# Class A amplification of large signals

Discussion in 'General Electronics Chat' started by Peter Pan, Feb 16, 2009.

1. ### Peter Pan Thread Starter Senior Member

Mar 24, 2005
122
0
Hello All,

I am a beginner in electronics and interested to know whether it is possible to amplify large (with amplitude about 10 V, for example) AC signals using bipolar transistors so that it would be A class amplification (i.e. almost without nonlinear distortions of signal's shape)?

Thank you.

Apr 5, 2008
15,806
2,389
3. ### mik3 Senior Member

Feb 4, 2008
4,846
63
You can if the input signal has a peak-peak voltage of less than 20mV. Also, you will need a common collector amplifier on its output to amplify current if this will be an audio amplifier.

Why don't you use an op amp?
Op amps are much more linear devices than single transistors.

4. ### Peter Pan Thread Starter Senior Member

Mar 24, 2005
122
0
Thank you for reply. So, bipolar transistors can do it only when input voltage is so small? It's a bit sad as I wanted to use 2N1711 for amplification of 5V RF pulses without distortion of pulse shape.

Well, maybe op amps are better I do not know yet. My task is actually this: I need to amplify sequence of rectangular RF pulses (where rf frequency of each pulse is 5 MHz and pulse is repeated every second) from 5 V level to 50-60 V level at 50 Ohm load without much distortion.
But how to do it is flexible and it can be done using bipolar or FET, or Op amplifiers - which active device and technique are actually the best for such task?

Last edited: Feb 16, 2009
5. ### mik3 Senior Member

Feb 4, 2008
4,846
63
The most efficient way to do it is with a half H bridge driver made of two transistors, MOSFET or BJT.

6. ### haruwn Member

Jan 30, 2009
11
0
hi there.......
Buddy m stuck with more or less same case as u are ..... stay in touch..... as soon as i could (InshaALLAH) figure out how to do so....... i would love to help u out..... Rememba in prayers......

7. ### Peter Pan Thread Starter Senior Member

Mar 24, 2005
122
0
This is AB class, isn't it? If so then ... large signal A-class amplification is impossibility?

8. ### Audioguru New Member

Dec 20, 2007
9,411
896
A class-A transistor produces pretty bad even-order harmonics. Most audio amplifiers are class-AB so they don't get very hot all the time like a class-A amplifier and so that the even-order harmonic distortion is cancelled. What remains is a small amount of odd-order distortion that is reduced with negative feedback.

An RF amplifier is sometimes tuned to reduce its distortion or it uses a lowpass filter at its output.

Apr 5, 2008
15,806
2,389
10. ### Peter Pan Thread Starter Senior Member

Mar 24, 2005
122
0
There will be relatively large second harmonic at 10 MHz in output of my amplifier and I am going to reduce it using several stages of low pass LC filter.

Last edited: Feb 16, 2009
11. ### KL7AJ AAC Fanatic!

Nov 4, 2008
2,047
295

Hi Peter:

Your best solution for this is a COMMON BASE linear amplfier. It is commonly used for R.F. power amplification when you have extra input power to "burn". Actually it's not lost...just added in series with the output power. It's a bit tricky to implement, but I have a lot of experience with this.

Eric

12. ### mik3 Senior Member

Feb 4, 2008
4,846
63
The pulses are rectangular and you will switch the high side transistor when the pulse is high and the low side transistor when the pulse is low. Like the CMOS gates output stage.

Mar 24, 2005
122
0
14. ### KL7AJ AAC Fanatic!

Nov 4, 2008
2,047
295

Hi Peter:

I think you'll like the common base circuit because the input and output signals are in phase.

15. ### Peter Pan Thread Starter Senior Member

Mar 24, 2005
122
0
Thank you but this does not seem crucial for this amplifier really. I am more concerned whether central frequency and rectangular shape of RF pulses (with rise times less than a millisecond) will be reproduced at output without much distortions.
I am actually quite interested about what Bertus has advised, it seems the solution!

Last edited: Feb 16, 2009
16. ### Peter Pan Thread Starter Senior Member

Mar 24, 2005
122
0
That's it! Although it's likely to be C class amplifier really but perhaps I was wrong assuming from the beginning that my design has to be Class A only. So I reverse it (about Class A) and get stick to this design for a while and will try to model it on LTspice first.

17. ### bertus Administrator

Apr 5, 2008
15,806
2,389
Hello,

When you look at the description, you have to set a bias current.
Greetings,
Bertus

18. ### Peter Pan Thread Starter Senior Member

Mar 24, 2005
122
0
Hi Bertus,

I have simplified this circuit a little (pic is attached) and set values of some inductors as a guess and then simulated using LTspice. A spectrum of output (an input is 5MHz pulse with duration of 10 us) looks very good (except noisy baseline) although output signal in time domain clearly has some unwanted lower frequency component (pictures are attached). Are any ideas where it comes from and whether it's possible to remove it?

File size:
29.7 KB
Views:
22
File size:
53.1 KB
Views:
25
• ###### linear mosfet power amplifier.jpg
File size:
107.3 KB
Views:
21
Last edited: Feb 18, 2009
19. ### bertus Administrator

Apr 5, 2008
15,806
2,389
Hello,

What is the output impedance from your signal source ?

In the schematic there is a transformer at the input to bring the impedance down.

What is the gate voltage ?

Greetings,
Bertus

20. ### Peter Pan Thread Starter Senior Member

Mar 24, 2005
122
0
It's TTL signal (I have attached pix of input signal and its spectrum) and it looks like that lower frequency component comes with source (input signal) as a side effect of amplitude modulation in the AND gate. I think it has to be filtered out first and then feed to the circuit for amplification.
I have tried this design without transformer as transformer is relatively bulky (in reality) and the circuit (at least its model) seems working well without it.

File size:
61.1 KB
Views:
19
• ###### in_spectrum.jpg
File size:
31.8 KB
Views:
15
Last edited: Feb 18, 2009