FM transmitter (mod4)

[chingyg]

What you mean by adding a heat sink? the 2n3386 got a built in heat sink doesn't?
Do you replace both Q2 and Q3 with it or just Q3?

 

[Hero999]

With a brand new 9V battery, Q3 in my FM transmitter draws about 50mA and dissipates 0.45W which makes it warm.

Q3 might fry with a 12V or 13.8V supply even if the value of its base resistor is increased.

Ah it'll go into thermal runaway.

Adding a 68Ω emitter resistor should prevent this, at 50mA it should drop 3.4V which will take the voltage across it back down to below 9V.

A more efficient solution would be to bias the transistor in class C so it doesn't dissipate quite so much power.

 

[audioguru]

BTW can I use 2N2222A for the 2N3904's coz i got a bunch of 2N2222A lying around at home.

A 2N2222A has different capacitance than a 2N3904 but should work. Try it.
But the frequency might not be in the FM broadcast band without changing the coils.

 

[transistor495]

What you mean by adding a heat sink? the 2n3386 got a built in heat sink doesn't?
Do you replace both Q2 and Q3 with it or just Q3?

Like this I mean:**broken link removed**

 

[chingyg]

A 2N2222A has different capacitance than a 2N3904 but should work. Try it.
But the frequency might not be in the FM broadcast band without changing the coils.

Would you perfer transistor495's 2N3866?
maybe I should stick to 2N3904?

 

[chingyg]

Like this I mean:**broken link removed**

ok I am going to try to see if I could find one.

No luck finding one. If I do use it, I'd probably just wrap it in thick foil or something.

 

[chingyg]

A more efficient solution would be to bias the transistor in class C so it doesn't dissipate quite so much power.

How would you do that?

 

[Hero999]

Well removing R8 would do that but the duty cycle will be low.

Here's how I'd do it.

The transistor is baised just below the point where it turns on and stays biased in this manner regardless of the power supply voltage.

You'll need more filtering on the output otherwise it will be littered with harmonics and interfere with other stations.

 

[chingyg]

ok, would this work?

Q3 is replaced by 2N3386, Q3 has 10R resistor on emitter.
Q1 is omitted, audio signal is from PC.

 

[Hero999]

That will work but you'll need to do the pre-emphasis on the computer.

 

[audioguru]

Q3 will operate in class-C if the 47k base resistor connects to ground instead of to +9V. I simulated it with a 12V supply and the output power became 0.09 times the power when it was the original circuit with a 9V supply.

 

[Hero999]

Q3 will operate in class-C if the 47k base resistor connects to ground instead of to +9V.

That's true.

I was wrong when I talked about removing R8, it's still required to pull the base down.

I simulated it with a 12V supply and the output power became 0.09 times the power when it was the original circuit with a 9V supply.

That's because the duty cycle would be very low. My circuit will give a duty cycle of just under 50%.

 

[chingyg]

Well removing R8 would do that but the duty cycle will be low.

Here's how I'd do it.

The transistor is baised just below the point where it turns on and stays biased in this manner regardless of the power supply voltage.

You'll need more filtering on the output otherwise it will be littered with harmonics and interfere with other stations.

what about a amplifier like this


C1 68pF
C2 10pF
C3 omited
L1 0.82uH
L2 0.47uH
RFC 10uH
Q1 2N5109

 

[chingyg]

Q3 will operate in class-C if the 47k base resistor connects to ground instead of to +9V. I simulated it with a 12V supply and the output power became 0.09 times the power when it was the original circuit with a 9V supply.

What simulator do you guys use?

 

[Hero999]

Yes, I've seen that schematic on the Internet before.

It will work but needs a fairly strong input signal.

It could be connected to the output of the origional mod 4 transmitter.

I use LTSpice; most people do around here.

 

[chingyg]

Yes, I've seen that schematic on the Internet before.

It will work but needs a fairly strong input signal.

It could be connected to the output of the origional mod 4 transmitter.

I use LTSpice; most people do around here.

that what I was thinking, so input is where the aerial is. and can the output be a 70cm long of wire (aerial) or should I connect a cap in series with it? I must construct it on vero board and keep leads as short as I could. then should I place every thing I a box and cover it with foil that is connected to the ground of the power input? and can I just use regulated 12V for the amplifier (it is 15V on the schematics)?

and for the 0.82uH and the 0.47uH inductor, do you use homemade aircore inductors. using the aircore inductor formula?
LTSpice?
I've got MultiSim 10 and proteus 7

 

[chingyg]

and I also run a coil calculator I got from the internet. It shows that the coil shown on the first page. the one with 9 turns, 5mm diameter and 10.16mm length has got a inductance of 0.23uH. Should I use that coil, or wind my own one that has got 0.1uH of inductance as specified on the diagram.

0.23uH (5-35pF) -> 148.4 Mhz - 56Mhz
0.1uH (5-35pF) -> 85Mhz - 225Mhz

These are from the LC equation, I dont know if it is correct or not, since as audioguru said that the stray capacitance got a huge influence on the behavior of the circuit.

 

[Hero999]

Yes, use the air coils and calculate their dimensions using the inductance formula.

The frequencies will be slightly lower due to the parasitic capacitance.

 

[chingyg]

Yes, use the air coils and calculate their dimensions using the inductance formula.

The frequencies will be slightly lower due to the parasitic capacitance.

So I should use the ones audioguru is using, in his circuit? his design worked after all.

BTW for those TO 39 transistor, Can I improvise the heat sink by wrap foil around it or glue a plate of aluminum on top?

About that Class C amplifier, I posted earlier, RF out can just be an antenna right? no need for another capacitor?

 

[chingyg]

This is what I would use.