Measuring physical length of transmisson line

I have a network analyser which can measure electrical length, delay time and phase of the "DUT" in addition to all S parameters.But it does not have the option to measure the physical length of the transmition lines.I know some analysers can tell it directly.
Now:
1. Since we know the speed of the signal which goes on the line (300.000 Km/sn X Velocity Factor) and the delay time i think we sould be able to calculate the physical length from the formula :
Way=speed x time
Is this a correct way of thinking?
Any idea on measuring physical length?

2. I also wanto see a HF antenna's gain vs frequancy.What kind of referance antenna should i use?Is there any wide band antenna that has the same response exactly for the whole 2-30 MHz on the market?
Thanks for any help...

it is more accurate to use RF software tools to do electrical->physical, since it will include all other effects (fringing fields...etc) into consideration.
I prefer using LinCALC in A_D_S, but you can download a free line-calculation utility from microwave office website I think (as far as I remember)..
also any RF simulation package will include such an utility, so check what tool you are using...

The AWR free transmission line program http://www.appwave.com/products/ (go to bottom of page and click the tx line title) is probably the best free program available. It calculates line loss of configruations that others, such as the free appcad program, do not do.

Rogers has has a free one here http://www.rogers-corp.com/mwu/mwi_java/user.htm that does more parameters, but on fewer line types.

Hi nec3,

according to above posts I recommend you using a software to calculate the physical length of a transmission line. Your formula is right in principal but it does not take er (epsilon r) in account. The dielectric constant affects the velocity of a wave on your board. The velocity will also be changed by housing, losses, ... --> so use a software!

For measuring an antenna gain you need a reflection free chamber (a little bit expensive). As a reference antenna you should use an isotropic radiator (theoretical, doesn't exist). YOu can also use a broadband dipole or something like this and take the antenna gain in account.

bye

Yes, in first approximation

Delay (seconds) = (length * sqrt(er)) / c

Anyway I think it's better to measure directly the physical length of the trasmission line.

Try to use configuration of two identical antennas separated by R in deaf chamber. Then:

abs(s21)^2=G^2×lambda^2×(1-abs(gANT)^2)^2/(4×pi×R)^2

where s21 is measured by network analyser, R is distance, lambda is wavelength, gANT is refl. coeff. of antenna and G is test antenna gain