another FM transmitter project

[Nepaliman]

Did you mistakenly calculate with 1 M resistor? Diagram has 10 M resistor. I tried once myself but I am being fool.

Oh! If it has low current consumption then it might has very low level output, isn't it?

And, If I need high level audio (high gain), than can I use 10K Rc and 1 Meg Rb (approximately)?


(I thought it is related to input signal level of base, but I was wrong). Now I guessed why base of jFET has 44 Meg, because its collector audio level should be low because collector goes to 'Mic in'.

 

[audioguru]

Did you mistakenly calculate with 1 M resistor? Diagram has 10 M resistor.

Yes, I was wrong and calculated with a 1M resistor. With a 10M resistor it will be biased better.

Oh! If it has low current consumption then it might has very low level output, isn't it?

No, if it is biased so that the output can swing symmetrically up and down almost 9V peak-to-peak then its output level will be very high when it has the extremely high impedance load of old-fashioned crystal headphones.

If I need high level audio (high gain), than can I use 10K Rc and 1 Meg Rb (approximately)?

A high gain transistor causes severe distortion. The gain is higher when the resistor values are higher, not when they are lower.

I guessed why base of jFET has 44 Meg, because its collector audio level should be low because collector goes to 'Mic in'.

A Jfet has a gate, which is not a base. Its load is usually connected to its drain, which is not a collector.
The gate of a FET is an extremely high resistance because there is no gate current so 44M is fine.
I do not know why the "mic in" is connected to the drain which is usually its output.

 

[audioguru]

Here is a simulation of a single transistor that is biased with a single resistor from its collector to its base which provides negative feedback. the negative feedback reduces its gain and distortion if the input source has some resistance.

When the input source resistance is 22k then the gain is only 19 and the distortion at high levels (but not clipping) is very high.

 

[Nepaliman]

Oh Yes! I know feedback R of opamp sets its gain (high gain if higher resistor) but I never thought this is same here . How great- current consumption is low but gain is high!

I followed your simulation and used 10Meg instead of 1Meg then-

Ib=3uA
Ic=300uA
Vb=0.5V
Vc=5V
and distortion is very high- Bad clipping. (I used load resistor 2k instead of 100k, I don't know about it but I tried to simulate like a 2k high z speaker )

From few years ago, I am thinking about a simple thing- you said 'input resistance' and simulated with 22k but how to determine (calculate) this resistance on real circuit? (real circuit has no input resistor) thank you.

 

[audioguru]

EVERY signal source has a certain amount of resistance. The output resistance of a transistor is the resistance of its collector resistor. The output resistance of an opamp circuit is almost zero ohms.

The single transistor amplifier that is biased with a resistor from collector to base has very high gain and very high distortion when fed from a very low resistance like this:

 

[Willen]

Um......then can I say- preamp stage of Mod4 has input resistance up to 10k (mic bias R)? How to learn about it through google?

 

[audioguru]

The 10k resistor that powers the electret mic is parallel to the output impedance of the Jfet inside the mic.
I measured a few electret mics powered from a 10k resistor and the combination was about 2.5k ohms.

 

[Willen]

Recently I prepared a metal housing for MOD 4 and I set the Tx inside this housing. Now It's so easy to handle the transmitter

 

[audioguru]

Did you connect the metal housing to 0V so it is a good shield?
Did the housing change the radio frequency much?

 

[Willen]

Did you connect the metal housing to 0V so it is a good shield?

Yes! Electrically housing has been connected to 0V.

Did the housing change the radio frequency much?

Did you mean its stability when I move near to Tx? I got no more stability improvements in this case. Even I Connected this OV ground to 'Earth Ground' to get VERY stable frequency but if I moved near, (2 or 4 cm near) Tx will change its frequency little. I am using cell phone as audio input using one meter audio cable (cell phone to Tx). So every time when I touch the cell phone, also frequency will change little, then Rx produce hiss (out of frequency). Happens same if I touched the battery. (battery wire is no more long than 1 feet.)

 

[Willen]

Hi AG,

I found a FM tx circuit. It has 0.1uF at its oscillator from base to Gnd (same as 470pf on Mod-4). Won't it ground (short) the audio signal instead of RF??

 

[audioguru]

Another project by a student who knows NOTHING about electronics. It is good that you spotted this horrible error.

The output impedance of Q1 is 10k (R4) so R6 is not needed since it simply makes the value of R4 a little higher.
You are correct, the resistors and C4 form a lowpass filter that has a cutoff frequency of 146Hz. All frequencies higher than 146Hz will be reduced.
It has no pre-emphasis so de-emphasis in an FM radio will cut high audio frequencies even more.

The extremely low values of R7 and R8 increase the cutoff frequency a little and attenuate all frequencies to 0.137 times the level if their values are much higher.

I used 470pF as a low impedance at 100MHz (3.4 ohms) but at 15kHz it is 23k ohms and barely affects it.

 

[Willen]

Here I found another more funny (I think) FM Tx designed by popular site 'circuitstoday'. He used C1= 0.047uF as a parallel of Mic. I thik it will cuts lots of audio. How to calculate it? (I can calculate cut-off freq. in lowpass and highpass filter, but couldn't get more idea here).

Another capacitor C4 has 0.2uF value. Double than previous circuit. Same fault like before? It has no coupling caps and C4 is connected to positive (instead of Gnd). I guessed it also shorts audio like before, isn't it?

Another capacitor C7 has been marked as 0.0022pf hahaha

 

[audioguru]

Another horrible FM transmitter.
The electret mic has an impedance of about 3.3k ohms and has R1 as 2.7k in parallel so the source impedance is 1.5k ohms. The 0.047uF capacitor cuts 2270Hz and all higher frequencies.
The 0.2uF capacitor C4 is parallel with 4.7k so it cuts 170Hz and all higher frequencies.

The battery is missing a parallel capacitor so when the battery runs down a little and its internal resistance rises then the circuit will probably suddenly stop working.

The value of the 0.022pF output capacitor is made by having the circuit in one city and the antenna in another city, fairly far apart. Hee, hee.
A 22pF output capacitor will work fine.

 

[Willen]

The battery is missing a parallel capacitor so when the battery runs down a little and its internal resistance rises then the circuit will probably suddenly stop working.

Oh! I heard such precaution first time and it looks like serious! I never thought that parallel caps has such huge role there. Can I understand little more what did you mean please?

The value of the 0.022pF output capacitor is made by having the circuit in one city and the antenna in another city, fairly far apart. Hee, hee.

hahahahahaha so funny!!!

 

[audioguru]

Oh! I heard such precaution first time and it looks like serious! I never thought that parallel caps has such huge role there. Can I understand little more what did you mean please?

Nearly all circuits have the supply current fluctuating due to signals.
An Energizer 9V alkaline battery has six AAAA alkaline cells in series inside. The datasheet for a AAAA cell lists its internal resistance at 0.225 ohms when new. When new the internal resistance of the 9V battery is 1.35 ohms. Then its impedance is low and the voltage is fairly steady when the circuit it powers has fluctuating current. But as the battery runs down its internal resistance rises to maybe 100 ohms so the voltage jumps up and down when there is fluctuating current. A 100uF capacitor has a reactance of 16 ohms at 100Hz and a 0.1uF capacitor has a reactance of 1.6 ohms at 1MHz and prevent the supply voltage from jumping up and down so are good supply bypass capacitors.

 

[Willen]

Hi AG,

wow I got few hundreds of 0.33uF tantalum (and many more) from Dick Cappels! Before I hardly bought 4 pieces of 330nF non polarized in expensive rate. BUT tanalum is 'Polarized'! Once you are in little doubt about replacing polarized 330nF caps on MOD-4. I am guessing to use positive lead in the Mic side and collector (Q1) side. Can't I? Why were you in doubt? Any mismatch please?

 

[audioguru]

I use boxed film capacitors for coupling audio on all my circuits. They are not expensive because their voltage rating is only 50V or 63V and they are small. In Asia they are made as "green caps" (but some are a brownish-reddish color). The mic voltage is probably more positive than the base voltage of Q1. The collector voltage of Q1 is more positive than the base voltage of Q2.

 

[Willen]

Hi again,
Before when I just have had just few BC547, I used to make 100MHz FM oscillator using this poor transistor which has just 100MHz as a typical transient frequency. Later I decided to I use 250MHz transistor like 2N3904, 2N4401 when I got. But seeing few signal RF transistor, I am thinking to use these RF transistor there as an 100MHz oscillator. Lets say a transistor MPSH34, which has 500MHz min fT and collector current is 50mA. It is said that it is a 'low noise OSC/Amp transistor'.

In you experience and opinion is it good to use such RF transistor as
an 100MHz oscillator than using 2N3904 general purpose? What might be the huge benefits please!

And also can I use it as a class A amplifier to amplify this oscillator's output RF? (maybe same as MOD-4 Q3) But its collector current is rated as only 50mA so I think it would be hard, wouldn't it? (assuming Vcc is 9V DC).

 

[audioguru]

An RF oscillator transistor usually does not need high power and works fine with a voltage gain of only a little more than 1 at the oscillation frequency.
But an RF amplifier might need high power and cooling and might need a high voltage gain.

To make an FM transmitter with a higher output power then you could use a more powerful transistors (and good cooling) for the oscillator and for the RF amplifier and change parts values for higher currents. Use a 12V or more supply instead of 9V.