Colpitts oscillatoroutput power for Am transmitter

[Francesco_bre]

Hi everyone, i'm new to electronics and to this forum. I'm trying everyday to learn more about electronics. In these days i'm experimenting with the colpitts oscillator and am transmitters. I successfully built a colpitts oscillator which resonates around 800 kHz. I also modulated the wave with an audio signal and received the signal with an am receiver. Obviously the range was very low, about 3 feets. My question is, how to increment the output power in order to transmit at higher distances? I also attached an antenna to the output of the colpitts but it wasn't much effective. I also noticed that if i change the inductor to some smaller value, in order to increase oscillating frequency, i get lower output in terms of peak to peak voltage of the produced sine wave. Can someone explain also this behaviour? Thanks in advance.
I leave the circuit diagram of the colpitts and the Ltspice file attached to the post.

 

[danadakk]

Your osc is not buffered, and antennae not matched I gather.

This might help with some basic principles :

https://ia601500.us.archive.org/9/items/QRP_Classics_The_Best_QRP_Projects_from_QST_and_the_ARRL_Handbook/QRP_Classics_The_Best_QRP_Projects_from_QST_and_the_ARRL_Handbook.pdf

Also I presume you are using std alum 10 uF caps, they are not very effective at
high freqs :

Regards, Dana.

 

[betwixt]

Also check the LC ratio in the tuned circuit. The inductance seems too low and the capacitors too high. The resonant frequency can be made from an infinite range of L and C values but there is an optimal range for maximum output. As Dana stated, change the 10uF caps for something more like 100nF ceramic. You may also find it beneficial to take the output from the transistors emitter pin, the level may be a little lower but the antenna itself will have far less effect on the tuned frequency.

Brian.

 

[Francesco_bre]

Your osc is not buffered, and antennae not matched I gather.

This might help with some basic principles :

https://ia601500.us.archive.org/9/items/QRP_Classics_The_Best_QRP_Projects_from_QST_and_the_ARRL_Handbook/QRP_Classics_The_Best_QRP_Projects_from_QST_and_the_ARRL_Handbook.pdf

Also I presume you are using std alum 10 uF caps, they are not very effective at
high freqs :

View attachment 184884


Regards, Dana.

Thank you for your reply Dana. I confirm that i am using standard Aluminum Capacitors, i will change them with some ceramic capacitors, even if they are of much lower capacitance. Thank you for the pdf book, i'll read it to understand better the principles.

 

[Francesco_bre]

Also check the LC ratio in the tuned circuit. The inductance seems too low and the capacitors too high. The resonant frequency can be made from an infinite range of L and C values but there is an optimal range for maximum output. As Dana stated, change the 10uF caps for something more like 100nF ceramic. You may also find it beneficial to take the output from the transistors emitter pin, the level may be a little lower but the antenna itself will have far less effect on the tuned frequency.

Brian.

Thank you Brian for your reply. I didn't take into account the optimal range, I didn't know. Can you suggest me how to find what is the better combination for a given frequency? As Dana said, i'll change the capacitors with ceramic ones. As you suggested, i'll try to take the output on the emitter.

 

[Francesco_bre]

I changed the capacitors to 100 nf ceramic ones and the output voltage increased a little bit. The sinewave amplitude was about 3 volts peak to peak. I connected the output to a bjt emitter follower configuration in order to buffer it. If i understood better, in order to get more range now, do I have to amplify this signal with an amplifier stage to drive the antenna?
Can i use a common emitter amplifier for this scope? Thanks

 

[danadakk]

I would first match the antenna with a matching network. If you receiver has
a signal strength meter on it that helps. Or build one :

https://www.arrl.org/files/file/QST/This%20Month%20in%20QST/2021/07%20July%202021/2021-07%20WHITE%20Sensitive%20field%20strength%20meter.pdf

RF Signal Meter Circuit - Homemade Circuit Projects

In this article we discuss the circuit details of a couple of very interesting and sensitive RF signal meters, which can be used for measuring the RF strength of the […]

www.homemade-circuits.com

Discussion of matching here :

Internet Archive: Digital Library of Free & Borrowable Books, Movies, Music & Wayback Machine

archive.org

Regards, Dana.

 

[BradtheRad]

There's a Colpitts configuration that includes the power supply in the resonating loop. Assuming the supply has low internal resistance it can provide greater current to the LCC combination. And even greater voltage.

 

[betwixt]

Regarding the LC ratio, as I said, increasing L and decreasing C or doing the opposite will still keep the same resonant frequency. 1/(2*pi*sqrt(L*C)) has infinite values to give the same result. However there is a range of values for L and C that work better than others, for example if you used a very large inductor you would need capacitors so small that the capacitance between the wiring itself would mask the tuning you wanted and probably the resistance of the wire would be enough to stop it oscillating.

For 800Khz you should be looking at an inductance of say 100uH and capacitors making 395pF total across it.

Brian.

 

[Francesco_bre]

I would first match the antenna with a matching network. If you receiver has
a signal strength meter on it that helps. Or build one :

https://www.arrl.org/files/file/QST/This%20Month%20in%20QST/2021/07%20July%202021/2021-07%20WHITE%20Sensitive%20field%20strength%20meter.pdf

RF Signal Meter Circuit - Homemade Circuit Projects

In this article we discuss the circuit details of a couple of very interesting and sensitive RF signal meters, which can be used for measuring the RF strength of the […]

www.homemade-circuits.com

Discussion of matching here :

Internet Archive: Digital Library of Free & Borrowable Books, Movies, Music & Wayback Machine

archive.org

Regards, Dana.

Thank you Dana for your answer. I've read about antenna impedance matching. In my case i'm using a simple wire as an Antenna that is long about 262 cm, that i keep high in the room, sticked to the ceiling. From what i understood i need to know what is the impedance of the antenna, that is not easy to calculate without specific equipment. I found some calculators online which give approximate results. The one i used calculated the capacitance, inductance and radiating resistance, which resulted in C=75 pf and Rr=0.021 Ohm. The calculator also gave me an inductance value of L=0.49 mH, but i think it is the inductance needed to cancel out the impedance given by the capacitance.
I tried to put the inductor with that value in between the output of the colpitts oscillator and the antenna, of which one end is connected to ground wire of my house. Anyway, this had no effect on the range. Moreover, i noticed that the emitter follower that i put together before, actually lowers the colpitts output when the 2 circuits are connected, that's why I preferred to connect directly the antenna to the colpitts output through the inductor. I thought that the emitter follower didn't absorb too much current in the base.

This is the site I used as calculator:

Whip antenna calculator | monopole antenna calculator

This Whip Antenna calculator calculates various parameters of whip antenna or monopole antenna.This calculator on the whip antenna mentions formula or equations used for calculations.

www.rfwireless-world.com

 

[Francesco_bre]

Regarding the LC ratio, as I said, increasing L and decreasing C or doing the opposite will still keep the same resonant frequency. 1/(2*pi*sqrt(L*C)) has infinite values to give the same result. However there is a range of values for L and C that work better than others, for example if you used a very large inductor you would need capacitors so small that the capacitance between the wiring itself would mask the tuning you wanted and probably the resistance of the wire would be enough to stop it oscillating.

For 800Khz you should be looking at an inductance of say 100uH and capacitors making 395pF total across it.

Brian.

Thank you Brian for your answer. I understand now what you mean, I'll try to arrange the inductor and capacitor in a better way as you suggested.

 

[Francesco_bre]

I changed the values of inductance and capacitance for the Lc tank. Now i am using ~100 uH inductor and 2x470 pF capacitors. This gave me a resonant frequency of about 1-1.2 Mhz.The output voltage from the Colpitts jumped up to a 9 volt peak to peak sine wave. Just by doing this i increased a lot the distance, i think up to 5-6 meters in my house with a lot of walls as obstacles.
Now my wire antenna has a big effect on distance, without it it didn't transmit no longer than 1.5 meter. I also noticed the little improvement on connecting the ground of the circuit to the earth's ground, the received audio was more clear.
So, in order to understand better, what was the practical reason for which the previous combination of Lc gave me only 3 volts peak to peak?. In the previous circuit i was using little inductor and bigger capacitors, so i think that the little capacitance between the wires, in my case the breadboard(i know it's the worst thing on which to build a radio transmitter, but i needed to change the circuit continuously to design it properly) wasn't disturbing the Lc tank, but maybe the inductor was too small for the Lc tank to oscillate properly? Please let me know if this can be the reason, or if it is another one.
I am a beginner in electronics so there are lots of things i don't know yet.
Anyway i still would like to try to increase the range, as it is an opportunity to learn how it can be done.
Thanks

 

[BradtheRad]

In the previous circuit i was using little inductor and bigger capacitors, so i think that the little capacitance between the wires

Large Farad value is associated with greater current, less voltage swing at a given frequency.

Large Henry value is associated with less current, greater voltage swing at a given frequency.

--- Updated Sep 10, 2023 ---

Also notice in your initial schematic, resistors reduce the amount of current coming from your supply. Your transmitter must draw its energy through those resistors.

Regardless whether your LCC tank has large current swings, its energy also is provided through those resistors.

 

[Francesco_bre]

Large Farad value is associated with greater current, less voltage swing at a given frequency.

Large Henry value is associated with less current, greater voltage swing at a given frequency.

--- Updated Sep 10, 2023 ---

Also notice in your initial schematic, resistors reduce the amount of current coming from your supply. Your transmitter must draw its energy through those resistors.

Regardless whether your LCC tank has large current swings, its energy also is provided through those resistors.

Thank you for your reply Brad. I understand what you are saying. So, in order to be able to draw more current, especially if i want to do just one stage, like only the colpitts oscillator modulated on the supply side by an audio signal, i can adjust the resistors.
If starting from this point, i want to increase the transmitted power with another stage, should I buffer the output and amplify it in another stage? What kind of simple stage can I use for this purpose? I've seen guys on the internet using a push pull amplifier configuration, can it be useful?

 

[dick_freebird]

Attaching the oscillator output to any load will degrade Q and
amplitude. Buffering will give you power gain and antenna match
(i.e. not-loss). Because the output is within the tank, your
antenna may be "pulling" the frequency if the match is anything
but pure-resistive; resistive, however, kills Q.

The "input filter" cap at left should be well bigger than the tank
capacitors and have good (but not perfect) Q - you don't want,
for example, the input filter SRF to land anywhere near fOsc.

 

[betwixt]

The main reason for buffering is to isolate the effects of the antenna on the oscillator frequency. If you attach the wire to the tuned circuit, it becomes part of that circuit and will adjust its resonant frequency. You may already have noticed that moving the antenna or touching it causes the tuning to shift. If you add a buffering stage, it carries the signal forward to the antenna but blocks the effect the antenna has on the tuned circuit. A buffer will not give you increased output power, for that you need a buffer followed by an amplifier stage. I still recommend you take the output to the buffer from the emitter pin.

An ideal buffer presents no load to the stage driving it but doesn't necessarily increase the amplitude. An amplifier increases the signal level and more importantly for, increases the power. Do not confuse amplitude with power, more volts doesn't mean more range and in any case, if you are using AM, you need linear amplification or the audio will become distorted.

Brian.

 

[Francesco_bre]

Thanks for the replies. Now i added the buffer stage with an emitter follower and on the output i get the same sine wave as the input. Now i would like to amplify the power given by this signal to drive the antenna. I'm trying to do it with a push pull amplifier, but first, i need to bias it in the proper way. To bias it, i'm trying to use a Vbe multiplier that I feed with the output of the buffer. The problem is that, even if the circuit (Vbe multiplier+ Piush-pull amplifier) works in simulation, actually my vbe multiplier is lowering the amplitude of the signal at one side, from 8V to 5V. I'm not sure if the Vbe multiplier must be arranged in the way I'm doing it at this point. i would like to have the same signal at both outputs of the multiplier, just shifted Dc wise. I leave attached the circuit I'm using, where the V4 sine wave is coming from the output of the buffer, and R5 is a 2k potentiometer. When i turn the potentiometer i can see the lower output shifting dc wise, while the other output also lowers in amplitude if i reduce the variable resistance.
I hope that you can help me with this. Thanks

 

[danadakk]

I found bias on emitter follower input not flexible, more work would have to be done.

As aside limitations of non licensed transmitters :

https://transition.fcc.gov/Bureaus/Engineering_Technology/Documents/bulletins/oet63/oet63rev.pdf

If you wanted to model antenna more accurately a Nanovna quite useful, and fairly low cost :

NanoVNA | Very tiny handheld Vector Network Analyzer

Very tiny handheld Vector Network Analyzer

nanovna.com

Regards, Dana.

 

[Francesco_bre]

View attachment 184903

I found bias on emitter follower input not flexible, more work would have to be done.

As aside limitations of non licensed transmitters :

https://transition.fcc.gov/Bureaus/Engineering_Technology/Documents/bulletins/oet63/oet63rev.pdf

If you wanted to model antenna more accurately a Nanovna quite useful, and fairly low cost :

NanoVNA | Very tiny handheld Vector Network Analyzer

Very tiny handheld Vector Network Analyzer

nanovna.com

Regards, Dana.

Thank you Dana. Can you explain me what does it mean for the bias to be not flexible?
I tried the buffer+push pull amplifier but something wasn't working as expected, probably i have to spend more time studying that type of amplifier.
Do you think that, in order to keep the circuit simpler, even if more inefficient, i can use directly the emitter follower as power amplifier to drive the antenna?
I would like to take the 8 volts peak to peak sine coming from the Colpitts, pass it through the emitter follower, but amplify the current in the output load, that is the antenna.
I tried this configuration using voltage divider on input to set the bias to the base,but as soon as i connect the colpitts output to the input of emitter follower, the sine wave of the Colpitts reduces in amplitude. Moreover the output of emitter follower is also more reduced in amplitude.
Probably i set up the resistors in biasing and on the emitter in the wrong way.

 

[danadakk]

What is your actual goal for power ?

Take a look at this for getting a few W :

Some more ref reading :

QRP Classics The Best QRP Projects from QST and the ARRL Handbook : Free Download, Borrow, and Streaming : Internet Archive

QRP Classics The Best QRP Projects from QST and the ARRL Handbook

archive.org

Regards, Dana.

--- Updated Sep 12, 2023 ---

Running sim lots of power lost in output coupling C, so here its
eliminated :

Note the above output transistors runnimg way over their Pdiss limits, so those
should be picked for higher power, and heatsink. Note using a P-P output direct
coupled creates a lot of static Pdiss.

I think you have to focus first on matching load to output drivers, 3904 and 3906 (higher
power versions of those.). Using a toroid.


Regards, Dana.