Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

THD-cadence simulation

Status
Not open for further replies.
J

jingxuelu

Junior Member level 3
Junior Member level 3
Joined
May 16, 2008
Messages
26
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,283
Activity points
1,454
frequency spectrum simulation cadence

Does anyone know how to simulate THD in circuit design environment of cadence ic6.1 version?

how to right the thd() expression with calculator?
 

cadence thd

I am not yet exposed to ic6.x.

What is done 5.x is: do an DFT of the output waveform which will give you the frequency spectrum of the output waveform. From the frequency spectrum you will be able to find the amplitudes of fundamental, harmonics and other noise components. There from you can calculate whatever you need.

sankudey
 

    J

    jingxuelu

    Points: 2
    Helpful Answer Positive Rating
thd cadence

No, I don´t recommend to do a DFT.
In this case you get the spectrum (roughly calculated !) including the start period at t=0. But that´s not what you want.

Therefore: Perform a TRAN analysis and measure as exact as possible the fundamental frequency fo. Then do a second TRAN analysis and command in addition "FOURIER" (just a click to enable) and write into the box the value of fo.

After the analysis the results of the first 9 harmonics (based on fo) are written down in the output file - including the calculated value of the THD !
 

    J

    jingxuelu

    Points: 2
    Helpful Answer Positive Rating
cadence calculator thd

after simulation, select a wave , and then select thd in the calculator, then you can print the value of thd
 

    J

    jingxuelu

    Points: 2
    Helpful Answer Positive Rating
cadence 2nd order harmonic

hi

i am attaching a doc on thd simulations

a very good reference..detailed step by step procedure given along with explanations

shweta
 
fourier thd

LvW said:
No, I don´t recommend to do a DFT.
In this case you get the spectrum (roughly calculated !) including the start period at t=0. But that´s not what you want.

Therefore: Perform a TRAN analysis and measure as exact as possible the fundamental frequency fo. Then do a second TRAN analysis and command in addition "FOURIER" (just a click to enable) and write into the box the value of fo.

After the analysis the results of the first 9 harmonics (based on fo) are written down in the output file - including the calculated value of the THD !
Click to expand...

Any thing wrong with DFT?
How does it differ by DFT (as you told 'roughly') and FOURIER? Aren't the algo's same?
Any way, you do DFT on a waveform which is the output of the TRAN.

I am really confused about the difference and their accuracy level.

sankudey
 

    J

    jingxuelu

    Points: 2
    Helpful Answer Positive Rating
simulated thd

No, certainly there is nothing "wrong" with the DFT.
However, if you like to calculate the THD the FOURIER expansion is the logical way because it directly follows the definition of THD. And this definition is based on a periodic waveform.

Now, if you command FOURIER/THD the simulation program considers only the last period which then is continued in order to create a real periodic waveform.
This is in contrast to the DFT, which transforms the whole signal including start-up signal changes, and so on....
This results in a corresponding spectrum, but that´s not what is needed for the THD.
Hope I could express myself clearly.
 
  • Like
Reactions: garm

    G

    garm

    Points: 2
    Helpful Answer Positive Rating
simulated thd vs frequency

Hi, shw, san, Lvw
thanks for your help, let me try one
 
cadence values for sin

I simulated THD using the last period, and ic6.x gave me a value of 9.4e-1 for 1vpp 500k sine output. so what is the unit of thd, 9.4e-1 is too large, but i cann't find very big harmonic distortion from dft simulation
 

discrete fourier transform thd

yes,use trans the use THD function in caculate
 

thd calculation using cadence

try to simulate an ideal wave to find how the thd software is written
 

4-40 thd

I also agree that there is nothing to be blamed for in using fourier analysis. Make sure to use the appropriate windows (triangular, hamming....etc..).

Alternatively, you could use the PSS analysis that gives you also the THD extracted from "main result".
 
way to caculate thd

i tried rectangular and hamming windown for 500k ideal sine wave (one cycle), rectangular windown can give me right result, but hamming window will give me wrong result(has big 2nd order harmonic), is there anyone who can explain this?
 

simulate thd with cadence

jingxuelu said:
I simulated THD using the last period, and ic6.x gave me a value of 9.4e-1 for 1vpp 500k sine output. so what is the unit of thd, 9.4e-1 is too large, but i cann't find very big harmonic distortion from dft simulation
Click to expand...

The THD is defined as a ratio between amplitudes and, therefore, it has "no unit".
Normally, the THD is given as an absolute number or in %.
 

thd of rectangle wave

No, I don´t recommend to do a DFT.
In this case you get the spectrum (roughly calculated !) including the start period at t=0. But that´s not what you want.

Therefore: Perform a TRAN analysis and measure as exact as possible the fundamental frequency fo. Then do a second TRAN analysis and command in addition "FOURIER" (just a click to enable) and write into the box the value of fo.

After the analysis the results of the first 9 harmonics (based on fo) are written down in the output file - including the calculated value of the THD !
Click to expand...

Hello LvW,

I use IC5x,
I don't find the "FOURIER" command in the the TRAN analysis toolbox,
Please can you help me, where I can find it.

Thanks.
 

how many periods are used to measure thd

In amplifiers, if the input signal frequency is low enough, I would suggest avoid transient sim (+DFT) or flex-balance (just like HB) in Spectre(RF).

The easiest way is to use DC simulation, and through VerilogA function and calculator to find the Fourier series coefficients. You can obtain much more insight in DC sim and avoid possible settling time issue from transient sim (or PSS shooting method), or strong nonlinearity limit from flex-balance.

However, if it's a high frequency signal (i.e. pass the frequency of your first pole), you have to resort to either transient or HB. Because this time you are facing Volterra instead of Taylor.
 

Hi guys,

I'm getting a large value at DC after doing the DFT. My fundamental freq is at 5.8GHz and i'm getting a resonance at DC with value way larger than at the fundamental freq.
Could you help me on this?
Thanks!

- - - Updated - - -

fredd said:
Hi guys,

I'm getting a large value at DC after doing the DFT. My fundamental freq is at 5.8GHz and i'm getting a resonance at DC with value way larger than at the fundamental freq.
Could you help me on this?
Thanks!
Click to expand...

I have found the culprit.
My output sine wave has DC value/offset. That's why i have values at freq=0Hz after DFT.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top