20
BTW, did you say which Impedance Analyzer you are using?
 
  • #21
It's ... HP 4191a impedance analyzer

I'll proceed to do Cat5 cable testing tomorrow on the machine.

Thank you for your help!
 
  • #22
Oh I found something from the Manual.

These are the equations that the machine use to calculate C:

1 / (2*pi*f*X)

X / [2*pi*f *(R^2 + X^2)]

So C decrease with frequency ... according to the machine... I'm going to go research on these 2 equations ... Thanks again Berkeman
 
  • #23
C should remain constant in both equations

If f increases by factor of 2, X decreases by factor of 2 ... so C remains constant.

Also, R is negligible in comparison to X at high frequency so it can be ignore... so C remains constant.

Oh, I forgot to mention that all of the capacitance value that the machine gave me has a negative (-) sign in front.

The manual explains: " If the test frequency exceeds the resonant frequency point, the measured inductance & capacitance of sample becomes a negative value. In such frequency region, the residual inductance or stray capacitance dominates the actual measurement value"

So, this is inline with expectation because cable capacitance is by definition stray capacitance. http://en.wikipedia.org/wiki/Capacitance
 

Hot Threads

Recent Insights

Back
Top