Twisted pair cable capacitance - How to best measure it

[AndersJ]

I am measuring the capacitance between the two wires of a twisted pair cable using 3 different methods:

1. Older HP LCR meter
2. New Agilent LCR meter
2. Pulse generator, RC and oscilloscope to determine the time from the pulse leading edge to 63.2% of full amplitude.

The cable length is 10 meters.
I believe I do not have excessive cabling or stray capacitances.
The oscilloscope probe adds 75ns, that I subtract from the oscilloscope reading.

For all 3 methods the result is consistent, 49 pF/m
The cable manufacturer insists that the capacitance is 40 pF/m.

Am I doing this measurement correctly?

[Whales]

The cable manufacturer insists that the capacitance is 40 pF/m.

What's their tolerance on that spec?  +/-50%, 20%, 10%, 5%, 1%?

What test frequency are they using for it?

If they don't know the answers to those questions then it's very difficult to verify their claims.

Also worth investigating given this is twisted pair: do they spec whether or not you put axial tension on the cable or not?  That will probably make the capacitance go up.

[AndersJ]

Tolerance:  +5%, -10%
Frequency: Not specified. When I measure, f makes little or no difference.
Tension:     Not specified. My cable is "coiled". Would C change if I lay it out on the floor?

[jonpaul]

Cable UTP, TP, part no, mfg, url pdf spec ?

Measure freq, guarded?

pulsr, probe...never heard of it

j

[AndersJ]

This is a CAN cable,
BUT,
please do not worry about detailed cable specifications.
This topic is meant to adress general concerns when measuring cable capacitance.

[PartialDischarge]

A CAN cable has a shield, how are you connecting it? that is going to affect the measurement!

[mag_therm]

With 10m, it is feasible to measure length accurately,
Then use vna to obtain s11 and tdr  . Measure 3 times with end
short, open, and with Z0.
Those secondary parameters can give primary parameters.

[David Hess]

I have used an LCR meter, impedance bridge, multimeter charge/discharge capacitance measurement, and RC oscilloscope measurement, and they all returned the same and correct result for cable capacitance.  I suspect you found the tolerance of the cable.

[jonpaul]

See excellent Belden specs and papers of Steve Lampen, re CAN, cable capacity and twisted pairs UTP, TP.

Most cable mfg very carefully control twist and Cper foot

j

[AndersJ]

PartialDischarge: The shield is disconnected. Connecting it to the Guard terminal lowers C by an order pf magnitude.
David Hess:        Your past experiences suggest I may be doing it right. Thank you.
jonpaul:             Thank you. I will try to find the documentation you suggest.

[mag_therm]

Production units were 100 percent measured by ATE at end of line.
That was a long time ago when I was assisting with ATE maintenance.
Not sure about now, whether all is measured or just samples.