Can a class-A amplifier always be considered as a linear amplifier in RF frequencies?

A class-A amplifier can always be considered as a linear amplifier in RF frequencies?

Re: Class-A is linear?

neazoi said:

A class-A amplifier can always be considered as a linear amplifier in RF frequencies?

Click to expand...

No, of course not "always". Linearity (THD) depends on amplitudes.

Re: Class-A is linear?

So if I see a clean non-distorted sinewave on my oscilloscope out of the amplifier, it does not mean that the amplifier is suitable for ssb operation that requires linear amplifiers??

Note, I am talking about RF amplifiers.

Re: Class-A is linear?

neazoi said:

So if I see a clean non-distorted sinewave on my oscilloscope out of the amplifier, it does not mean that the amplifier is suitable for ssb operation that requires linear amplifiers??

Click to expand...

Are you really able to decide if a signal on the screen is "clean and non-distorted"?
Strictly speaking, no amplifier is 100% linear.
You always must accept a certain degree of non-linearity (measured as THD).

Re: Class-A is linear?

Thank you very much,

So, Is there any way to check if a class-A RF amplifier is linear or not? (accepting a degree of non linearity) ?

Re: Class-A is linear?

neazoi said:

Thank you very much,

So, Is there any way to check if a class-A RF amplifier is linear or not? (accepting a degree of non linearity) ?

Click to expand...

The classical way to specify and to check linearity uses the degree of distortion (THD in %). It can be measured (resp. simulated) with a spectrum analyzer (resp. FOURIER transform).

Re: Class-A is linear?

Is there any way with the X-Y input of the oscilloscope?

Re: Class-A is linear?

neazoi said:

Is there any way with the X-Y input of the oscilloscope?

Click to expand...

Yes, but as far as I can see - only in theory: If you shift the phase by 90 deg. (sin to cos) a clean sinusoidal signal would produce a Lissajous picture (circle) via the X_Y inputs. But I don`t know how a distorted signal would look like. And, more than that, how one should quantify the result? It`s really only a theoretical game.

Re: Class-A is linear?

neazoi said:

Thank you very much,

So, Is there any way to check if a class-A RF amplifier is linear or not? (accepting a degree of non linearity) ?

Click to expand...

A spectrum analyser (if you have one) would show up the harmonics. An FFT on the oscilloscope can do a similar thing.

Keith.

Re: Class-A is linear?

I am not quite confident of how linearity is related to the harmonics.
a non linear amplifier will produce more harmonics but how can you measure how linear the amplifier is?

The Oscilloscope on the other hand can display the main waveform and by using the X-Y you can notice a pervect circle if the sinusoidal is undistorted.

Another guy told me that if you connect the X input on the input of the amplifier and the Y input, to the output of the amplifier then you must see a straight line about 45 degrees. But can this be used for linearity issues?

Class-A is linear?

I guess the true test would probably be intermodulation distortion. Inject two signals into the amplifier and look for the unwanted intermodulation products with a spectrum analyser. Looking in one of my books on RF power amplifier design it talks about both - intermodulation products and harmonic products. Both will show up with an intermodulation distortion test.

Keith.

Re: Class-A is linear?

The normal method IS to use an X-Y display. Connect the drive signal to the X input and the output signal to the Y input. Use appropriate attenuators of course! Feed a signal into the amplifier, it doesn't matter too much what the signal is but a fairly clean sine way would be best.

If the amplifier is linear you should get a diagonal line because the input X and output Y will be directly proportional to each other if the amplifier is linear. You should expect a little twist or loop in the trace because of phase shifting but the line should still be fairly straight. Non linearity will show as a kink in the line, usually at one end where compression starts to occur.

The other method is to use a normal oscilloscope time trace but feed a composite signal of two non-harmonically related audio tones into the transmitter while looking at the output waveform. Google "ssb two tone" for lots of information.

Brian.

Re: Class-A is linear?

Quote betwixt: The normal method IS to use an X-Y display.

The "normal" method? In contrast, I consider the spectrum analyzer method as "normal" - and, moreover, it is the most accurate method!

You should expect a little twist or loop in the trace because of phase shifting but the line should still be fairly straight.

A line with a loop and - at the same time - "fairly straight"? Betwixt, can you please explain?

Re: Class-A is linear?

Ok - my bad choice of words, sorry !

What I intended to mean is that the line not be absolutely straight because of phase shifts in the amplifier. Phase shifts open the line into an elongated loop, the width of the loop being proportional to the amount of phase shift. Non-linearity will show as the whole pattern becoming asymetric, both sides of the line or loop bending in the same direction.

I agree that non-linearity will show as unwanted products in the output spectrum but that assumes the products were not already there when fed into the linear amplifier. Look at the worst case scenario, you feed a perfect square wave into a linear amplifier, finding harmonics in the output would be considered good news !

Brian.

Re: Class-A is linear?

betwixt said:

........... Look at the worst case scenario, you feed a perfect square wave into a linear amplifier, finding harmonics in the output would be considered good news !

Brian.

Click to expand...

Yes, that`s true. Therefore, both spectrums (at the input and at the output) have to be measured and compared to each other. The difference in THD then is a measure of the amplifier non-linearity