FM converter (mixer) -- I don't understand

[michael1978]

Hello,

Can somebody help me,to explain me this circuits
this is converter, but i really don't understand how it works( i understand mixers, you put rf to base and local oscillator to emiter, but this is a litlle bit complicated)

 

[anorlunda]

I can't speak for others, but to me your images are too fuzzy to read.

 

[michael1978]

I can't speak for others, but to me your images are too fuzzy to read.

Hi, i know, that is true, but about image i alse put pdf file which is big picture, you can zoom image on pdf file, my image in my pc is big, but here to small.

 

[Tom.G]

Th PDF you attached is quite clear.

The "CONVERTER" stage is a combination Local Oscillator and a Mixer. If you are after more details I'll leave that to others here who have more recently worked with such circuits.

 

[michael1978]

Th PDF you attached is quite clear.

The "CONVERTER" stage is a combination Local Oscillator and a Mixer. If you are after more details I'll leave that to others here who have more recently worked with such circuits.

Goodmorning Tom, i know, but i don't understand the formula, i know which make local oscillator and rf signal, and if they input to mixer Transistor, i try to do experiments in simulator ltspice, but i get another result,
So i like to help me, if you can...thnx.

 

[jim hardy]

Mixing is at its simplest just multiplication.
There's plenty of scholarly treatments out there, my first two hits were
ftp://ftp.analog.com/pub/cftl/ADI%20Classics/Basic%20Linear%20Design%20(Linear%20Circuit%20Design%20Handbook),%202007/Chapter_4_RF_IF_Circuits.pdf
https://www.digikey.com/en/articles/techzone/2011/oct/the-basics-of-mixers

but your mixer is simpler than those high performance IC's.

Thought experiment
What if i could modulate a resistor's value at one frequency , f1
and force current of another frequency f2 through it ?
By ohm's law the voltage across the resistor would be their product f1f2, wouldn't it ?

end thought experiment...

" Transistor " is a compound word , a combination of 'transfer' and 'resistor'.
I think of the transistor as a resistor (collector to emitter) whose value is modulated by base current.
So does this Berkeley professor , (properly giving credit, it's at)
http://rfic.eecs.berkeley.edu/~niknejad/ee242/pdf/ee242_mixer_fund.pdf

and that's what your single 2N3904 mixer is doing
(though emitter and collector are swapped from that professor's explanation)

Oversimplifying a little just to get started:::

2N3904's collector current is local oscillator frequency
and RF comes in on the emitter
base is effectively 'grounded' for signal by C14
so emitter-base current is RF input
and collector current gets multiplied by collector resistance which varies with base current at RF input frequency
and collector voltage is the product of the two frequencies
C17B sets Local Oscillator frequency
D3 trims local oscillator frequency to keep it centered on the station being received.

It really does work. Designers in early 20th century did it with four prong vacuum tubes.
https://www.radiomuseum.org/tubes/tube_30.html

it really is that simple, in concept.
You should go through the trigonometry of multiplying cosines having different frequencies , if only to appreciate the genius of our predecessors.

Then it's time to refine the description to include C3 and C23...

 

[Guineafowl]

When’s the electronics textbook coming out, Jim?

 

[michael1978]

Mixing is at its simplest just multiplication.
There's plenty of scholarly treatments out there, my first two hits were
ftp://ftp.analog.com/pub/cftl/ADI%20Classics/Basic%20Linear%20Design%20(Linear%20Circuit%20Design%20Handbook),%202007/Chapter_4_RF_IF_Circuits.pdf
https://www.digikey.com/en/articles/techzone/2011/oct/the-basics-of-mixers

but your mixer is simpler than those high performance IC's.

Thought experiment
What if i could modulate a resistor's value at one frequency , f1
and force current of another frequency f2 through it ?
By ohm's law the voltage across the resistor would be their product f1f2, wouldn't it ?

end thought experiment...

" Transistor " is a compound word , a combination of 'transfer' and 'resistor'.
I think of the transistor as a resistor (collector to emitter) whose value is modulated by base current.
So does this Berkeley professor , (properly giving credit, it's at)
http://rfic.eecs.berkeley.edu/~niknejad/ee242/pdf/ee242_mixer_fund.pdf
View attachment 233015

and that's what your single 2N3904 mixer is doing
(though emitter and collector are swapped from that professor's explanation)

Oversimplifying a little just to get started:::

2N3904's collector current is local oscillator frequency
and RF comes in on the emitter
base is effectively 'grounded' for signal by C14
so emitter-base current is RF input
and collector current gets multiplied by collector resistance which varies with base current at RF input frequency
and collector voltage is the product of the two frequencies
C17B sets Local Oscillator frequency
D3 trims local oscillator frequency to keep it centered on the station being received.

It really does work. Designers in early 20th century did it with four prong vacuum tubes.
https://www.radiomuseum.org/tubes/tube_30.html
View attachment 233018

it really is that simple, in concept.
You should go through the trigonometry of multiplying cosines having different frequencies , if only to appreciate the genius of our predecessors.

Then it's time to refine the description to include C3 and C23...

Hi Jim, thanks for so much info,
so you say local oscillator is at collector(but it most be also inductor not only c17B to set frequency)
and rf at base and emitter
and output is taken from collector? right...
but what C3 and L1, L4 and C16

 

[michael1978]

https://www.physicsforums.com/attachments/233040 I try to build one mixer in ltspice, but the formula and mixer i don't understand i need help
I build this one

 

[jim hardy]

but what C3 and L1, L4 and C16

C3 provides feedback from output to input so that the amplifier will oscillate.
L1 just blocks AC signal from being short circuited to positive supply line

L4 and C16
(i'd appreciate a hand from one of our Ham Radio or @analogdesign guys here )
form a tank circuit in series with emitter
which i believe to be tuned via C17A to track RF amplifier , so as to siphon away RF signals not at desired frequency.
so its impedance is frequency dependent

I know nothing of spice
but your spice circuit shows no path for DC to leave 2N3904's emitter
while your real circuit shows one, L4 and R15.

There's clever AC coupling E-B via the tank circuit and C23
so this seemingly simple stage employs two feedbacks
hence will not lend itself easily to pencil and paper analysis.

 

[tech99]

I think C23 is just a decoupling capacitor, because C1 is such a large capacitance for 100MHz.
The input circuit of the SPICE circuit may not be correct. At LO frequency remember that the input circuit L4 C16 is just 10.7MHz above resonance, so is capacitive, maybe around 40 Ohms, and also C9 is only 5pF, which gives isolation from the generator.
In operation, circuits using 3-electrode devices as self-oscillating mixers are called autodyne mixers. The usual idea is that the tank circuit for the LO can be placed in the collector circuit in series with the IF transformer. Sometimes, feedback is obtained via a tickler coil in the emitter circuit. In this case it is via C4 between collector and base, with base grounded. I have found these oscillators very temperamental.
Some mixer circuits use a tap on the input coil where the LO voltage is zero, by bridge action, so that LO radiation is minimised. It is just possible that is happening here.
Currents at LO and signal frequencies flow in the transistor. One explanation is that the large LO current swing modulates the gain of the transistor and this causes the signals to be multiplied. Another explanation relies on the non-linearities of the transistor.
My personal view is that the circuit is a complete mess!

 

[jim hardy]

My personal view is that the circuit is a complete mess!

I wondered myself about that C14/C23 ratio - thirty-ish ? 3% voltage divider?

Thanks for the help . I'm more accustomed to 6BE6 pentagrid converters.

 

[dlgoff]

When’s the electronics textbook coming out, Jim?

Put me down for two.

 

[michael1978]

C3 provides feedback from output to input so that the amplifier will oscillate.
L1 just blocks AC signal from being short circuited to positive supply line

L4 and C16
(i'd appreciate a hand from one of our Ham Radio or @analogdesign guys here )
form a tank circuit in series with emitter
which i believe to be tuned via C17A to track RF amplifier , so as to siphon away RF signals not at desired frequency.
so its impedance is frequency dependent

I know nothing of spice
but your spice circuit shows no path for DC to leave 2N3904's emitter
while your real circuit shows one, L4 and R15.

There's clever AC coupling E-B via the tank circuit and C23
so this seemingly simple stage employs two feedbacks
hence will not lend itself easily to pencil and paper analysis.

View attachment 233049

Goodmorning JIM thank you very much for help, and for time...i understand now

 

[michael1978]

I think C23 is just a decoupling capacitor, because C1 is such a large capacitance for 100MHz.
The input circuit of the SPICE circuit may not be correct. At LO frequency remember that the input circuit L4 C16 is just 10.7MHz above resonance, so is capacitive, maybe around 40 Ohms, and also C9 is only 5pF, which gives isolation from the generator.
In operation, circuits using 3-electrode devices as self-oscillating mixers are called autodyne mixers. The usual idea is that the tank circuit for the LO can be placed in the collector circuit in series with the IF transformer. Sometimes, feedback is obtained via a tickler coil in the emitter circuit. In this case it is via C4 between collector and base, with base grounded. I have found these oscillators very temperamental.
Some mixer circuits use a tap on the input coil where the LO voltage is zero, by bridge action, so that LO radiation is minimised. It is just possible that is happening here.
Currents at LO and signal frequencies flow in the transistor. One explanation is that the large LO current swing modulates the gain of the transistor and this causes the signals to be multiplied. Another explanation relies on the non-linearities of the transistor.
My personal view is that the circuit is a complete mess!

Hi thank you for help...…..i understand now.

 

[Guineafowl]

Put me down for two.

Seeing as the OP is now happy I’ll risk a quick hijack...

Too many technical books are written by authors who are ‘too clever’ - they have forgotten, or don’t understand, what it’s like to NOT know something.

Many accounts are also written with the goal of showing how clever the author is, rather than conveying understanding. It takes little intellectual effort to regurgitate your knowledge, but lots of effort to shape it into a neat package for a newbie. Turgid, impenetrable textbooks are icons of laziness.

For example - I wanted to know what a phase-locked loop is and how it demodulates FM signals into audio. I read one account online, full of highly technical language and Dickensian sentences, and was none the wiser. I read another, and was shocked at how simple the concept was. I had to read it again because I couldn’t believe the difference.

When a Jim Hardy post comes up, you know you’re going to follow along and end up grasping a concept that you didn’t before. It would be a shame if that gift weren’t immortalised in print.

 

[sophiecentaur]

When a Jim Hardy post comes up, you know you’re going to follow along and end up grasping a concept that you didn’t before. It would be a shame if that gift weren’t immortalised in print.

Hear hear!

 

[sophiecentaur]

I read one account online, full of highly technical language and Dickensian sentences, and was none the wiser.

To be fair, it is not always the purpose of published papers to educate the uninitiated. To start off an advanced paper with a noddy introduction to a subject is not always appropriate. I think you are expecting a lot when you think of how things have changed since the Internet. In the old days, all you had were Learned Papers, Text Books and Constructor Mags. Now you can choose exactly what level you want from hundreds of offerings without even going down to your local shop or library.
"Eeh, you were lucky!" (Monty Python)

 

[dlgoff]

When a Jim Hardy post comes up, you know you’re going to follow along and end up grasping a concept that you didn’t before. It would be a shame if that gift weren’t immortalised in print.

Okay @jim hardy. You need to do this. Shoot; if it would be okay with @Greg Bernhardt, you two could just publish a book from your helpful post. I'd love to have one just because I know you guys.

 

[jim hardy]

Thanks guys for all the kind words.
But I'm the straggler here , my math is not up to snuff for PF . I am lucky you tolerate me.

old jim

 

[michael1978]

Hi, i build this one like Notch filter is good? of i make mistake ? THIS OVER L4 AND C16, THEY FORM NOTCH FILTER , what frequency have to be?

 

[michael1978]

I think C23 is just a decoupling capacitor, because C1 is such a large capacitance for 100MHz.
The input circuit of the SPICE circuit may not be correct. At LO frequency remember that the input circuit L4 C16 is just 10.7MHz above resonance, so is capacitive, maybe around 40 Ohms, and also C9 is only 5pF, which gives isolation from the generator.
In operation, circuits using 3-electrode devices as self-oscillating mixers are called autodyne mixers. The usual idea is that the tank circuit for the LO can be placed in the collector circuit in series with the IF transformer. Sometimes, feedback is obtained via a tickler coil in the emitter circuit. In this case it is via C4 between collector and base, with base grounded. I have found these oscillators very temperamental.
Some mixer circuits use a tap on the input coil where the LO voltage is zero, by bridge action, so that LO radiation is minimised. It is just possible that is happening here.
Currents at LO and signal frequencies flow in the transistor. One explanation is that the large LO current swing modulates the gain of the transistor and this causes the signals to be multiplied. Another explanation relies on the non-linearities of the transistor.
My personal view is that the circuit is a complete mess!

Hi, may i ask you, must L4 and C16 tuned to 10.7Mhz? thanks.

 

[dlgoff]

Okay @jim hardy. You need to do this. Shoot; if it would be okay with @Greg Bernhardt, you two could just publish a book from your helpful post. I'd love to have one just because I know you guys.

Thanks guys for all the kind words.
But I'm the straggler here , my math is not up to snuff for PF . I am lucky you tolerate me.

old jim

BTW: I could tolerate you all day, any day.

 

[michael1978]

BTW: I could tolerate you all day, any day.

but help me ;-)

 

[jim hardy]

may i ask you, must L4 and C16 tuned to 10.7Mhz?

Myself i think it's tuned to the radio station being received.

But I'm not certain about that .
If it's not tunable by C17A i'd think then it should be tuned to the middle of the FM band.

 

[michael1978]

Myself i think it's tuned to the radio station being received.

But I'm not certain about that .
If it's not tunable by C17A i'd think then it should be tuned to the middle of the FM band.

Hello Jim
So Fm band is 88 to 108Mhz middle and is 98MHZ so i have to tune to 98 MHZ, thank you very much JIM.

 

[NascentOxygen]

I would say that L2 with C17A give the variable RF tuning, and L1 with C17B give the tunable local oscillator. C9 is a coupling capacitor. I don't know what to say about the fixed-frequency tuned circuit L4-C16.

 

[michael1978]

I would say that L2 with C17A give the variable RF tuning, and L1 with C17B give the tunable local oscillator. C9 is a coupling capacitor. I don't know what to say about the fixed-frequency tuned circuit L4-C16.

Thank you man, thank you for info,
Maybe must be tuned to if 10.7Mhz,
i don't know, i am building one double conversion fm superhet receiver, i buy all the componens, so i going to finish, after i goona build this one,
ok NascenOxgen

 

[Svein]

Reminds me - I once built an FM receiver somewhat like that. I remember doing it on the copper side of a blank PCB with a thin brass shield across the transistor and a coil consisting of about 10 turns of a copper wire wound around a pencil. The components were given small copper islands (scratched out with a scalpel) to live on. It was fun to build, but not very stable and not very economical.

 

[michael1978]

Reminds me - I once built an FM receiver somewhat like that. I remember doing it on the copper side of a blank PCB with a thin brass shield across the transistor and a coil consisting of about 10 turns of a copper wire wound around a pencil. The components were given small copper islands (scratched out with a scalpel) to live on. It was fun to build, but not very stable and not very economical.

Helllo, i buy the pc board and i put rf amplifier componenet
after i start to build local and mixer, i put components, i start to put if transformer 10,7mhz, and if transformer he can't go inside pc board, so i leave it,(i make nervous), but i hope tomorow to start again, to build

 

[jim hardy]

I would say that L2 with C17A give the variable RF tuning, and L1 with C17B give the tunable local oscillator. C9 is a coupling capacitor. I don't know what to say about the fixed-frequency tuned circuit L4-C16.

I'm beginning to come around to that line of thought too.

 

[michael1978]

I'm beginning to come around to that line of thought too.

Goodmorning Sir,
I think today i going to do a little bit, because my mother have a birthday and we have a little bit like a party to do...……
Greetings Jim.

 

[Tom.G]

I hazard that L4 - C16 functions to isolate the received signal from the Q3 oscillator feedback network. That is, it blocks the received signal from Q3-B and ensures that it drives Q3-E.

Cheers,
Tom

 

[Svein]

Sigh. Once 10.7 MHz was a high frequency. You needed shielded, tuned MF transformers and high frequency tubes. Now the mainstream CPUs use a clock frequency of at least 3500MHz! Thus, an entire FM receiver can be put inside an IC like this (http://circuitscheme.com/fm-receiver-with-tda7021t.html):

 

[michael1978]

I hazard that L4 - C16 functions to isolate the received signal from the Q3 oscillator feedback network. That is, it blocks the received signal from Q3-B and ensures that it drives Q3-E.

Cheers,
Tom

Hi Tom,

i take the book from library and i read, he explaing almost everything, but L4 and C16 not, Strange

 

[michael1978]

Sigh. Once 10.7 MHz was a high frequency. You needed shielded, tuned MF transformers and high frequency tubes. Now the mainstream CPUs use a clock frequency of at least 3500MHz! Thus, an entire FM receiver can be put inside an IC like this (http://circuitscheme.com/fm-receiver-with-tda7021t.html):
View attachment 233269

Hi, i buy long time ago, one chip tda7000 i think, but is easy to build, but you don't learn nothing, so now i already everything i know, how to build self, just a little bit, after i can use ic chip

 

[jim hardy]

That's a great way to learn, building from scratch.
ARRL Radio Amateur Handbook is a great reference. It will give you tips about layout . At high frequency stray capacitance can bring surprises.

 

[jim hardy]

I hazard that L4 - C16 functions to isolate the received signal from the Q3 oscillator feedback network. That is, it blocks the received signal from Q3-B and ensures that it drives Q3-E.

Cheers,
Tom

That's sort of along the lines i had thought.
It'd make more sense to me were it tuned to the station being received.
That's why i suggested set it midband.

 

[michael1978]

That's a great way to learn, building from scratch.
ARRL Radio Amateur Handbook is a great reference. It will give you tips about layout . At high frequency stray capacitance can bring surprises.

Thank you Jim i read a lot of times, but difficult to understand, but this book electronic communications for technecians is very goooood, i borow and i read.

 

[jim hardy]

Thank you Jim i read a lot of times, but difficult to understand, but this book electronic communications for technecians is very goooood, i borow and i read.

You're doing, actually building something.
You learn far more by doing than by reading about doing.

Spread those wings and fly, young man !

 

[michael1978]

You're doing, actually building something.
You learn far more by doing than by reading about doing.

Spread those wings and fly, young man !

Thanks Jim

 

[michael1978]

I hazard that L4 - C16 functions to isolate the received signal from the Q3 oscillator feedback network. That is, it blocks the received signal from Q3-B and ensures that it drives Q3-E.

Cheers,
Tom

Hello Tom
i try to simulate in ltspice, i tune L4 and C16 to 98mhz,? i get this result, is ok?

I hazard that L4 - C16 functions to isolate the received signal from the Q3 oscillator feedback network. That is, it blocks the received signal from Q3-B and ensures that it drives Q3-E.

Cheers,
Tom

HI Tom,
i try to experiment, and what do you think is this mixer good, and result good, output from collector
see

 

[Tom.G]

Does not look right. What are the parts values in the original circuit of C17B and of L1?
Also please simulate the circuit with source V1 having zero output and show the waveform and FFT traces again.

Cheers,
Tom

 

[michael1978]

Does not look right. What are the parts values in the original circuit of C17B and of L1?
Also please simulate the circuit with source V1 having zero output and show the waveform and FFT traces again.

Cheers,
Tom

Sorry tomi don't understand with V1 having zero output

i build the same result, wha do you think maybe wronk circuits? because also in the book they don't say nothing about L4 and C16

see picture, alsoi put the file if you need, i don't know anymore what to do, you have to rename cricut.txt circuit.asc

 

[michael1978]

I hazard that L4 - C16 functions to isolate the received signal from the Q3 oscillator feedback network. That is, it blocks the received signal from Q3-B and ensures that it drives Q3-E.

Cheers,
Tom

Hi Tom
i get this result without L4 AND C16

 

[Baluncore]

I think you need to draw the circuit in a sensible way.
C17a is used to tune the RF front end.
C17b is used with L1 to set the local oscillator frequency. C17b is labelled FM RF but it is actually the LO.
You should be able to see the Local Oscillator in the circuit as drawn here.

Varactor D3, (not shown), is used to trim the LO frequency, it moves the IF signal to center of FM detector slope.

 

[michael1978]

I think you need to draw the circuit in a sensible way.
C17a is used to tune the RF front end.
C17b is used with L1 to set the local oscillator frequency. C17b is labelled FM RF but it is actually the LO.
You should be able to see the Local Oscillator in the circuit as drawn here.

View attachment 233461
Varactor D3, (not shown), is used to trim the LO frequency, it moves the IF signal to center of FM detector slope.

First I apologize, i did want to make you tired, for the last time, you explain very good in details,
I apologize,
and thank you very muchhhhhhhh, you are a doctor

 

[michael1978]

I think you need to draw the circuit in a sensible way.
C17a is used to tune the RF front end.
C17b is used with L1 to set the local oscillator frequency. C17b is labelled FM RF but it is actually the LO.
You should be able to see the Local Oscillator in the circuit as drawn here.

View attachment 233461
Varactor D3, (not shown), is used to trim the LO frequency, it moves the IF signal to center of FM detector slope.

Hi Balounce one more time me, you forget bout C23, i done see, where you have to put in circuite between ? C5 is C23
Thank you.

 

[Baluncore]

It is difficult to simulate frequency mixers in spice. You can independently simulate the RF input impedance or the IF output network in the frequency domain, or oscillation of the LO in time, but not both time and frequency domains in the same run.

The way your example circuit was drawn effectively disguised the mixer in the LO where; Two signals are present on the non-linear emitter of Q3. One is the continuous oscillation of the LO, the other is the RF coupled into the emitter through C9. One of the cross-modulation products is selected by the tuned circuit in the collector, as the IF.

It is simple to understand once you look at the untangled circuit from the right direction.

Replace C5, 100 pF with C23, 300pF. I think I was thrown by having to guess the missing value for C14. What was the value of C14 before you put the solid black line of misunderstanding through it ?

 

[michael1978]

GoodMoring Sir, ithe value of C14 IS 5nf, and this circuit i take from the book, he explain almost everything but not all, i replace C5 with C23 300pFTHANK YOU...…………….