Can we adjust W and L for txn line circuit?

[kellyza]

Hi all...

When I converted my lumped element circuit into txn line circuit, the output obtained was not as optimum as I got in the lumped element circuit. It is very hard to get the same or more or less the same results. What I did is I adjusted the W and L of the txn line and it works. I really not sure why do we need those adjustment. Can we simply do that? and why? :???:

Thanks

 

[vfone]

The conversion from lumped to distributed circuits is not as simple. Is not enough to replace a lumped component with an equivalent microstrip one.
Check the theory for Richards Transformation and Kuroda Identities.

https://www.ittc.ku.edu/~jstiles/723/...rmation .pdf
https://www.ittc.ku.edu/~jstiles/723/...ntities .pdf

 

[kellyza]

Hi vfone,

Thanks for reply... I actually simply converted from lumped to distributed circuit using txn line calculator by taking into account the properties of the board only. Then I ran the simulation accordingly.

Thanks

---------- Post added at 02:39 ---------- Previous post was at 02:37 ----------

I can't open the link posted in the thread.

 

[vfone]

You know, while ago I've tried the same thing using a dedicated simulation filter software (which have a direct conversion from lumped to distributed) and I was not happy with the results. The final circuit needs a lot of tuning to get decent values and dimensions for components.

https://www.ittc.ku.edu/~jstiles/723/handouts/Richards Tranformation .pdf
https://www.ittc.ku.edu/~jstiles/723/handouts/Kurodas Identities .pdf

 

[kellyza]

Yes I agree and it is now happen to me... I need to tune most of the dimensions to get what I want. Then what did you do to solve it? Btw thanks for the links.

 

[vfone]

Nothing else than hard tuning, following basics of the theory.
Anyway, do not expect the final dimension of the circuit to be the same (or even closer) if you want to keep the performances.
good luck

 

[kellyza]

Thank you very much... I really appreciate it. Mind me asking you other question. What should I consider if I couldn't get the same output when the circuit simulated by EM simulator? We should get more or less the same result right?

Thanks in advance.

 

[vfone]

The choice of each component depends mainly on the types of particular circuit, the acceptable losses or Q factors, the power handling, and the operating frequency.
Have to keep in mind few things as, a λ/4 (or a λ/2) microstrip resonator have more than one resonance, and also remember that the losses in distributed circuit arise due to a number of mechanisms, which are different than lumped circuits, for example dielectric substrate and radiation losses.
Anyway, what you are doing is fine, using an EM simulator, instead of a linear one.

 

[kellyza]

Thanks vfone for the explanation. As far as I'm in the right track, I think everything is fine. I'll try to minimize all the losses as low as I could. Please share with me and others if you have more knowledge in txn line design circuit. I'm quite new in this area and I am willing to learn and gain more knowledge and experience. For your info I'm using MWO as my simulation tool. Thanks again for your kindness.

 

[kellyza]

Hi vfone... also Hi to all...

May be you could give me some comments on this thread

https://www.edaboard.com/threads/189716/

Thanks.

 

[kellyza]

Hi all,

I always believe that there are a light at the end of a tunnel. I will find my ways out no matter how. Hoping that somebody out there can give me the clue and guide me to reach at the end of the tunnel.

Thanks
-Kellyza-