The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> How to do a distortion Analysis of a Differential Amplifier?
https://designers-guide.org/forum/YaBB.pl?num=1244804507

Message started by saket on Jun 12^th, 2009, 4:01am

Title: How to do a distortion Analysis of a Differential Amplifier?
Post by saket on Jun 12^th, 2009, 4:01am

I am looking for ways to reduce the third harmonic from the output of a differential amplifier. Are there any papers detailing distortion analysis? any good books which relate to this topic that anyone can recommend?

Any help will be deeply appreciated!

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by buddypoor on Jun 15^th, 2009, 7:02am

The simplest method is certainly to reduce the level of the input signal because THD gets larger when the output voltage swing is large.
LvW

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by sos on Jun 23^rd, 2009, 4:55am

A possibly more useful response...

Distortion can occur throughout the circuit, but it is likely to be either at the input or at the output. Improving distortion first requires deciding where the dominant source is.

Odd harmonic distortion is often caused by output load currents. Change the load impedance and see what happens.

Output distortion is often dependent on output stage bias current. Try experimenting with the bias.

Input distortion in a feedback amplifier implies that you have too much error voltage applied to the input stage, given its construction. Higher bias current can help, you can also improve distortion by degenerating the input, but these also change open loop gain.

In a feedback amplifier, higher open loop gain will usually decrease distortion whether the source is at the input or output. The trick, of course, is how to increase the open loop gain and retain stability.

Steve

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by raja.cedt on Jun 23^rd, 2009, 7:55am

hi,
1.how can you say that odd harmonics depends on output load current
2.Increasing loop gain will reduce distortion but what i saw in many cases it is limited by many other parameters like PM,settling time....

Thanks,
Rajasekhar.

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by pancho_hideboo on Jun 23^rd, 2009, 9:56am

raja.cedt wrote on Jun 23^rd, 2009, 7:55am:

1. how can you say that odd harmonics depends on output load current

This changes a load line for output transistor, so it is quite natural that odd harmonics depends on output load current.

sos wrote on Jun 23^rd, 2009, 4:55am:

Odd harmonic distortion is often caused by output load currents.
Change the load impedance and see what happens.
Output distortion is often dependent on output stage bias current.
Try experimenting with the bias.

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by Berti on Jun 24^th, 2009, 6:20am

Quote:

raja.cedt wrote on Yesterday at 7:55am:
1. how can you say that odd harmonics depends on output load current

This changes a load line for output transistor, so it is quite natural that odd harmonics depends on output load current.

I am not convinced that this statement is generally true. Saket didn't give any details. What is the load current in a SC circuit?

I agree with most of the answers but I wouldn't recommend to try improving the distortion by trial-and-error. Take a look at the following paper: G. Palumbo and S. Pennisi, High-frequency harmonic distortion in feedback amplifiers, IEEE Trans. Circuits and Systems, March, 2003.

They provide a mathematical framework to do distortion analysis without the need for complicated Volterra series. Maybe you gain more insight this way.

Good luck!

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by pancho_hideboo on Jun 24^th, 2009, 6:25am

Berti wrote on Jun 24^th, 2009, 6:20am:

I am not convinced that this statement is generally true. Saket didn't give any details. What is the load current in a SC circuit?

Who refers to SC circuits in this thread ?
And even for SC circuits, we can define load line and current.

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by raja.cedt on Jun 24^th, 2009, 6:26am

hi,
lets say my opamp is driving a voltage buffer, does it mean opamp output don't have odd harmonics? (i am asking this based on the previous ans ...load current will create odd harmonics , in this scenario opamp is not driving any load current )

Thanks,
Rajasekhar.

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by pancho_hideboo on Jun 24^th, 2009, 6:31am

raja.cedt wrote on Jun 24^th, 2009, 6:26am:

lets say my opamp is driving a voltage buffer, does it mean opamp output don't have odd harmonics?

Output impedance of OPAmp is not zero actually.

And apart from final output of OPAmp, there are loads in voltage gain stages in every OPAmp.

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by aaron_do on Jun 24^th, 2009, 7:10am

Hi all,

I think you can look at a basic single stage CMOS amplifier as a V-I converter (diff pair for example) followed by a I-V converter (the load + output resistance). Both the V-I and the I-V converters will be nonlinear and will create odd harmonics.

cheers,
Aaron

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by pancho_hideboo on Jun 24^th, 2009, 7:16am

aaron_do wrote on Jun 24^th, 2009, 7:10am:

Both the V-I and the I-V converters will be nonlinear and will create odd harmonics.

I think odd harmonics of V-I converters is dominated by input nonlinearity not output.

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by sos on Jun 24^th, 2009, 8:54am

raja.cedt wrote on Jun 23^rd, 2009, 7:55am:

hi,
1.how can you say that odd harmonics depends on output load current

I'm assuming that the output stage has push and pull capability. If a signal distorts similarly for positive and negative signals, that shows up in the spectrum as odd harmonics.

2.Increasing loop gain will reduce distortion but what i saw in many cases it is limited by many other parameters like PM,settling time....

Increasing loop gain doesn't always work, but sometimes it is easy to try.

Thanks,
Rajasekhar.

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by sos on Jun 24^th, 2009, 8:56am

raja.cedt wrote on Jun 24^th, 2009, 6:26am:

My suggesting output load current assumed you had a load. If there is no load, other sources of distortion are at fault.

Steve

Title: Re: How to do a distortion Analysis of a Differential Amplifier?
Post by Berti on Jun 24^th, 2009, 11:28pm

I think we are talking all about more or less the same. But people are providing different explanations since Saket didn't specify whether "amplifier" refers to an OTA or an OpAmp, and its is not clear whether he drives resistive or capacitive loads/feedback or open-loop etc.

Maybe Saket can clarify that point.?

Cheers