# THD using pss analysis for a fully differential amplifier

1. ## THD using pss analysis for a fully differential amplifier

I want to find the THD of fully differential amplifier. I am using cadence virtuoso. I have found three options in simulating the THD in cadence which are:

1. using PSS analysis---- I select pss analysis then got ---direct plot --- forum forum then select voltage--- I place no. of harmonics i.e. 10 (I have kept)---fundamental frequency---select net as--differential and then select the two output nodes of my differential amplifier---It shows it in the form of 10 vertical straight lines because I have set the no. of harmonic as 10. I am taking the difference between the first harmonic and third harmonic which shows 58 dB.

2. Second option is I am doing transient analysis. On this transient response of the fully differential output I apply DFT fusing calculator and then take 20 dB---- I again get vertical lines as harmonics this time I get 53 dB which is the difference between first and third harmonic.

3. Third option is I have done transient analysis----apply THD using calculator ---take 20 dB--- This time it doesn't show waveform, it shows in form of numbers -10 dB, which I think is wrong.

I havent placed any port in my circuit The only source is the input sources of differential amplifiers. Which one is correct? Can anyone help.

i ALSO want to calculate SNR. There is no SNR option in calculator? Anyother way to find SNR?

•

2. ## Re: THD using pss analysis for a fully differential amplifier

Can you understand a definition of THD, SNR, SINAD ?
THD is not ratio of funadmental-component and 3rd-order-harmonic.

Don't rely on Direct Plot Form in Cadence ADE blindly.
Use skill functions directly with understanding THD definition surely.

Originally Posted by ICdesignerbeginner
1. using PSS analysis
............................................
It shows it in the form of 10 vertical straight lines because I have set the no. of harmonic as 10.
I am taking the difference between the first harmonic and third harmonic which shows 58 dB.
If you use Shooting-Newton-PSS, no. of harmonic as 10 is enough.
However, if you use HB-PSS, it is too few.

Originally Posted by ICdesignerbeginner
2. Second option is I am doing transient analysis.
On this transient response of the fully differential output I apply DFT fusing calculator
............................................
I again get vertical lines as harmonics this time I get 53 dB which is the difference between first and third harmonic.
Originally Posted by ICdesignerbeginner
3. Third option is I have done transient analysis
----apply THD using calculator ---take 20 dB---
This time it doesn't show waveform, it shows in form of numbers -10 dB, which I think is wrong.
Don't use function without understanding its definition.

"thd()" returns value as "%".

So 20*log10(10**(-10/20)/100)=-50dB

Originally Posted by ICdesignerbeginner
I havent placed any port in my circuit
Why do you require "port" ?

Originally Posted by ICdesignerbeginner
The only source is the input sources of differential amplifiers.
Do you surely set common mode bias ?
Can you understand the followings ?

There are two modes regarding input drive, common mode and differential mode.

Also there are two modes regarding output, common mode and differential mode.

Originally Posted by ICdesignerbeginner
Which one is correct?
Any of three could be correct, as far as you can understand THD correctly and you can do simulation correctly.

Originally Posted by ICdesignerbeginner
i ALSO want to calculate SNR.
There is no SNR option in calculator?
Anyother way to find SNR?
Simply calculate SNR based on its definition.

Here noise power can be evaluated by PSS/Pnoise.
Distortion's power can be evaluated by PSS.

If you would like to use Transient Analysis, you have to use Transient-Noise Analysis.

Noise power which is evaluated by conventional Transient Analysis is no more than numerical noise not physical noise.

•

3. ## Re: THD using pss analysis for a fully differential amplifier

Originally Posted by pancho_hideboo
Can you understand a definition of THD, SNR, SINAD ?
THD is not ratio of funadmental-component and 3rd-order-harmonic.
I am not taking ratio I am taking difference between first and third harmonic. I have seen paper showing difference between third and first harmonic.

Originally Posted by pancho_hideboo
Don't rely on Direct Plot Form in Cadence ADE blindly.
Use skill functions directly with understanding THD definition surely.

If you use Shooting-Newton-PSS, no. of harmonic as 10 is enough.
However, if you use HB-PSS, it is too few.
In cadence version which I am using dosent have shooting-Newton and HB-PSS option so I am just using PSS analysis

Originally Posted by pancho_hideboo
Don't use function without understanding its definition.

"thd()" returns value as "%".

So 20*log10(10**(-10/20)/100)=-50dB
Using calculator THD gives in % but I have taken 20 log with it so its in dB. and thats coming as -10
Please correct me if I am wrong.

Why do you require "port" ?

Originally Posted by pancho_hideboo
Do you surely set common mode bias ?
Can you understand the followings ?

There are two modes regarding input drive, common mode and differential mode.

Also there are two modes regarding output, common mode and differential mode.
yes I have placed the common mode and differential mode.

Originally Posted by pancho_hideboo
Any of three could be correct, as far as you can understand THD correctly and you can do simulation correctly.

Simply calculate SNR based on its definition.

Here noise power can be evaluated by PSS/Pnoise.
Distortion's power can be evaluated by PSS.

If you would like to use Transient Analysis, you have to use Transient-Noise Analysis.

Noise power which is evaluated by conventional Transient Analysis is no more than numerical noise not physical noise.
How can I use Distortion power can you help me in This? I want to calculate input referred noise?

To get SNR , I divide my input signal with noise power/input referred noise and get SNR?

Is this noise power and spectral noise density the same?

what does this 0_waveform/n_thd/nil means? what is /nil?

•

4. ## Re: THD using pss analysis for a fully differential amplifier

Post thread after learning very basic things surely.

Originally Posted by ICdesignerbeginner
I am not taking ratio I am taking difference between first and third harmonic.
You can not undetstand very basic thing at all.

Difference of dB value is ratio of linear value.

Again, can you understand THD surely ?

Originally Posted by ICdesignerbeginner
I have seen paper showing difference between third and first harmonic.
Surely see "paper".

Originally Posted by ICdesignerbeginner
In cadence version which I am using dosent have shooting-Newton and HB-PSS option so I am just using PSS analysis
No, you can not understand anything at all.
Surely see GUI of PSS.
Or see "spectre -h pss".

Originally Posted by ICdesignerbeginner
Using calculator THD gives in % but I have taken 20 log with it so its in dB. and thats coming as -10
Please correct me if I am wrong.
Are you kindergarten pupil ?
Surely see my append.

Originally Posted by ICdesignerbeginner
How can I use Distortion power can you help me in This?
If you don't include higher order harmonic distortion as noise, you don't evaluate higher order harmonic distortion power.
SNR is defined without distortion in attached figure.

Simply, S/N or S/(N+D).

Originally Posted by ICdesignerbeginner
I want to calculate input referred noise?
I can not understand what you want to mean at all.

Originally Posted by ICdesignerbeginner
To get SNR , I divide my input signal with noise power/input referred noise and get SNR?
No.
However your input level is linear opeartion region, it is close to true SNR.

Originally Posted by ICdesignerbeginner
Is this noise power and spectral noise density the same?
No.

--[[ ]]--