Frequency response for a transistor amp.

[walid1]

Finally, after a long discussion with Mr. Audioguru, i discover that the frequency response is a relation (curve) of freq vs voltage gain (Av).

The following Fig is the 1st stage of guru's FM Tx:

Av = RC/re, this is assumed at very high freq that makes Xc4 = 0 so shortened the RE
But at audio freq (30Hz to 15KHz), Xc4 changes from about 35k at 30Hz to about 70 ohm at 15KHz.
So, RE (=470 ohm) can not be considered as shorted by C4 at all freqs
RE//XC4 range from 61 ohm at 15kHz to 464 ohm at 30 ohm

Av = Rc/(re+[RE//XC4]) changes with freq

I wrote a MATHCAD commands to look at this effect of C4, her is it:

And her is the results:

the strange thing is that the freq response is look like this:

with some knees (3dB)
my curve has not a knee

From my MATHCAD curve, I can see that C4 bossts the hi freq signals more low freq signal

please comment and explain this difference for me

http://NOTE: I'm sorry i made MATHCAD for C4 = 100n and discuss for C4=150n
It'll take long time to correct this and to upload the new fig

thanks

 

[audioguru2]

Hi Walid,
FM radio stations and TV stations boost high audio frequencies like that. It is called "pre-emphasis". FM radios and TVs use a similar de-emphasis to cut the treble audio frequencies back down to normal and cut hiss at the same time. If a transmitter doesn't boost the treble audio frequencies then a radio will cut them and it will sound bad, like a stereo with its treble control turned all the way down.

America invented FM broadcasting and used a very high amount of pre-emphasis. Later when Europe began FM broadcasting their newer microphones and recorders had better high frequency response and the treble boost caused overloading of the transmitters. So Europe decided on a reduced amount of pre-emphasis. America kept their higher amount. Radios and TVs are also different between Europe and America.

The knee where the frequency response is +3dB is 2.1kHz (RC= 75us) for America, and is 3.2kHz (1khz (RC= 50us) for Europe.

 

[walid1]

thank you guru


Summary of what guru said:
(1)Pre-emphasis: a process is being carried out at a transmitter to boost high audio frequencies
(2)De-emphasis: a process is being carried out at a reception side to cut the treble audio frequencies back down to normal and cut hiss at the same time.
(3)Why do they boost the treble audio frequencies at the transmitter? If a transmitter doesn't boost the treble audio frequencies then a radio will cut them and it will sound bad, like a stereo with its treble control turned all the way down.

I have a few questions:

(1)Why remained an American used the same pattern despite the tremendous progress of technology?
(2)Why they need to boost then cut, if they didn't cut at the reciever they haven't to boost at Tx?
(3)

The knee where the frequency response is +3dB is 2.1kHz (RC= 75us) for America, and is 3.2kHz (1khz (RC= 50us) for Europe.

(a) 2.1kHz: Is it from MATHCAD curve or from your design and calculations?
(b) RC = 75us: RE * C4 = 470 * 150n = 70.5us?
© 3.2kHz (1khz (RC= 50us) : RC = 50us is ok but what are these two frequencies?

(4) Our region (Israel-Palestine) is closer to Europe. If I did my calculations to design the first stage pre amp And went to the selection of this C4 value, then I have to use (RE * C4 = 50us)? 
(5) The following Fig is drawn using a cwCAD III

I want to watch the freq response But I do not know how to do it, can u please tell me.
I want that to compare between Math calc and swCAD results.
(6) You told me once:

The frequency response of the preamp is flat from 60Hz to about 600Hz and is boosted above about 1.5kHz due to the selectable pre-emphasis capacitor.

How the freq response be flat if we know: Xc4 decrease when freq increase
XC4// RE  is also decrease with f
so Av increase linearly with f ?!
thank you audioguru

 

[audioguru2]

(1)Why remained an American used the same pattern despite the tremendous progress of technology?

If they changed the corner frequency of pre-emphasis then all radios made up to then would sound wrong.
It works good. They use compressors to reduce overload.

(2)Why they need to boost then cut, if they didn't cut at the reciever they haven't to boost at Tx?

Pre-emphasis and de-emphasis is done to reduce hiss. Then the signal to noise ratio is excellent.

(a) 2.1kHz: Is it from MATHCAD curve or from your design and calculations?

2.1kHz is made with an RC of 75us.

(b) RC = 75us: RE * C4 = 470 * 150n = 70.5us?

It is 6% low but it sounds fine.

© 3.2kHz (1khz (RC= 50us) : RC = 50us is ok but what are these two frequencies?

The frequency response is boosted +3dB at these frequencies. The frequencies are a standard of broadcasting.

(4) Our region (Israel-Palestine) is closer to Europe. If I did my calculations to design the first stage pre amp And went to the selection of this C4 value, then I have to use (RE * C4 = 50us)?

I don't know the broadcasting standards in your region. Ask the engineer at a local FM radio station.
 

(5) The following Fig is drawn using a cwCAD III
I want to watch the freq response But I do not know how to do it, can u please tell me.

I have never made a frequency response sweep in SwCADIII. Its "help" section shoud say how.

How the freq response be flat (from 60Hz to about 600Hz) if we know: Xc4 decrease when freq increase
XC4// RE  is also decrease with f
so Av increase linearly with f ?!

Below the corner frequency, the reactance of the capacitor is high so it does not affect the frequency response.
Above the corner frequency, the frequency response rises at exactly 6dB per octave or 20dB per 10 times the frequency. 6dB is a doubling of the level. An octave is a doubling of the frequency. 20dB is 10 times the level.

 

[Kevin Weddle]

Walid, I think the response is flat where the reactance equals the resistance. The lowest value always carries the most weight. But when they approach each other in value, the impedance is severely affected by both.

 

[audioguru2]

No Kevin.
At the frequency where the reactance of the capacitor is equal to the resistor then the response is boosted exactly 3dB (times 1.414 in level). Just like in the curves from Walid.

 

[Kevin Weddle]

I think the simulation is wrong. I don't see where you would get a change in response at Xc = 13000.

 

[audioguru2]

I think the simulation is wrong. I don't see where you would get a change in response at Xc = 13000.

What is Xc=13000?
Walid used the 0.1uF capacitor for European treble boost in his simulation. It is in parallel with the 470 ohm emitter resistor and its reactance equals 470 ohms at 3400Hz where the response is boosted +3dB.