VHF RF Amplifier Design with 2N2222 - I have some issues

Thread Starter

Lalo Mendez

Joined Oct 16, 2018
10
<br class="Apple-interchange-newline"><div></div>

508/5000

I am trying to design a preamplifier for the VHF band of 88-108 MHz. I have made an impedance adaptation using 3 elements taking as central frequency 100 Mhz and a bandwidth of 25 MHz Q = 100MHz / 25MHz = 4. I would like to know how I can improve this design, I think it is amplifying very little. I do not know if I have done badly the impedance adaptation matching in the smith chart or if I am ignoring something in this amplifier design. Is there a way to increase the power output? In this design.
upload_2018-10-16_16-53-22.png
These parameters I have taken from the data sheet. (https://www.onsemi.com/pub/Collateral/P2N2222A-D.PDF)

upload_2018-10-16_17-8-51.png
And these are the admittance parameters that I've got. These formulas I have taken from this book "Estado solido en ingeniera de Radiocomunicacion, Herbert L. Krauss" (pag 125).

upload_2018-10-16_17-13-27.png

And this is the final result I have to design the impedance matching network. The equations taken from "RF Circuit Design 2nd Edition, Chris Bowick" (pag 130-131)
upload_2018-10-16_17-28-45.png
 

Attachments

BR-549

Joined Sep 22, 2013
4,938
"I think it is amplifying very little." How do you know? Is this a sim? or a real circuit? Can you put a test signal in and measure the out?
 

danadak

Joined Mar 10, 2018
4,057
Datasheet at 20 mA shows an Ft of just 250 Mhz, so there is not much G left
at FM radio frequencies. I think you should be thinking Ft of 1 Ghz or there
abouts.

Are you looking for it as a PA or a small signal amp ?

Regards, Dana.
 

Thread Starter

Lalo Mendez

Joined Oct 16, 2018
10
Datasheet at 20 mA shows an Ft of just 250 Mhz, so there is not much G left
at FM radio frequencies. I think you should be thinking Ft of 1 Ghz or there
abouts.

Are you looking for it as a PA or a small signal amp ?

Regards, Dana.
Hello Dana! A Small Signal Amp
 

Bordodynov

Joined May 20, 2015
2,697
I have comments. An amplifier with one transistor without feedback (this case) should not be fed 100mV. If you apply 10mV, then nonlinear distortion can reach 10%. It will be better if you use AC-analysis, and not TRAN-analysis. Also, could you give access to the scheme in LTspice format?
 

Sensacell

Joined Jun 19, 2012
2,625
The transistor (2N2222) is a bad choice for this frequency range, choose one with an Ft at least 10X higher.
It simply has very low gain at these frequencies.
 

Thread Starter

Lalo Mendez

Joined Oct 16, 2018
10
I have comments. An amplifier with one transistor without feedback (this case) should not be fed 100mV. If you apply 10mV, then nonlinear distortion can reach 10%. It will be better if you use AC-analysis, and not TRAN-analysis. Also, could you give access to the scheme in LTspice format?
Thanks for the comments! Here you have the file.
 

Attachments

Thread Starter

Lalo Mendez

Joined Oct 16, 2018
10
The transistor (2N2222) is a bad choice for this frequency range, choose one with an Ft at least 10X higher.
It simply has very low gain at these frequencies.
Thanks for the observation! What happens is that the area where I live is difficult to find RF transistors, they are scarce. Another transistor that I could get is the 2N5179, this should work better. I will try to make a design with this one.
 

Thread Starter

Lalo Mendez

Joined Oct 16, 2018
10
I think you can optimize the input circuit of the filter and reduce the input impedance to 50 Ohms, as well as increase the output voltage.
View attachment 161800
Nice, thank you help me with the circuit! Could you explain how I can optimize it?
Another question hehe, how can I calculate the capacitor values (C3 = 200pF) and resistance (R7 = 10 ohms). I had used a 115nF value.
 

Bordodynov

Joined May 20, 2015
2,697
Some 35 years ago I was calculating the elements of the circuits manually (analytically). Then I did not have a spice. Now I use Spice (LTspice) and use the pick method. In this scheme, I used the phenomenon of quasi-resonance in the emitter of the transistor. I also applied a two-transistor circuit configuration, which is equivalent to doubling the cut-off frequency. At the input of the high-frequency amplifier, it is desirable to use a high-pass filter. This will protect the emitter junction of the input transistor from breakdown. I made television, antenna amplifiers and when connected to a collective antenna, I burned a couple of amps. This was due to the TV's switching power supply.
2018-10-19_10-29-01.png
 

SLK001

Joined Nov 29, 2011
1,548
Some 35 years ago I was calculating the elements of the circuits manually (analytically). Then I did not have a spice. Now I use Spice (LTspice) and use the pick method. ]
What is the "pick method"? Also, how confident are you in the results? Your results look too good for a 100 MHz amp (no distortion in the output). Are you using D1 and D2 as input limiters?
 

Bordodynov

Joined May 20, 2015
2,697
"Pick" for me, is the substitution of various values of elements and observation of how the characteristics change (in this case the frequency response), and then the choice of those values that showed the best result. Since the loho transistor 2N2222 is suitable for gaining 100 MHz, I decided to add a second transistor. The mode of both transistors I decided to choose 10mA. For this, I used two diodes (i.e. diodes not for the limiter). Then I decided to use the back coupling to stabilize the gain and increase the input resistance. At 10mA, the dynamic resistance of the emitter is approximately 3 ohms. At first I chose 20 Ohm, but the gain turned out to be small and I stopped at 10 Ohm.
I also decided to use quasi-resonance to increase the gain. This phenomenon is due to the fact that from the emitter side, the input resistance is inductive in nature (at high frequencies). At first I tried to use 100 pF, but I received a small gain at the bottom of the frequency range. 200pF is better. I did not reduce the input impedance of the amplifier, because it is always necessary. For the absence of reflections in the cable, it is enough to make its coordination at one end. Especially if there will be a cable.I agree that it is necessary to use higher frequency transistors. This should be done on the assumption that in this case there will be less noise. I advise you to explore this scheme "moving" the values of the elements.
 
Top