Audioguru FM Tx

[audioguru2]

Hi Walid,
You should never assume Vbe in a transistor circuit that has a low emitter voltage. It is typically 0.62V on my datasheet.

I used the parts I had. I didn't have 320k for R2 nor 60k for R3 so the divider current to base current ratio is 20:1 instead of 10:1 which better controls the effect of different current gain in different transistors.

My Mod-3 version of the circuit had an extra emitter resistor to raise the emitter voltage so that different Vbe in different transistors wouldn't affect the collector voltage too much. But I didn't have space for it and a large bypass capacitor for it.

The pre-emphasis at 15kHz in North America is about 7.2 times the gain at lower frequencies. With its lower frequencies gain of about 18 then the max gain of the transistor of about 130 results in a pre-emphasis which just makes it and sounds perfect.

You have used hFE (DC gain) of 230 in your calculation for Zin instead of Hfe (AC gain) of 100 so the actual impedance is less. Since the pre-emphasis capacitor affects only high frequencies then the correct value for input impedance occurs only at about 30kHz and higher. At lower frequencies the input impedance is much higher. ;D

 

[walid1]

Hi audioguru
because I have 4 figures, I'll make this question into two.

From What you said:"You should never assume Vbe in a transistor circuit that has a low emitter voltage. It is typically 0.62V on my datasheet."
I can understand:
1) Are u tell me to go to datasheets only when determining VBE and never to assume it?
You teach me before that it is not different in electronic calculations. What the problem if it 0.6, 0.7 or 0.8 volts if we know that resistors are 5% error.
2)From the beginning of this discussion till now you frequently said datasheets datasheets datasheets datasheets datasheets datasheets datasheets datasheets datasheets datasheets...
I finally look at these datasheets of FAIRCHILD 2n3904 transistor.
Fig.1 and 2 are the relation btn VBE and IC, one in saturation and the other in ON.
What I can understand is that u choose the 25 degree curve.
I dont know the difference btn them or from any of them u take your values.

View attachment 37907

View attachment 37908

 

[walid1]

In fig3 seems to be that figure u use to determine hfe vs the operating IC.
Figure 4 is to determine hie.
tell me how exactly u use them, and if VCE or frequency is defferent from specified in the curve, how can I use them.
thanx.

View attachment 37909

View attachment 37910

 

[walid1]

audioguru: "I didn't have 320k for R2 nor 60k for R3 so the divider current
            to base current ratio is 20:1 instead of 10:1 which better
            controls the effect of different current gain in different
            transistors."
walid: who said that u use R2=320k and R3=60k, and if u do so the divider
current to base current ratio will be 5:1 and not 20:1.
And Why say that, I can't understand u!

 

[walid1]

Hi audioguru;

you said: My Mod-3 version of the circuit had an extra RE to raise VE so
          that different Vbe in different transistors wouldn't affect the
          VC too much.

The question: Are u mean a bypassed RE added in series to a not bypassed RE?

you said: But I didn't have space for it and a large bypass capacitor for it.
walid: I don't know what do u mean!

 

[walid1]

Hi audioguru,

u said: The pre-emphasis at 15kHz in North America is about 7.2 times the
        gain at lower frequencies.With its lower frequencies gain of about
        18 then the max gain of the transistor of about 130 results in a
        pre-emphasis which just makes it and sounds perfect.

My questions:
(1) What the meaning of The pre-emphasis from the view of electronic?
(2) What kinds of audio at 15KHz?
(3) Why u relate the 15kHz to North America region only?

 

[walid1]

Good night Audioguru,

U said: You have used hFE (DC gain) of 230 in your calculation for Zin
        instead of Hfe (AC gain) of 100 so the actual impedance is less.
walid: Why to use AC hfe, u never said that before?       
      And how to find this AC hfe?


And u said: Since the pre-emphasis capacitor affects only high frequencies
            then the correct value for input impedance occurs only at about
            30kHz and higher. At lower frequencies the input impedance is
            much higher.
walid: I can't understand anything from this.

 

[audioguru2]

Hi Walid,
1) Resistors are only 5% but a Vbe of 0.8V is a whopping 33% higher than 0.6V.
2) The graph for Vbe ON Voltage is different from the graph for Vbe Saturation Voltage and you should use the correct one (ON Voltage).
3) The graphs in my Fairchild datasheet don't have figure numbers. It has a graph for hFE (DC) where it is 230 from a collector current from 0.1mA to about 8mA then drops at higher currents, and a different graph for Hfe (AC) where it is 70 at 0.1mA and rises to 180 at 10mA. At the 0.34mA in my transmitter it is exactly 100. There is another graph for Hfe at radio frequencies at 10mA.
4) A graph shows the AC input impedance Hie at various collector currents. It doesn't agree with Hfe x 26mV/IE and I don't know why. I trust the graph.
5) I wanted a divider current to base current ratio of 10:1. I didn't have 286k and 54k resistors to give a divider current of 14.7uA, so I used resistors that give a divider current of 26.3uA for a ratio of 17.8.
6) The Mod-4 bypass cap is small because it is to bypass only high audio frequencies. In my Mod-3 circuit, I had an additional RE to raise the emitter voltage and it was bypassed with a huge electrolytic cap for bypass of all frequencies. The additional resistor and huge cap didn't fit on my Mod-4 Veroboard.
7a) Pre-emphasis boosts (more gain) high audio frequencies. It can electronically be done by using the capacitive reactance of a bypass cap for a transistor's emitter resistor. At low audio frequencies the capacitor has no effect like it isn't there. At high frequencies the cap is an AC short.
7b) Young people up to about 30 years old can hear up to 20kHz. FM broadcasting is limited to 15kHz maybe because it is old. Musical instruments' harmonics and voice sillibant sounds extend past 15kHz to 20kHz.
7c) FM broadcasting was started in North America and they decided to use a huge amount of pre-emphasis. Later, microphones and electronics improved and when the Europeans tried it their transmitters were overloaded with the boosted high frequencies so they decided to use 67% of the amount used in North America. Of course, FM radios in North America have more de-emphasis (treble and noise cut) than the ones in Europe.
8) I discussed the difference between hFE (DC) and Hfe (AC) above.
9) I discussed the reactance change of the pre-emphasis capacitor at low and high frequencies above. (Smiley)

View attachment 37911

 

[walid1]

Hi audioguru
(1) In my calculations above hie= 10.9Kohm at Ic=0.3mA, and Fig 4 above reveal this it is about 10Kohm, so I can say that my calcuklations are true. what would u say?
(2) I have the same datasheets from FAIRCHILD. If u look at Fig.2 above u may notice that VBE at IC=0.3mA = 0.63V and not 0.62V if u want to be more accurate.
(3) Why i have not the AC hfe curves in my datasheets as u, mine is also from TAIFCHILD as yours, anyway are u mean that hfe in hie=hfe*re is the AC hfe and not DC hfe?
(4) When i face a problem like you in not having the resistors that make the desired ratio 10:1, is this the best solution to disturb the design and make the ratio near 20:1?
(5) What did u mean by The Mod-4 bypass cap, i think you mean the the Ce cap in your fourth version (C4 in the first design). Are u made 4 versions of this Tx, why?
(6) If I want to boost a high audio frequencies of 10KHz using a a bypass cap for a transistor's emitter resistor, how can I calculate its value?
    my answer: Its Xc must be <= 0.1RE at the desired frequency, correct me.
(7) I see from all what u tell me that millions of things must be taken into the mind when trying to design a preamp circuit, and man can't do this. so I think one must use a computer tool to do this huge work for him!
(8) below u can see a graph (Fig.freq responce ver02) represents the frequency response of this preamp, i use a pspice to do it, what u comment about it.
(9) In my calculations above u comment only about VBE, and Hie only. Can I assume that the remaining calc. are true?
(10) I ask u this question:"Why C2, it is strange, may be for negative feedback"
    and u answer:"C2 shorts the base to emitter at RF frequencies to prevent RF pickup at the mic from being rectified by the transistor which would upset its bias. "
    I'm not understand it as u expect, but I'll do my explaination and u correct me:
At RF freq. like 100Mhz, Xc2= 16 ohm = short circuit
At AF freq. like 2KHz, Xc2 = 800 Kohm = open circuit
Now if the MIC picks up any RF power from the TX itself or from other Tx's, it give it to the base of Q1 which rectified them into DC and this DC will change the Q1's bias.
I'm sure this is wrong wrong explaination, because this rectified RF power is Dc and C2 can't pass DC to ground! I don't know, u may tell me.
Another point about this: I never see a like this cap btn B and E, how this idea come to u?
  thank u audioguru, I feel that your student in advance.... may be in one day have the ability to answer and help people like u do.

View attachment 37914

 

[walid1]

Hi
I put C4 = 50nF instesd of 100nF in the Pspice commands that follows, to see how this affect on the pre-emphasis:

Audioguru FM Tx 1st stage (Audio preamp), V- divider configuration Ver.02
VCC 1 0 5V
Vmic 8 0 AC 1
*Vmic 8 0 SIN( 0 5M 1K 0 0 0 )
Rmic 8 7 3.3K
R2 1 2 150K
R3 2 0 39K
Q1 3 2 4 Q_2N3904
R4 1 3 10K
R5 4 0 470
RL 6 0 21.6k
C1 7 2 330N
C2 2 4 100P
C3 3 6 330N
C4 4 0 50N
.MODEL Q_2N3904 NPN (Is=6.734f Xti=3 Eg=1.11 Vaf=74.03 Bf=416.4 Ne=1.259
+                  Ise=6.734 Ikf=66.78m Xtb=1.5 Br=.7371 Nc=2
+                  Isc=0 Ikr=0 Rc=1 Cjc=3.638p Mjc=.3085 Vjc=.75 Fc=.5
+                  Cje=4.493p Mje=.2593 Vje=.75 Tr=239.5n Tf=301.2p
+                  Itf=.4 Vtf=4 Xtf=2 Rb=10)
.TRAN 1ns 2Ms
.OP
.AC DEC 20 1Hz 20kHz
.probe 
.end

the freq. response curve shown below (Fig.freq responce C2=50n).

My questions:
(1) I don't so any resonable difference btn this curve and the simillar one above, comment.
(2) In our calculations above we conclude that IC =0.3 mA and IB =1.3 uA, and we correct IB to be 2.6uA to maintane a ratio of 10:1,
then, I calc Hie to be = DC hfe*re = 230*83.3 =  about 19Kohm and audioguru correct me to Hie = AC hfe*re and tell me that AC hfe from datasheets = 100, so, Hie=100*83.3 = 8.3K ohm.
Also we conclude that the intended v.gain = 10 and Zout =RC =10k.

the following is the result of the program above:


NAME        Q1       
MODEL        Q_2N3904 
IB          9.91E-06 <=== ours =1.3 or 2.6uA
IC          2.05E-04 <=== ours = 0.3 mA
VBE          6.24E-01 ===> its very good as exactly audioguru said
VBC        -2.22E+00
VCE          2.85E+00 ===> about VCC/2 = 2.5V
BETADC      2.07E+01 <=== 20.7 it is so far from 230
GM          7.91E-03  $$$$ I don't know what this
RPI          3.24E+03  <=== it is Hie = 3.24K and our was 8.3K
RX          1.00E+01  unknown
RO          3.71E+05  the trasistor o/p Z and when this in //RC ++> Zout=RC
CBE          8.69E-12
CBC          2.38E-12
CJS          0.00E+00
BETAAC      2.57E+01  this Ac hfe far from 100
CBX          0.00E+00
FT          1.14E+08

the voltage gain appears in the second Figure (Fig.Av 1st stage). From this figure we can say that Av = 43.8/5 = about 9 very close to 10, I accept this.
Why these differences? pease comment to know where to depepnd on Pspice and where not.
thank u.

View attachment 37915

 

[audioguru2]

(1) In my calculations above hie= 10.9Kohm at Ic=0.3mA, and Fig 4 above reveal this it is about 10Kohm, so I can say that my calcuklations are true. what would u say?

Close enough, who cares?

(2) I have the same datasheets from FAIRCHILD. If u look at Fig.2 above u may notice that VBE at IC=0.3mA = 0.63V and not 0.62V if u want to be more accurate.

Close enough, who cares?

(3) Why i have not the AC hfe curves in my datasheets as u, mine is also from FAIRCHILD as yours, anyway are u mean that hfe in hie=hfe*re is the AC hfe and not DC hfe?

Correct. Impedance and Hfe is for AC, resistance and hFE is for DC. Go to http://www.fairchildsemi.com/ds/2N/2N3904.pdf and download their newest datasheet.

(4) When i face a problem like you in not having the resistors that make the desired ratio 10:1, is this the best solution to disturb the design and make the ratio near 20:1?

A divider to base current ratio of 10:1 is good and a higher ratio is a little better if your project can work with the resulting lower input impedance due to the lower resistances of the divider.

(5a) What did u mean by The Mod-4 bypass cap, i think you mean the the Ce cap in your fourth version (C4 in the first design).

The bypass cap is the pre-emphasis making cap C4 in my Mod-4 circuit.

(5b) Are u made 4 versions of this Tx, why?

I made the 1st one to see if my guess about its wrong biasing was correct, and by how much. I was correct because it didn't work with a new 9V battery and it didn't work when the battery was a little old. I made Mod-2 to fix it but it sounded bad and drifted its RF frequency all over the place. I fixed those problems in Mod-3 then simplified and perfected it into Mod-4. They were fun.

(6) If I want to boost a high audio frequencies of 10KHz using a a bypass cap for a transistor's emitter resistor, how can I calculate its value?
    my answer: Its Xc must be <= 0.1RE at the desired frequency, correct me.

No. For the frequency response to rise 3dB at 10kHz, the Xc of the cap must be equal to RE.

(7) I see from all what u tell me that millions of things must be taken into the mind when trying to design a preamp circuit, and man can't do this. so I think one must use a computer tool to do this huge work for him!

There is nowhere near a million things to think about when designing such a simple circuit. Maybe only 30 things. Man has been designing circuits much more complicated without computers for a million years. I don't use a sim program because I don't trust them and I like to exercise my mind.

(8) below u can see a graph (Fig.freq responce ver02) represents the frequency response of this preamp, i use a pspice to do it, what u comment about it.

It doesn't have enough levels and frequencies grid lines to see much and I dis-agree with its roll-off of low frequencies starting at about 200Hz. I designed and measured its frequency response to be absolutely flat to 60 Hz. Its -3dB frequency was about 15Hz for it to transmit really good bass sounds.
Maybe Spice didn't use the microphone's and R1's output impedance of about 3.3k ohms in series with the input cap. 

(9) In my calculations above u comment only about VBE, and Hie only. Can I assume that the remaining calc. are true?

You are the man and got the remaining calc's correct. C'mon over here and I'll slap your back or something and give you a cold beer if you want. Sorry, I've run out of gold stick-on stars.

(10).............
I'm sure this is wrong wrong explaination, because this rectified RF power is Dc and C2 can't pass DC to ground! I don't know, u may tell me.

The base-emitter junction of a transistor is a perfect rectifier of strong RF. When the RF swings positive, the transistor's junction is forward biased and C1 charges with its input positive. When the RF stops or swings negative then the charge on the transistor side of the cap is too negative. Therfore the transistor gets cut-off.

Another point about this: I never see a like this cap btn B and E, how this idea come to u?

It was difficult because my transmitter made strong RF all over my workshop/lab. I tried the cap from the input to ground and a few other places but it didn't stop the strong RF from messing-up the transistor's biasing. When I stuck the cap between b and e then it was perfect.

I feel that your student in advance.... may be in one day have the ability to answer and help people like u do.

Good, you're learning. Next time somebody asks 10,000 questions I'll call you to answer them. (Smiley)

 

[audioguru2]

I don't trust Spice so I don't use it.
You need to understand exactly what it is saying. Some things are for the transistor alone, and other things are for the entire circuit. Sometimes the model of a transistor from a manufacturer is incorrect.
A Spice sim on another electronics chat site has a circuit similar to this one and shows a common power transistor producing an output voltage higher than its supply! The transistor has 3 models from different manufacturers. Only one of them is correct and they are all different.

The different pre-emphasis caps produce a frequency shift of one octave from each other since the values of the caps are 2:1. One octave is difficult to see on the graph.

The base current is 1.3uA with a typical transistor. It is never 2.6uA unless the transistor is saturated. The ratio of the divider current to the base current should be 10:1 or more to ensure that the Vb voltage doesn't change much when a transistor has lower gain than typical. ;D

 

[walid1]

Hi Audioguru

THE VCO STAGE
TANK CIRCUIT (LC)
L=100nH, C=5-35pF, so the freq range from 87MHz at c=35P to 225MHz at 5p.
To make the coil of 100nH, u may wound 5.5 turns of 1mm enameled copper wire on 5mm in dia former, the turns must be close so that the length of the coil = 5-6 mm.
When I looked to the photo of your actual TX, I noticed that u wound 10 turns on about 2.5mm in dia former and of length = 10mm
(1) What the difference btn using the two coils. Notice that my coil is half in number of turns and in length and double in diamete.
(2) If-as u told me- MIC provid Q1 by 5mV signal, and we know that Av about 10, so the sig at the base of Q2 is about 50mV, this 50mv voltage applied to base of Q2, and we know that VB=VE+0.72=2.25+0.72=2.97V, so, Vb changes btn 2.97V+0.05 and 2.97V-0.05.
From all that I see that u waste the current. You need that the Q2 not saturated nor cutoff, so VB need to be only 0.72+0.05=0.77V or 0.8V. What u say?

 

[walid1]

Hi audioguru
once u told me in one of your replies that u geuss VE to be 2.25V and ...
From that deduce your design decisions which are:
IC=10mA
VE=2.25V
Vcc=5V
Q2 = 2N3904 has hfe=230 at the operating current (IC) ....

If I use these decesions with the same 2n3904, then I measure IC and IB and they give a different hfe, say 160 and not 230. You know that at this case IB and IC became different, What to do ?

 

[audioguru2]

THE VCO STAGE
TANK CIRCUIT (LC)
L=100nH, C=5-35pF, so the freq range from 87MHz at c=35P to 225MHz at 5p.

I didn't design the tank, I just copied it and many other ideas from Rod Elliot's project: http://www.sound.westhost.com/project54.htm
I didn't measure the inductance and I just used little trimmer capacitors that I had.
The minimum capacitance of the trimmers is definately not only 5 pF due to stray capacitance of the wiring and the transistor's output capacitance and even the effect on the tank of the positive feedback capacitor. I didn't measure the frequency range and couldn't tune it over the FM band because there are too many powerful FM stations around me. My FM dial is full.

To make the coil of 100nH, u may wound 5.5 turns of 1mm enameled copper wire on 5mm in dia former, the turns must be close so that the length of the coil = 5-6 mm.
When I looked to the photo of your actual TX, I noticed that u wound 10 turns on about 2.5mm in dia former and of length = 10mm
(1) What the difference btn using the two coils. Notice that my coil is half in number of turns and in length and double in diameter.

I think that your coil would allow the circuit to tune higher frequencies.

(2) If-as u told me- MIC provid Q1 by 5mV signal, and we know that Av about 10, so the sig at the base of Q2 is about 50mV, this 50mv voltage applied to base of Q2, and we know that VB=VE+0.72=2.25+0.72=2.97V, so, Vb changes btn 2.97V+0.05 and 2.97V-0.05.
From all that I see that u waste the current. You need that the Q2 not saturated nor cutoff, so VB need to be only 0.72+0.05=0.77V or 0.8V. What u say?

You are talking about AM (and the associated FM) modulation the VCO transistor. You forgot that it is also a high-level VHF oscillator that needs a 10mA collector current to have a large VHF output voltage swing.

then I measure IC and IB and they give a different hfe, say 160 and not 230. You know that at this case IB and IC became different, What to do ?

Yes, every transistor has a different hFE. The circuit has an emitter resistor which provides DC negative feedback so the change in the transistor's operating point is fairly small when the hFE is changed. ;D

 

[Alun]

I think that your coil would allow the circuit to tune higher frequencies.

Higher than 225MHz?
With a 2N3904?
I don't think so as it only has a GBWP of 300MHz, maybe slightly higher but the gain needs to be considerably greater than one, especially at these frequencies where the Q of the tank circuit might not be that high.

 

[audioguru2]

Hi Alun,
We were comparing my 10.5 turns, small diameter coil with his 5.5 turns, larger diameter coil.
Using the same tuning capacitor then his would tune to higher frequencies than mine because his coil's inductance would be less than mine.
I don't think either circuit would tune to as high as 225MHz because of stray capacitance. ;D

 

[Alun]

Hi Alun,
We were comparing my 10.5 turns, small diameter coil with his 5.5 turns, larger diameter coil.
Using the same tuning capacitor then his would tune to higher frequencies than mine because his coil's inductance would be less than mine.

In theory yes, but too bad it probably wouldn't oscillate because the trnaistor won't have enough gain a such a high frequency.

I don't think either circuit would tune to as high as 225MHz because of stray capacitance. ;D

Yep too bad, the stray capacitances in the transistor (and eleswhere in the construction) have acted together and conspired against our very (or even ultra) high frequency goals. ;D

 

[walid1]

Hi all my friends, and special HIIIIIIIIIIIII to audioguru

Now and finally i am going to build this FM Tx , yesterday i put all the components on the table and today afternoon i'll put them together
my question is if i build my own receiver using TDA7000 or TDA7088 to receive signals from that TX ...
another way, I want to talk with my friend 100 meter far away, i need 2 units each contain TX and Rx
If I tune Tx1 to 100MHz and Rx1 to say 80MHz, so, Tx2 to 80MHz and Rx2 to 100MHz
I know that to each fundamental freq there are some harmonics,
when, some day i build my first transmitter i hear myself in many places along the 88-108MHz band
I am worry from that in the sme unit the sender hear himself
If this occur my project will fail ans so my efforts so please tell me and advice me how to do this and what must i do

thanks

 

[audioguru2]

if i build my own receiver using TDA7000 or TDA7088 to receive signals from that TX ...
another way, I want to talk with my friend 100 meter far away

The TDA7000 and probably also the other IC are very simple and cheap. They don't make an FM radio as good as the home stereo tuner and car radio that I tested my FM transmitter's range with.

If I tune Tx1 to 100MHz and Rx1 to say 80MHz, so, Tx2 to 80MHz and Rx2 to 100MHz
I know that to each fundamental freq there are some harmonics

Those cheap "radios" have an IF frequency of only 70kHz, so image frequency multiples of it might cause interference. There is a big gap between 80MHz and 100MHz so it might work fine.

when, some day i build my first transmitter i hear myself in many places along the 88-108MHz band
I am worry from that in the sme unit the sender hear himself

Good point! ;D
The cheap radios overload easily with strong stations. Your transmitter will be a very strong station to your cheap "radio"and cause overload. The 1st stage of a good FM tuner is tuned and has automatic-gain-control to avoid overload.

If this occur my project will fail and so my efforts so please tell me and advice me how to do this and what must i do

Adding a tuned circuit to the input of each "radio" will help prevent overload. Adding AGC would be very difficult.