and before the RF cops think about me...

[mstechca]

OMG!

Good news and bad news for me.

The good news is that I can make any sort of transmitter.

The bad news is that it is transmitting on frequencies I dont want it to transmit on!

and whoever invented 1/(2*pi*sqr(L*C)) should go back to the drawing board because it doesn't work for me in all cases.

The following is the circuit I used. (see picture)

I find that if I use the capacitor between base and ground, then I am able to fine tune the transmitter until the receiver picks up the signal clearly.

BUT, I have tried the following, and it interfers with TV station #12 and yet, the station is local!

I tried removing the base-ground capacitor.

I tried increasing the capacitance in the tank circuit further.

I even tried increasing the resistor to 50K and decreasing the cap from base to +ve to 680pF.

I'm trying to transmit audio from my computer.

my computer alone causes no interference.

 

[Roff]

Your oscillator is oscillating around 100MHz, as you probably intended. However, the output is extremely nonlinear (nonsinusoidal), which creates large harmonics at 2, 3, 4, ... times the resonant frequency of your tank. Channel 12 is at 205.25 MHz, and the 2nd harmonic is probably the highest powered one. This is the reason you have to be careful to add a filter between a high-powered transmitter's output and the antenna. Below are time domain and frequency domain graphs of a ~100MHz oscillator which is very similar to yours.

 

[audioguru]

MStechca,
I don't know why you are using a 2N4402 transistor connected as a diode, instead of a resistor.
I don't know why you are using only the collector-emitter capacitance of the 2N3904 transistor, which changes with supply voltage and temperature changes, instead of a small feedback capacitor.

I think (hope) I solved the interference by harmonics problem in my FM transmitter circuit by adding a tuned output RF amp. Maybe Ron can sim it to see.

 

[Pyroandrew]

OMG!

and whoever invented 1/(2*pi*sqr(L*C)) should go back to the drawing board because it doesn't work for me in all cases.

This formula gives you the frequency in ω (Omega), that is radians per second NOT Hz (cycles per second).

I'm trying to transmit audio from my computer.

my computer alone causes no interference.

Your going to have to apply a lot of knowledge to stop noise, tracking down noise sources is a very laborious task, let alone solving them. As said before You need to take out the harmonics above the intended carrier. By doing this your carrier will become more sinusoidal, cause a lot less interferance and your channel bandwidth will be nice and small, if you keep the FSK ratio small that is.

 

[Roff]

OMG!

and whoever invented 1/(2*pi*sqr(L*C)) should go back to the drawing board because it doesn't work for me in all cases.

This formula gives you the frequency in ω (Omega), that is radians per second NOT Hz (cycles per second).

Not true. I think you forgot that the "2*pi" term converts to Hz.

 

[Pyroandrew]

My mistake, you are right, the "1/(2Pi)" term does indeed do the oposite.

 

[mstechca]

MStechca,
I don't know why you are using a 2N4402 transistor connected as a diode, instead of a resistor.

I'm just setting up a power control. Right now, base and collector of the PNP are grounded for full power. Soon, I will do base of PNP to ground, and then colector to ground.

I don't know why you are using only the collector-emitter capacitance of the 2N3904 transistor, which changes with supply voltage and temperature changes, instead of a small feedback capacitor.

Sorry, my mom was so mad at me last night for causing interference, I actually forgot to add a 5.6pF capacitor between collector and emitter of the NPN.

I think (hope) I solved the interference by harmonics problem in my FM transmitter circuit by adding a tuned output RF amp. Maybe Ron can sim it to see.

The section you highlighted with a red square is what I am using as my tank circuit as well.

Isn't there an extra filter I can add on to cut the harmonics off completely?

 

[zachtheterrible]

Isn't there an extra filter I can add on to cut the harmonics off completely?

Couldn't he add a low-pass filter to cut off everything above 100MHZ or so? This would nesecitate a buffer though because the capacitance or inductance would muck up the LC circuit I would imagine.

 

[atferrari]

What is your intended frequency?

Have you checked that what you are trying to do is legal?

 

[audioguru]

Here is an FM transmitter circuit with an RF output stage with a complex lowpass filter:

 

[zachtheterrible]

thats a pi network. Now ive got a question.

Why not just shunt a capacitor to ground for the lowpass filter? The higher frequencies (harmonics would shunt to ground) and the lower frequency (the one you want) would not shunt to ground.

 

[audioguru]

Hi Zach,
A simple single RC lowpass filter has a very gradual cutoff slope. If the 100MHz fundamental frequency output voltage is reduced by it to 0.7 (-3dB) then the 200MHz 1st harmonic (called the 2nd harmonic) is reduced to only 0.5 (-6dB). With the dual LC lowpass filter, the 1st harmonic is reduced to 0.0625. Since it reduces the harmonics so well, its cutoff frequency can be increased so that it doesn't reduce the fundamental frequency very much. :lol:

 

[mstechca]

That low pass filter set in the circuit is a little strange to me.

I can only tell that L6 and 47p make a low pass filter.

L5 and 56p seem to make a low pass filter. I think 1nF is a coupling capacitor because of its high value. L4 and 15pF is another low pass filter?

Can you please show me all the filters near the antenna? Thanks.

 

[audioguru]

That low pass filter set in the circuit is a little strange to me.

Why? Its function is to reduce harmonics and it uses multiple stages to do its job very well.

I can only tell that L6 and 47p make a low pass filter.

Correct, the 200MHz harmonic is reduced to 1/4.

L5 and 56p seem to make a low pass filter.

Correct. Together with the other lowpass filter, the 200MHz harmonic is reduced to 1/16th.

I think 1nF is a coupling capacitor because of its high value.

Correct.

L4 and 15pF is another low pass filter?

No, at 100MHz they are in parallel and are a tuned circuit. They are in parallel because the two 1n0 decoupling caps have a very low RF impedance therefore the positive supply and ground are at the same RF potential.

Can you please show me all the filters near the antenna? Thanks.

We just finished discussing them. :lol:

 

[mstechca]

Here's the interesting part.

It seems that the 1nF coupling capacitor almost makes my circuit useless. I changed it to 15pF and I have to add a capacitor between + and gnd just to get the transmitter working. Also, with this setup, I wont be able to provide optimum transmitter results with a capacitor connection between base of the NPN and ground.

When I tested everything, it seemed to stop TV interference, however, I turned on the radio which is about 1/2 meter away from the transmitter and flicked through the entire FM band, and came across the transmitter signal only twice.

Why is it that when low pass filters are chained on, they go down to 1/4, and 1/16? why those fractions?

and also, when I introduced the coupling capacitor, there were some oscillations being transmitted to the receiver. It makes me think that somehow, a number of capacitors are being added.

Is there a way I can do this without a coupling capacitor from collector to the filter section?

If I were to transmit at 150Mhz, and I set some high pass filters to 120Mhz or so, how many of them do I need to prevent a radio about 1/2 a meter away from picking up harmonics of the transmitter?

 

[audioguru]

It seems that the 1nF coupling capacitor almost makes my circuit useless. I changed it to 15pF and I have to add a capacitor between + and gnd just to get the transmitter working.

The circuit will work fine without the coupling cap. Its function is to prevent shorting the battery if the antenna touches ground.
Your transmitter has two caps in series between + and ground so adding another one won't make any difference unless the two have wiring much too long.

Also, with this setup, I wont be able to provide optimum transmitter results with a capacitor connection between base of the NPN and ground.

Your oscillator is operating in a common base mode so of course it needs a cap from its base to ground.
The coupling cap to the lowpass filters has nothing to do with the oscillator's base, unless your circuit has ground wiring much too long.

When I tested everything, it seemed to stop TV interference, however, I turned on the radio which is about 1/2 meter away from the transmitter and flicked through the entire FM band, and came across the transmitter signal only twice.

The transmitter has only a single fundamental RF frequency. Any more frequencies picked-up by a nearby radio are caused by the radio giving its "image" frequency pick-up due to its poor design or by the radio's front end overloading also due to its poor design.
Try receiving your transmitter with a good quality home stereo tuner or a car radio and the transmitter will be picked-up at only one spot on the dial.

Why is it that when low pass filters are chained on, they go down to 1/4, and 1/16? why those fractions?

The inductive and capacitive reactances of an LC lowpass filter cause its output response to drop at the rate of 12dB per octave. -12dB is exactly 1/4 of the voltage. One octave is double the frequency.
The load impedance for an LC lowpass filter is very important. With a load impedance too high, the filter is a series resonant circuit which is like a dead short to ground. That is why the two filters in the article are different. The 2nd filter is the load for the 1st one, and the antenna is the load for the 2nd one.

when I introduced the coupling capacitor, there were some oscillations being transmitted to the receiver.

What kind of oscillations?

Is there a way I can do this without a coupling capacitor from collector to the filter section?

Certainly, but keep the antenna away from ground.
Calculate the capacitive reactance of a 1nF cap at 100MHz. Its 1.6 ohms is pretty much like a short piece of wire.

If I were to transmit at 150Mhz, and I set some high pass filters to 120Mhz or so, how many of them do I need to prevent a radio about 1/2 a meter away from picking up harmonics of the transmitter?

You don't want highpass filters. They would pass the harmonics. You want lowpass filters to attenuate harmonics.
The required number of filters and their cutoff frequency is determined by how dirty (distorted) is the fundamental frequency and its level.
A simple LC lowpass filter isn't perfect, it attenuates its cutoff frequency too, then has much more attenuation for harmonics. :lol:

 

[mstechca]

the oscillation is a high-pitched, one-tone sound.

Why would a high-pass filter pass harmonics? I'm thinking of using it because I want to define a frequency range in which the transmitter can operate well in, so that I dont have to mess up anything else.

 

[audioguru]

Hi Mstechca,
The high pitched sound might come from the switching frequency of your computer's power supply beating with the squelch oscillator in your super-regen receiver.

I think you should replace the diode-connected PNP transistor in your transmitter circuit with a resistor for much better linearity of the transmitter's RF waveform. As a rectifier, the PNP-diode might be messing-up the biasing of the RF oscillator.

By its definition, a highpass filter passes high frequencies and attenuates frequencies below its cutoff frequency. The high frequencies that it passes are the harmonics of the fundamental frequency. You don't want to pass the harmonics that cause TV interference, you want to attenuate them with a lowpass filter.
A fixed capacitor in series or in parallel with the variable capacitor will limit the range of frequencies the transmitter can operate in. :lol: