[SOLVED] transmission line calculator querry

[sohkha]

Hi all,
Transmission line calculators in ADS and AWR are valid over what frequency range? and Can i have formula for finding length width and spacing for coupled microstrip line?

thanks

 

[vlsitechnology]

I am sorry I don't know about this, If anyone know please help him out.

 

[vfone]

Theoretically there are no frequency limits, but practically the limitations are the physically dimensions of the structures.

 

[sohkha]

THANKS for reply. Well i am trying to get UWB frequency response. so if i change my substrate hight and microstrip thickness to get the allowable physical dimensions, it will distort my response. Any idea how to manage width and spacing greater the 250 microns while my response is not changed.

 

[tony_lth]

What is your UWB freq bandwidth?
250 microns is too easy even for China Vendor. Since you are in UK, I think 100 microns should be used.

 

[volker_muehlhaus]

so if i change my substrate hight and microstrip thickness to get the allowable physical dimensions, it will distort my response.

Thinner substrate -> smaller line width, smaller gap width
Thicker substrate -> bigger width, bigger gap width

Smaller epsilon -> longer lines/resonators
Bigger epsilon -> shorter lines/resonators

 

[sohkha]

My bandwidth is 7.5 GHz with center frequency at 6.85 GHz. We have facility to design 250 micron and above. I have found these conditions with ADS line calc.
30000/Height(mm) > Freq(MHz)
18 > Er > 1
10 > Width/Height > 0.1
10 > Spacing/Height > 0.1
metal thickness < 0.1*Height and < 0.2*Spacing.

 

[tony_lth]

you can try RO4350B substrate, 0.5oz metal thickness, 10mils height.

---------- Post added at 07:55 ---------- Previous post was at 07:54 ----------

or 20mils height.

 

[edajeff]

Those calculators are usually based on fixed formulas and only give you a rough estimate. I would suggest using a dedicated transmission line field solver like SEMSTRA **broken link removed**. It gives you much more accurate result in just a few seconds. You can also generate a circuit model for your transmission line.

 

[volker_muehlhaus]

Those calculators are usually based on fixed formulas and only give you a rough estimate.

I don't agree.

The calculators usually use closed form equations, whichs should be more accurate than the numerical solution of a discretized EM model. The cross sectional EM solvers are more versatile and can do many different configurations, but they are not more accurate.

 

[edajeff]

I don't agree.

The calculators usually use closed form equations, whichs should be more accurate than the numerical solution of a discretized EM model. The cross sectional EM solvers are more versatile and can do many different configurations, but they are not more accurate.

Those closed form equations are constructed either from analytical solutions or numerical solutions.

For the former case, the derivation of analytical solutions always involves some assumptions, which can have significant discrepancy from the real world case, such as dielectric material inhomogeneity and frequency dependence, conductor surface roughness, and ground plane size. The equations constructed from that at most have similar accuracy.

For the latter case, needless to say, if the numerical solutions are not accurate, the resulting closed form equations will not be accurate either. What’s important is how close its modeling is to the real world case.