The simplest FM transmitter and receiver module

[Eshal]

Eshal, the reactance of a capacitor is its impedance at a certain frequency. The formula is 1 divided by (2 x pi x f in Hz x C in Farads).
So at 100MHz my 100pF capacitor has a reactance of 16 ohms which is a short to the base-emitter of the mic preamp transistor at 100MHz.
But at 20kHz the reactance is 80k ohms which has no effect.

OMG... how simple it was . Thank you for help audioguru

and i am also thankful for Dan mils too that due to him i consider this formula.

tomorrow i have microcontroller lab. its night here. i need to be offilne but i will continue my struggle for this transmitter and receiver.

thank you all

 

[Dan Mills]

One of the things they NEVER seem to teach is a feeling for when you need to do the maths and when a simple guestimate is going to be close enough.
That 100pf cap is almost certainally in the second category, as in fact is most electronic design (A dirty little secret, most working engineers do not calculate most values in a circuit because there is usually no need)! The art is in knowing which parts can be anything within a decade or so and which need you to get your math on.

I would strongly reccommend a copy of "The Art of electronics" and read the chapters on RLC circuits, it is not too math heavy and will make a nice compliment to your circuit theory book (which I don't know, but are usually rather dense). Circuit theory is one of those subjects where it is very possible to get totally bogged down in the details and miss the big picture and the simplifying assumptions that can be made when things are orders of magnitude apart.

You will however need to understand the theory in order to pass the course, even if it comes out only once in a blue moon once you are working in the field.

Regards, Dan.

 

[Eshal]

Hello to all again.
How are you all?

@audioguru
Hello sir, how are you? I was thinking about your circuit. I am still stuck the mic preamp design. You used voltage divider bias. I have several questions:

1) why did you use voltage divider bias, there are many biasing techniques are available. Like, emitter feedback bias, collector feedback bias, base bias. Why did you choose voltage divider bias?

2) we know if we speak in mic so we can't get good result for transmission that's why we use preamp in order to amplify the mic signal. Right? But how much should we amplify? For example, if we speak in mic and its output is 1Vp-p then how much increase this signal at preamp before sending it to the next stage i.e. RF oscillator stage.

Thank you sir.

- - - Updated - - -

Hi Dan Mills
I agree with you. That's why I read circuit theory at my own with several books and the most favorite one is circuit theory by charles alexender.

 

[Dan Mills]

1: He needed the presence of the emitter resistor to provide the r part of the rc timeconstant required for the preemphasis, and once you have that a voltage divider is about as good as anything else as a means to bias the stage.
It is a simple mic amp in a very crude transmitter, no need to get particularly clever about it.
We all have mental scrapbooks full of neat little circuit fragments for things like this that we just pull out and use without thinking too much about them, and the nice thing about the potential divider bias is that everything about it is trivial to calculate more or less in your head.

2: This is actually a rather difficult question as it depends on how close the sound source is to the microphone and how loud the source is, as this transmitter is not very good and really generates a mixture of am and fm (it relies on changes in the capacitance of the transistor due to changing bias to produce the fm), this was probably determined by experiment. Better designs would sample the audio level and use it to automatically adjust the gain to suit (As well as using a proper varicap diode to handle the modulation).

BTW: 1V would be increadibly high for a microphone output, 0.01V would be more likely and much lower then that is not uncommon.

Regards, Dan.

 

[Audioguru]

Hi Eshal,
The electret mic needs a preamp with a fairly high input impedance so I biased it with high value resistors in a voltage divider. A collector feedback bias has a fairly low input impedance which will load down the level from the mic.

The output from an electret mic is about 10mV (0.01V) when you speak about 10cm from it. I wanted my mic to be VERY sensitive so my preamp has a lot more gain than is needed for speaking close to it. It picks up voices in another room and picks up my stereo playing in its room.

Some simple FM transmitters connect the mic signal directly to the RF oscillator without a preamp. Then they have no pre-emphasis and an FM radio sounds muffled without high audio frequencies.

Here is a simulation of my preamp with a very high 70mV peak input which happens when pre-emphasis boosts high frequencies 4 times to 8 times. I show the small amount of distortion it produces at the very high level.
The input is 70mV peak at 1kHz and the output is 1.25V so the voltage gain is 1.25V/0.07V= 17.9 times.

 

[Eshal]

Hi Eshal,
The electret mic needs a preamp with a fairly high input impedance so I biased it with high value resistors in a voltage divider. A collector feedback bias has a fairly low input impedance which will load down the level from the mic.

The output from an electret mic is about 10mV (0.01V) when you speak about 10cm from it. I wanted my mic to be VERY sensitive so my preamp has a lot more gain than is needed for speaking close to it. It picks up voices in another room and picks up my stereo playing in its room.

Some simple FM transmitters connect the mic signal directly to the RF oscillator without a preamp. Then they have no pre-emphasis and an FM radio sounds muffled without high audio frequencies.

Here is a simulation of my preamp with a very high 70mV peak input which happens when pre-emphasis boosts high frequencies 4 times to 8 times. I show the small amount of distortion it produces at the very high level.
The input is 70mV peak at 1kHz and the output is 1.25V so the voltage gain is 1.25V/0.07V= 17.9 times.

What I got from your all discussion from the post#25 is that we should choose transistor which could have voltage gain of approximately 20 times. Right?

1: He needed the presence of the emitter resistor to provide the r part of the rc timeconstant required for the preemphasis, and once you have that a voltage divider is about as good as anything else as a means to bias the stage.
It is a simple mic amp in a very crude transmitter, no need to get particularly clever about it.
We all have mental scrapbooks full of neat little circuit fragments for things like this that we just pull out and use without thinking too much about them, and the nice thing about the potential divider bias is that everything about it is trivial to calculate more or less in your head.

2: This is actually a rather difficult question as it depends on how close the sound source is to the microphone and how loud the source is, as this transmitter is not very good and really generates a mixture of am and fm (it relies on changes in the capacitance of the transistor due to changing bias to produce the fm), this was probably determined by experiment. Better designs would sample the audio level and use it to automatically adjust the gain to suit (As well as using a proper varicap diode to handle the modulation).

BTW: 1V would be increadibly high for a microphone output, 0.01V would be more likely and much lower then that is not uncommon.

Regards, Dan.

I understand your statement. But sir I am still confuse a little bit. On what criteria should we choose the Transistor for the amplifier. Obviously we can't pic any transistor from the store and fix it in the preamp circuit. There should be some criteria. I want to know that criteria sir.

 

[Eshal]

:-( no reply.

 

[Dan Mills]

Busy......

For something like this we choose a cheap, generic, small signal part that has sufficient ratings and that ideally we have in stock and has a reasonably high HFE and an Ft that is not totally ridiculous (GHz Ft parts can be very hard to use in low frequency designs because they tend to oscilate due to parasitics).

I cannot remeber what Audioguru uses but a 2N2222 or similar would probably be my default choice (Just because I have a bag of about 20'000 of them), even the venerable BC109 would do for the audio stage.

Regards, Dan.

 

[Eshal]

(GHz Ft parts can be very hard to use in low frequency designs because they tend to oscilate due to parasitics).

Explain this please.

Overall, I can understand that for audio stages we should use high hFE and low frequency transistors. Right?
And 2N2222 and BC109 may be recommended for audio stages like audio preamp because they have high hFE and low frequency. Right?

Thank you very much for your reply.

 

[Dan Mills]

A high Ft transistor has gain extending potentially way up into the microwave region, and at such frequencies a few mm of lead length can become an inductor, this means that if great care is not taken with the layout when using such parts you can form inadvertent oscilator circuits at frequencies far removed from anyhting you intended.

The 2N2222 and BC109 are what I would use for something like this because I have thousands of them in the junkbox, but almost ANY low frequency small signal silicon NPN would do in this position.

Regards, Dan.

 

[Audioguru]

Most modern transistors in the normal epoxy TO-92 package work up to several hundred MHz and have fairly high hFE.
I selected the 2N3904 because it has fairly low noise, it works to at least 300MHz, its hFE has a small range from 100 to 300 and it is inexpensive. I could not find small quantities available for a BC549 but it would also work fine.

I used old BC109 transistors in metal cases about 48 years ago when I was an engineer at Philips (my fist job). I saw and worked with their very first compact cassette tape recorder/player (Philips invented it) and red LED. I still have some BC109 transistors.

The transistor part number does not affect the audio voltage gain. The voltage gain is not affected by the current gain. The ratio of the collector resistor and its load to the unbypassed emitter resistor (plus about 90 ohms inside the transistor at this current) determine the voltage gain with a maximum of about 200 times.
My simulation shows a voltage gain of 17.9 times at 1kHz. The collector resistor is 10k with no load and the emitter resistor is 470 ohms so the voltage gain should be 10k/(470 + 90)= 17.9 times.

 

[Eshal]

My simulation shows a voltage gain of 17.9 times at 1kHz. The collector resistor is 10k with no load and the emitter resistor is 470 ohms so the voltage gain should be 10k/(470 + 90)= 17.9 times.

Sir, you previously wrote this too but unfortunately I tried to understand this but I was not able to get this statement. I am a dumb girl. :-(

- - - Updated - - -

What does this statement means? What does this statement implies to? I am not getting what are you trying to teach me by solving for gain of 17.9 times. :-(

 

[Dan Mills]

Ok, to a first order a transistor operating in the linear region controls the emitter current such that the emitter is maintained ~0.7V below the base voltage (Actually it doesn't, but this is close enough).

Now consider what happens if we raise the base voltage by say 1mV, well the transistor will pass enough current to maintain the emitter at a voltage ~0.1mV higher then it was so the base - emitter voltage stays at 0.7V, what does a 1mV change in emitter voltage do? Well the emitter resistor is 470 ohms, and we seem to be assuming 90 ohms for the intrisic emitter resistance, so raising the emitter voltage by 1mV will raise the current in those resistors by 1mV/(470+90 ohms), but that current has to come from somewhere... With a reasonably high Hfe part almost all of that current comes from the collector, so the collector current rises by the same amount.

Now the collector load is a 10K resistor, so a the question becomes how much does a change in current of 1mV(560ohms) cause the voltage across a 10K resistor to vary?

The answer is clearly that it varies by 10,000 ohms * 1mV / 560 ohms = 17.85mV.

Thus an input change of 1mV has causes an output swing of 17.85mV, note however that increasing the base voltage has increased the voltage drop across that resistor and therefore lowered the voltage on the output, making the output swing really -17.85mV, gain is then -17.85mV/1mV = -17.85.

This is one of the really nice things about this topology, the voltage gain is highly predictable and trivial to calculate, being merely - the ratio of the collector and emitter impedances.

HTH.

Regards, Dan.

 

[Eshal]

Hello sir.

Can gain of an amplifier can be negative?

 

[Audioguru]

Can gain of an amplifier can be negative?

Yes, when it is an attenuator.
Mathematically, my transistor preamp has a (negative) gain of -17.9 because its output in inverted from its input. But the output level is still 17.9 times larger than the input level.

 

[Eshal]

When we use common emitter configuration then always we get inverted output with respect to the input. I know this sir.

And why we are using R4=10K and R5=470Ω sir?

Thank you for your response sir.

 

[Dan Mills]

Linear gain figures like we are working with can be negative, if the stage is inverting (Attenuators have linear gains of magnetude < 1).
For gains expressed in dB negitive means attenuation of course and there is no way to express the phase relationships.

Ok, I want of voltage gain of about 20 times, input is in the mV region, and I decide I want about 200uA standing bias, lets walk through the design.
Further constraint is that I need to have a simple way to implement 50 or 75us premeph in a sane way.

Ok the premeph is most easily done across the emitter degeneration resistor, so it is going to be a fixed bias design with emitter degeneration, that settles that.

For best available output swing, we need the collector to sit at just above 1/2 supply rail, so for a 200uA stage current and a 5v rail, that argues for a collector resistor of ~R=E/I = 2.5V/200uA = 12.5K, 10K is standard and is close enough.

Now the gain is approximately given (As I demonstrated earlier) by the ratio of Zc/Ze so, rearranging for a gain of 20 we get Re = 10K/20 = 500 ohms, 470 seems like a good value that is readily available.

Finally, the base bias, for 200uA standing current, we need the emitter voltage to be E=IR = 200uA * 470 ohms = 94mV (call it 0.1V close enough), so the base voltage needs to be 0.7+0.1 = 0.8V.

Hfe of somewhere better then 100, so base current will be < 200uA/100 = 2uA, potential divider chain current should be > 10 times base current, so say 25uA flowing in the bias chain.
R=E/I = 0.8/25e-6 = 32k, and R= (5 - 0.8)/25e-6 = 168K which more or less matches what audioguru designed.

By third year this stuff should really be trivial?

Regards, Dan.

 

[Eshal]

Hello friend.
You are supposing 200uA of current. Is it the Ib (base input current)? If so then why did you choose 200uA? why not above or below this value?

Actually, I am asking just stupid question because I want to know how an electronic engineers like you know all these things.

 

[Audioguru]

Hi Eshal,
The collector and emitter current was selected to be 200uA because then the emitter resistor value is 470 ohms which allows the pre-emphasis capacitor values to be standard, small and inexpensive 100nF or 150nF 5% film type. If the collector/emitter current is 2mA then the emitter resistor will be 47 ohms and the pre-emphasis capacitor values must be 1uF and 1.5uF which have very poor tolerance (50%) if they are electrolytic type or are very large and expensive if they are 5% film type.

If the collector/emitter current is less at only 20uA then the collector/emitter resistor value will be 100k which will be the output impedance of the transistor preamp. Then its output level will be reduced by the lower input impedance of the oscillator.

When I designed the circuit I thought about all these things in less than 1 minute due to my experience.

Guess why there are two different pre-emphasis values for FM radio in the world?

 

[Eshal]

The collector and emitter current was selected to be 200uA because then the emitter resistor value is 470 ohms which allows the pre-emphasis capacitor values to be standard, small and inexpensive 100nF or 150nF 5% film type. If the collector/emitter current is 2mA then the emitter resistor will be 47 ohms and the pre-emphasis capacitor values must be 1uF and 1.5uF which have very poor tolerance (50%) if they are electrolytic type or are very large and expensive if they are 5% film type.

If the collector/emitter current is less at only 20uA then the collector/emitter resistor value will be 100k which will be the output impedance of the transistor preamp. Then its output level will be reduced by the lower input impedance of the oscillator.

Hello sir, again the same thinking which is disturbing me. It is your experience that you think to select the 200uA of collector and emitter current. But how would the fresh engineer like me come into known that how much current is best suited. So that's why I always ask you for calculation in the mathematical form. Can you demonstrate this 200uA of current's calculation mathematically?

When I designed the circuit I thought about all these things in less than 1 minute due to my experience.

I can understand that you are an experienced one. You know, I think on this forum I am the only one who is not experienced otherwise all are experienced.

Guess why there are two different pre-emphasis values for FM radio in the world?

Hmmmm..... because they choose different current values for collector and emitter current?

Regards,
Princess

- - - Updated - - -

By the way, the calculations of Dan Mills and designing of Audioguru are very impressive. I love these knowledge my dear gentlemen... hehehehehehe