| Author |
Message |
myem
Joined: 12 Dec 2003
Posts: 132
|
 25 Dec 2003 4:34 simulate 50-ohm transmission line |
|
|
|
dear all ~
how to know a 50-ohm transmiion line is good or not good ???
when I simulate 50-ohm transmission line with EM simulator~~
what kinds of charts should I observe ???
smith chart or cartesian ????
|
|
| Back to top |
|
 |
yuppi77
Joined: 18 Dec 2003
Posts: 33
|
| 25 Dec 2003 7:27 |
|
|
|
Hi myem,
When simulating a 50-ohm transmission line, the important thing to observe is would be the s-parameters.
You could view it either with the smith chart or the cartesian.
|
|
| Back to top |
|
|
caocao
Joined: 23 Nov 2003
Posts: 117
Helped: 3
|
| 26 Dec 2003 15:05 Re: simulate 50-ohm transmission line |
|
|
|
HI myem
I think when you use transmission line which character impedance is 50ohm ,you should konw the load impedance that is 50ohm or not .Then you can estimate the transmission line is good or not in Smith chart.
| myem wrote: |
dear all ~
how to know a 50-ohm transmiion line is good or not good ???
|
|
|
| Back to top |
|
|
activewei
Joined: 19 Sep 2002
Posts: 127
Location: UK
|
| 12 Jan 2004 5:26 |
|
|
|
| In fact, EM simulation of 50 ohm line is just a rough estimation (of course it is more accurate than just S-parameter simulation). At higher frequency (>10 GHz), the computation will still have some difference between simulation and measurement. Smith Chart is the best way of knowing how close your line closed to 50 ohms.
|
|
| Back to top |
|
|
jupitorcuu
Joined: 24 Apr 2003
Posts: 84
Location: chinaBJ
|
| 14 Jan 2004 7:54 |
|
|
|
| There is no perfect 50Ohm in the world, return loss under -25dB is OK!
|
|
| Back to top |
|
|
Scholl
Joined: 22 Oct 2003
Posts: 6
|
| 02 Mar 2004 11:39 |
|
|
|
| smith chart is the best.
|
|
| Back to top |
|
|
Element_115
Joined: 23 Mar 2002
Posts: 304
Helped: 20
|
| 02 Mar 2004 23:46 Re: simulate 50-ohm transmission line |
|
|
|
Just put a plot of S11 (Should be the same as S22) and S21.
S11 should be lower that ~ -35dB and S21 ~<0dB.
Assuming you have a 50 ohm 1 wave length line at your
desired Freq.
|
|
| Back to top |
|
|
flyhigh
Joined: 12 Nov 2003
Posts: 507
Helped: 25
Location: somewhere over the rainbow
|
| 03 Mar 2004 21:53 Re: simulate 50-ohm transmission line |
|
|
|
Hi,
just to join this interesting discussion with a somewhat changed point of view. When simulating in EM simulator and ploting S parameters on a Smith chart, you will for sure see what simulator "thinks" is a 50 Ohm line, this might not neccessarily be the 50 Ohm in practice for a number of reasons.
First, as stated above, deffinition od characteristic impedance is not unique even in theory. Different simulators use different methodologies for computing Zo so the results might be different. Difference can increase when accounting finite metal thickness. Furthermore, user can contribute to an error by uncareful definition of ports, meshing or other computational parameters. Also, some amount of structure idealization considering uniform line width and absence of surface roughness are neccessary to introduce in simulations. Finally, EM simulators are just the tool, not the absolute true!
To completely disapoint everybody, not even measurements are a good proof that any line has 50Ohm impedance, it is for certain a proof for pure TEM lines, for hybrid wave lines this is a mater of luck and chance.
flyhigh
|
|
| Back to top |
|
|
