Hi,

Now I'm trying LCR parameter fitting to S-parameter given.

This technique is probably widely used in simulator world.

See attached workspace.

Reference S-parameter is Taiyo-Yuden's SMT capacitor in Genesys lib.

And I made equivalent circuit for this SMT cap (L-C-R serial resonator) and tried to make LCR fit to S-parameter using Genesy Optimization.

This should be easy problem. When I tried in ADS, ADS quickly gave the answer on it, on the other hand, Genesys couldn't solve it and continued optimization routine eternaly.

Is it Genesys optimizer's limit or my Genesys has someting wrong?

Please advise me.

Thanks in advance.

Ume,

[file]fitting_example.wsx[/file]

Now I'm trying LCR parameter fitting to S-parameter given.

This technique is probably widely used in simulator world.

See attached workspace.

Reference S-parameter is Taiyo-Yuden's SMT capacitor in Genesys lib.

And I made equivalent circuit for this SMT cap (L-C-R serial resonator) and tried to make LCR fit to S-parameter using Genesy Optimization.

This should be easy problem. When I tried in ADS, ADS quickly gave the answer on it, on the other hand, Genesys couldn't solve it and continued optimization routine eternaly.

Is it Genesys optimizer's limit or my Genesys has someting wrong?

Please advise me.

Thanks in advance.

Ume,

[file]fitting_example.wsx[/file]

[1] Your equations specify:

S11diff=mag(Linear1_Data.S[1,1])-mag(Linear1_Data.S[3,3])

S21diff=mag(Linear1_Data.S[2,1])-mag(Linear1_Data.S[4,3])

Note that in the equation variable view S11diff and S21diff are real values. Because you're using mag in the equation the difference is a real number and when you try to optimize im(S11diff) it gets nothing.

[2] You're optimizing real and complex parts of S11diff and S22diff separately. I think you want the magnitude to be minimum (as per your equations).

So, I set the equations to:

S11diff=Linear1_Data.S[1,1]-Linear1_Data.S[3,3]

S21diff=Linear1_Data.S[2,1]-Linear1_Data.S[4,3]

to retain the complex data. Then I reduced to two optimization goals, minimizing mag(S11diff) and mag(S21diff).

The optimizer takes 27 quick rounds to hit a minimum with my severely de-tuned starting point.

Note that because the optimizer never reaches a perfect fit and you have no minimum error specified it will run forever, but the best answer is reached in about 5 seconds on my computer.

Mark

[file]fitting_examplea.wsx[/file]