🚀 go-pugleaf

In article <1992Oct9.201011.15028@cbfsb.cb.att.com> feg@cbnewsb.cb.att.com (forrest.e.gehrke) writes:

>In article <1992Oct8.120225.6247@fys.ruu.nl> vreeburg@fys.ruu.nl (Jurriaan Vreeburg) writes:
>>
>>>Put two pieces of 50 Ohm cable in parallel. It should be exactly the same.
>>>Gary
>>
>>After reading Gary's answer I had some doubt of its correctness. It implies
>>that by putting two 50 Ohm coax cables in series, one is able to create a 100 Ohm
>>coax cable, which is certainly not true.
>>
>Gary didn't say anything about connecting the two pieces in series.

>>I think that putting two cables in parallel doesn't change the impedance, since
>>the impedance is defined as:
>>
>>                   L
>>Z = square_root ( --- )
>>                   C
>>L: inductance per unit length
>>C: capacitance per unit length
>>
>>What happens if we put two cables in parallel: both the inductance and the capacitance
>>increase by a factor of two -> Z doesn't change!!! Putting two cables in series doesn't
>>change the impedance as well, since L & C are defined per unit length, which don't
>>change at all in that case.
>>
>You had better rethink the above statements.  Since when does paralleling
>two inductances double the inductance?

>Forrest Gehrke feg\@dodger.att.com

I have used both series and parallel lines extensively - both principles 
work. Parallel lines are simple enough ; if you use coaxial cable you just 
combine inners and braids. You can even combine different Zo cable and get 
the expected parallel characteristic impedance.

Series lines are obtained in practice with coaxial cable by laying the 
cables side by side and joining the braids of both at input and output. The 
signal and load is then connected between the inners and the cable will 
behave with characteristic impedance which is the sum of the two individual 
cable impedances. Very easy to measure with Network Analyzer or impedance 
bridge by measureing the reactance of a cable under OC and SC termination. 
The square root of the product is equal to the Zo. Choose a cable length 
which is approximately eighth wavelength at measurement frequency so that 
the inductive and capacitive reactances are more or less equal (to avoid 
multiplication of a small with a large number).

Andre Fourie