how to calculate THD ??

This ought to work - I've no idea what you've done wrong. It should work - I did this:

 thd(VT("/out") 0 2.0/VAR("frq") 128 VAR("frq"))

Note that I had a design variable called frq, which is what I was sweeping. The transient was run for 2*frq (actually not that easy to do in IC5141, I think, but I was running from a netlist), and the fundamental was frq.

Maybe you should contact customer support?

Andrew

For calculating THD in ADE I found the equation as:

THD=sqrt(sum(f^(abs(Ai)^2),i=1:infty))/abs(Af)*100%

where `f' is the fundamental frequency and Ai are the complex coefficients.  I attached a picture below; I'm confused about the notation; typically I see THD calculated as:

THD=sqrt(sum((abs(Ai)^2),i=1:infty))/abs(Af)*100%

Dear Colub,

Cdub said:

I'm confused about the notation; typically I see THD calculated as:

THD=sqrt(sum((abs(Ai)^2),i=1:infty))/abs(Af)*100%

I believe the expression and your understanding are both the same. Reading the definition provided by Cadence f^Ao is indeed the amplitude of the fundamental (it calls it the DC component), and F^Ai are the amplitudes of the higher ith components. Hence the numerator represents the sum of the powers of the higher harmonics ( note f^Ao is specifically excluded from the summation). Do you agree Colub?

Shawn

Thank You for your reply, I am still a bit confused; if "F^Ai are the amplitudes of the higher ith components" then the denominator of the expression should be F^Af right? Have I misinterpreted your reply?

Cdub said:

I am still a bit confused; if "F^Ai are the amplitudes of the higher ith components" then the denominator of the expression should be F^Af right? Have I misinterpreted your reply?

Dear Cdub,

Referring to the text in the manual:

"The computation uses the dft function. Assume that the dft function returns complex coefficients A0, A1…, Af , ... .

Please note that fundamental frequency f is the frequency contributing to the largest power in the signal. A0 is the complex coefficient for the DC component and Ai is the complex coefficient for the ith harmonic where . Then, total harmonic distortion is computed as:"

 Hence the notation in the denominator, Af, is consistent with the frequency contributing to the largest power in the signal. Typically, this will be the amplitude of your fundamental.

Is this any more clear (sorry about my prior confusion!)?

Shawn

Hi Shawn! Thanks again for the reply. It seems like the notation of the numerator is not consistent with the notation of the denominator. Is there something I might be missing?

Thanks Again!

Cdub said:

It seems like the notation of the numerator is not consistent with the notation of the denominator. Is there something I might be missing?

Dear Cdub,

I agree - the description appears inconsistent to me too. The ocean syntax, however, makes it reasonably clear to me anyway...

Good luck!

Shawn

Cdub said:

It seems like the notation of the numerator is not consistent with the notation of the denominator. Is there something I might be missing?

Dear Cdub,

I agree - the description appears inconsistent to me too. The ocean syntax, however, makes it reasonably clear to me anyway...

Good luck!

Shawn

Ah The Ocean reference is better, I guess there is an error in the "Virtuoso Visualization and Analysis XL User Guide". I attached the ocean reference image below, the f is part of the index, makes much more sense: