Simple Breadboard 16MHZ AM transmitter

[truffaldino]

Hello,

I was trying to design and build a simple solderless breadboard 16MHz transmitter (schametics of final stage is shown on image atached) but incountered a problem with antenna.

I was thinking that antenna is just a piece of wire sticked into the transmitter, but it turned out that designing an antenna is a science.

On my schematics and in LTSPICE simulations I presented antenna as a capacitor+resistance. Then I put inuctance and additional variable capacitor in series to make the antenna resonant.

My antenna is just about 7meters of wire lying on the floor in my room. Tuning capacitor is a cheap 20pf tiny trimmer cap and inductance is 10uH small choke.

According to LTSPICE simulations the energy otput of a resonant antenna is about 0.5watt, but in reality my transmitter range turned out to be a couple of steps from antenna.

I will be grateful if you could explain in simple terms how to tune the antenna using minimal means. As far as I understood I'm using "end feeded dipole antenna" and it has quite high reactance. I thought I could cancel this reactance with the inductance choke, but this simple solutions seems to be wrong.

Thank you in advance

 

[c_mitra]

At this frequency, the wavelength corresponds to 1.8 km; a lamda/4 antenna will still be too long. We just use an untuned antenna but there are other loop antenna for various breeds. For 0.5W power you can just use a simple wire hanging from the window (but not lying on the floor or within inches from the wall). This piece of wire acts like a capacitor (the other plate is the mother earth) and also an inductor (the value can be computed)- about a few 10 pF /meter of capacitance and a few 10 nH/meter of the wire.

If you have a power meter, you can tune by adjusting the 20 pF cap till you get max power. If you shift your antenna, you will have to work out this again. A small inductor (resonant) may be put in parallel with this cap. The antenna shall feed on this inductor.

Currently your antenna is lying on the floor which increases the capacitance greatly. You need to stretch it across the opposite corners (about 1-2 ft below the ceiling). Still better is to stretch it outdoors (like a clothes wire) but the ends should be attached to insulators.

 

[betwixt]

8-O 16MHz = wavelength of about 18m not 1.8Km !

A quarter wave antenna should be about lambda/4 = 4.5m but you will need a better matching network than that to maximize the power transfer to it. Incidentally, it isn't clear from your schematic where the antenna is connected, I assume you realize R4 is the antenna and not a physical resistor in series with it.

Brian.

 

[truffaldino]

Thanks Guys,

Yes, I modeled antenna by 50 Ohm resistor according to equation for dipole radiation resistance R=20*pi^2(L/lambda)^2. I asumed that the radiation resistance is the same for symmetric dipole or the end-fed one. I also measured antenna capacitance with capmeter, it is about 100pf when it is on the floor.

So, I added about 10pf (20pf trimmer) cap and 10uh inductor in series (this is about 16MHZ serial resonance) to make antenna resonant.

I also made a simple power detector (two germnium diodes, capacitor piece of wire and ammeter) to tune antenna.: I have trimmed capacitance to get maximum ammeter current and the current change is indeed of order of magnitude when trimming. But when I swith receiver on, there seems to be a little effect from this capacitor. Signal I hear does not change much when I trim capacitance and I hear signal only if I am close to antenna.

So, indeed, the better matching network is needed. My question is what could be simple working mathing network, how to design it?

 

[c_mitra]

8-O 16MHz = wavelength of about 18m not 1.8Km !

Sorry for the mistake and thanks for correcting.!!

 

[betwixt]

Truffaldino, unfortunately its a little more complicated than just adding a resonant element in series with the antenna. The first problem is not knowing the driving impedance which in your circuit would be difficult to predict. Second problem is related, how much power (strictly speaking current) are you providing at the base of the transistor as this will determine how it operates. The 2N3904 would have to be driven into saturation to get any usable power from it, the graphs on it's data sheet end at IC = 100mA and that only equates to about 0.4W DC input power so you would never get 0.5W out from a 5V supply.

I would suggest looking at a higher rated transistor and an 'L' or better still 'pi' matching network as that would allow a degree of matching in both directions, to the transistor and to the antenna. It also has higher attenuation of harmonics which would be important if you are saturating the transistor.

I also measured antenna capacitance with capmeter, it is about 100pf when it is on the floor.

Interesting: it isn't the correct way to do it but for curiosity sake, what did you measure the capacitance against?

Brian.

 

[chuckey]

As your aerial is shorter then a quarter wave its basic impedance will be less then 50 ohms and will have a small capacitor reactance in series with it. Now as its laying on the floor there is always the extra 100PF, and how its distributed. Suppose your aerial has a 20 PF as its reactance. Resonating the current "aerial" would results in 5 times more current flowing directly then flowing through the 20 PF and the radiation resistance.
To resonate the aerial look at the current taken by you PA transistor. As you get your aerial resonant, the current should dip. Make sure that you can tune through the dip. If the aerial is reflecting a high resistance, then the current should be low. As you match the aerial, you are transforming its resistance to a lower value, so the current will rise. Then you have to make sure that its still tuned (for a dip). So you have to change the match and tuning gain and walk these components to their best position.
Frank

 

[c_mitra]

Even if your antenna is not tuned and you may not be able to put out maximum power out of the simple antenna, you should be able to get a decent reception anyway. Please tell us about the range and quality of the reception (you should be able to hear it on a SW radio). With some more power, you should be able to cover the world.

 

[chuckey]

Not with the aerial laying on the floor.
Frank

 

[c_mitra]

So, indeed, the better matching network is needed. My question is what could be simple working mathing network, how to design it?

First you need to take it out in the open, stretch it nicely using some porcelain insulators and use a coax (just leave the braid unconnected) to feed it. With very little effort, you can make significant improvements to this simple antenna. But for the purpose of testing, this will do for the time being.

 

[chuckey]

I have just looked at your circuit again, that 10PF in series with the aerial is far to small. As explained the aerial already looks like a capacitor. If you just added the correct inductance in series, the two would cancel out and allow the resistive part of the aerial to be presented to the transistors collector.
The problem now becomes matching the low resistance of the aerial to the impedance of the transistor. As its a weedy transistor, it will not be taking much current, so the impedance will be high, say 1K. This you will have to match to the 20 ohms of the aerial. So some how you need a 7:1 transformer. The normal way to do this is via a PI network. This is a series L with a capacitor to earth at end. The ratio of the capacitors give you the transformer ratio. The value of the two capacitors in series has to resonant with the L at the working frequency.
Wiki says :- https://en.wikipedia.org/wiki/Antenna_tuner it deals mainly with L sections , but its a start.

Frank