1. LTspice
  2. Теми

Can LTspice show Cross Over distortion of a class B type amplifiers...

 

If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
 
Any thoughts on this would certainly be appreciated..... Thanks....
 

1.Use.TRAN analysis, not .AC, because that doesn't show non-linearity.

2. Check that the current through the output devices is very small with no input signal.

On 2024-12-30 10:36, dlwarr911 via groups.io wrote:
If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
 
Any thoughts on this would certainly be appreciated..... Thanks....
-- 
OOO - Own Opinions Only
Best Wishes
John Woodgate
Keep trying

Virus-free.www.avg.com
 

...and apply a very low input signal when running the transient test.

Le 30/12/2024 à 11:43, John Woodgate a écrit :

1.Use.TRAN analysis, not .AC, because that doesn't show non-linearity.

2. Check that the current through the output devices is very small with no input signal.

On 2024-12-30 10:36, dlwarr911 via groups.io wrote:
If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
 
Any thoughts on this would certainly be appreciated..... Thanks....
-- 
OOO - Own Opinions Only
Best Wishes
John Woodgate
Keep trying

Virus-free.www.avg.com

 

You can also do this with a .DC analysis. With this it's also very easy to do a "wingspread" plot, which plots the incremental gain, which is very revealing.

--
Regards,
Tony


On 30/12/2024 11:43, John Woodgate wrote:

1.Use.TRAN analysis, not .AC, because that doesn't show non-linearity.

2. Check that the current through the output devices is very small with no input signal.

On 2024-12-30 10:36, dlwarr911 via groups.io wrote:
If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
 
Any thoughts on this would certainly be appreciated..... Thanks....
-- 
OOO - Own Opinions Only
Best Wishes
John Woodgate
Keep trying
 

On Mon, Dec 30, 2024 at 05:41 AM, <dlwarr911@...> wrote:

Can LTSpice show Cross Over distortion of a class B type amplifiers...

Yes.

If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
To "see" it, as a distorted-looking sinewave, you need to make the distortion bad enough so that it stands out.  Visually looking at sinewaves, they generally need to be pretty bad for humans to see it with our eyes.
 
Use a .TRAN analysis.
 
It might help to disable LTspice's waveform compression (".options plotwinsize=0") so that it does not gloss over small levels of distortion that are hard to see.
 
You could add a .FOUR to print out the harmonic levels (in the .LOG file).  Or run an FFT.  They will help you see the distortion effects (as harmonic or IM distortion) even if the sinewave looks pretty good visually.
 
In other words, do the same thing you would do if you had an amplifier on a bench in front of you.  Hooked up to a scope, you might or might not see the distortion unless you look hard.  Or use the right test equipment (e.g., distortion analyzer).
 
You could in principle painstakingly compare the output with an undistorted sinewave, but getting its amplitude and phase just right could be challenging.
 
Andy
 
 

If your amp is in a high gain negative feedback loop, it can be harder to see the crossover distortion because the feedback look "eliminates" most of it.  Getting rid of that feedback would make it easier to see.  Higher frequencies might show it more because there is less open-loop gain.
 
Did you ask this question because you tried a simulation and couldn't see the distortion?  Or were you just wondering whether simulations show it?  Simulations do indeed show crossover distortion because they simulate what the amplifier actually does in real life with real hardware.  The degree to which you can see it depends on many things - not only in simulation but in real hardware.
 
Andy
 
 

I have uploaded an example schematic featuring the Quad 303 output stage.

--
Regards,
Tony


On 30/12/2024 11:56, Tony Casey wrote:

You can also do this with a .DC analysis. With this it's also very easy to do a "wingspread" plot, which plots the incremental gain, which is very revealing.

--
Regards,
Tony


On 30/12/2024 11:43, John Woodgate wrote:

1.Use.TRAN analysis, not .AC, because that doesn't show non-linearity.

2. Check that the current through the output devices is very small with no input signal.

On 2024-12-30 10:36, dlwarr911 via groups.io wrote:
If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
 
Any thoughts on this would certainly be appreciated..... Thanks....
-- 
OOO - Own Opinions Only
Best Wishes
John Woodgate
Keep trying
 

To evaluate and quantize crossover distortion you can use a notch filter (RC bridge) to eliminate the fundamental frequency out of the output voltage.

On Mon, Dec 30, 2024 at 1:15 PM Tony Casey via groups.io <tony=ritecom.com@groups.io> wrote:
I have uploaded an example schematic featuring the Quad 303 output stage.

--
Regards,
Tony


On 30/12/2024 11:56, Tony Casey wrote:
You can also do this with a .DC analysis. With this it's also very easy to do a "wingspread" plot, which plots the incremental gain, which is very revealing.

--
Regards,
Tony


On 30/12/2024 11:43, John Woodgate wrote:

1.Use.TRAN analysis, not .AC, because that doesn't show non-linearity.

2. Check that the current through the output devices is very small with no input signal.

On 2024-12-30 10:36, dlwarr911 via groups.io wrote:
If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
 
Any thoughts on this would certainly be appreciated..... Thanks....
-- 
OOO - Own Opinions Only
Best Wishes
John Woodgate
Keep trying
 

Yes, I know. I have already uploaded such a schematic example using one a few years go. It doesn't explicitly show crossover distortion, though - just the residual.

--
Regards,
Tony


On 30/12/2024 16:46, aledelaplaza via groups.io wrote:

To evaluate and quantize crossover distortion you can use a notch filter (RC bridge) to eliminate the fundamental frequency out of the output voltage.

On Mon, Dec 30, 2024 at 1:15 PM Tony Casey via groups.io <tony=ritecom.com@groups.io> wrote:
I have uploaded an example schematic featuring the Quad 303 output stage.

--
Regards,
Tony


On 30/12/2024 11:56, Tony Casey wrote:
You can also do this with a .DC analysis. With this it's also very easy to do a "wingspread" plot, which plots the incremental gain, which is very revealing.

--
Regards,
Tony


On 30/12/2024 11:43, John Woodgate wrote:

1.Use.TRAN analysis, not .AC, because that doesn't show non-linearity.

2. Check that the current through the output devices is very small with no input signal.

On 2024-12-30 10:36, dlwarr911 via groups.io wrote:
If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
 
Any thoughts on this would certainly be appreciated..... Thanks....
-- 
OOO - Own Opinions Only
Best Wishes
John Woodgate
Keep trying
 

In order to tune crossover distortion, the best is visualizing linearity. Distortion percentage doesn't indicate what to adjust.

Le 30/12/2024 à 16:46, aledelaplaza via groups.io a écrit :

To evaluate and quantize crossover distortion you can use a notch filter (RC bridge) to eliminate the fundamental frequency out of the output voltage.

On Mon, Dec 30, 2024 at 1:15 PM Tony Casey via groups.io <tony=ritecom.com@groups.io> wrote:
I have uploaded an example schematic featuring the Quad 303 output stage.

--
Regards,
Tony


On 30/12/2024 11:56, Tony Casey wrote:
You can also do this with a .DC analysis. With this it's also very easy to do a "wingspread" plot, which plots the incremental gain, which is very revealing.

--
Regards,
Tony


On 30/12/2024 11:43, John Woodgate wrote:

1.Use.TRAN analysis, not .AC, because that doesn't show non-linearity.

2. Check that the current through the output devices is very small with no input signal.

On 2024-12-30 10:36, dlwarr911 via groups.io wrote:
If I build a Class B amplifier (all have cross over distortion) What do I have to do, to see cross over in a trace...
All I get is a perfect looking sign wave... What would be the analysis and steps to do this...??
 
Any thoughts on this would certainly be appreciated..... Thanks....
-- 
OOO - Own Opinions Only
Best Wishes
John Woodgate
Keep trying

1 - 10 з 10