Hi,
I just got a bit interested in radio design, very basic few-transitor type and I have some questions regarding the basic concepts I would like to ask:
1. I have been reading about superhet receivers and I believe I understood the basic concept behind one, what bugs me is the hardware solution to concept of mixer and its link to audio mixers, also the term of mixing signals. If I understood correctly audio mixer should add signals in time domain producing only two original signals when observed in frequency domain, while the standard RF mixer should produce original signals plus the sum and difference of the signals. Now the question is : how is audio mixing different from RF mixing from point of circuit design ? Trying to simulate a basic audio mixer circuit which should simply "add" voltages I can see that a 100 Hz (difference of 1k and 900Hz) envelope is present which is clearly not the point of audio mixer. Also by searching the internet for audio mixers and RF mixers I could not see any general difference behind the two circuits: both can be accomplished by driving multiple signals to BJT base ( 2 audio signals in audio mixers and RF signal + local oscilator for RF mixers). Only obvious difference is the frequency of the input signals but that should not be of any importance for the basic design of the circuit?
2. From my intuitive point of view, any signal "adding" should produce a "new" signal if we assume that AC signal can be frozen in any time and observed as DC signal, so the resulting signal would be (V1+V2)/2 if the circuit is like the one on the picture (green-op amp output, red and blue - signal generators).
I seem to be missing something here so I was hoping for some explanation. Thank you in advance.
Also I would be very gratefull if you could point me to some literature regarding design of radio circuits which includes explanation of building blocks, I cant seem to find any on the internet, mostly bare schematics.

 

Hello,

In RF mixers they make use of non-linear elements.
The signals get multiplied rather than added as in your circuit.
Have a look at the attached PDF's for more info on RF mixers.

Bertus

 

Audio mixing is usually just algebraically adding the two signals. RF mixing is multiplying the two signals, COS(a) and COS(b). If you apply the trigonometry multiplier identity, you will see that you get not only COS(a) and COS(b), but COS(a-b) and COS(a+b).

 

The confusion is from the same word, mixing, referring to two completely different functions, as SLK001, stated.

 

bevandjero.......An audio mixer sums all the amplitudes and all the frequencies.

A RF mixer takes the differences between 2 frequencies. Notice that there are differences....not just one difference. There are two results(differences)....called heterodynes. So when we mix 2 frequencies....we get two new frequencies.

https://en.wikipedia.org/wiki/Heterodyne

 

Audio mixing doesn't add the frequencies -- the frequencies remain unchanged.

RF mixing produces both sum and difference frequencies (and their harmonics in most cases).

 

OK, thank you for the replies, I understand the mathematical difference between summing and mixing and I was uncautious to use the word "add" frequencies, and I was hoping to get a more in depth explanation of the mixing process particularly how it is done in hardware. From the pdf-s attached by bertus (which I hope to read thoroughly very soon), the basic mixer diode gives me some explanation. If I am right - for the mixer to work properly in a first circuit on the web page - http://www.radio-electronics.com/in...gn/mixers/bjt-transistor-mixer-multiplier.php - the transistor should not be biased in forward-active region, instead it should resemble class B amplifier, and the local oscillator signal should be dominating transistor bias ( turning it on and off), with the smaller input RF signal on top of oscillator signal. This is what produces the "mixing"? If so -> where exactly should the transistor bias point be? Please correct me if I got this all wrong.

 

Hello,

As said, for RF mixing, you will need a non-linear element.
This can be a diode or transistor.

Bertus

PS you might also interested in the following thread I made:
RF related links

 

One of the things that makes understanding and analyzing mixers difficult is that you can't use a linear model for the mixing element -- something that we usually go out of our way in order to make a reasonable thing to do by making the higher order terms negligible. But since mixing is an inherently non-linear process, we have to use a non-linear model -- in other words, we operate the device such that it is the higher-order terms that dominate the response.

 

Hi,
I just got a bit interested in radio design, very basic few-transitor type and I have some questions regarding the basic concepts I would like to ask:
1. I have been reading about superhet receivers and I believe I understood the basic concept behind one, what bugs me is the hardware solution to concept of mixer and its link to audio mixers, also the term of mixing signals. If I understood correctly audio mixer should add signals in time domain producing only two original signals when observed in frequency domain, while the standard RF mixer should produce original signals plus the sum and difference of the signals. Now the question is : how is audio mixing different from RF mixing from point of circuit design ? Trying to simulate a basic audio mixer circuit which should simply "add" voltages I can see that a 100 Hz (difference of 1k and 900Hz) envelope is present which is clearly not the point of audio mixer. Also by searching the internet for audio mixers and RF mixers I could not see any general difference behind the two circuits: both can be accomplished by driving multiple signals to BJT base ( 2 audio signals in audio mixers and RF signal + local oscilator for RF mixers). Only obvious difference is the frequency of the input signals but that should not be of any importance for the basic design of the circuit?
2. From my intuitive point of view, any signal "adding" should produce a "new" signal if we assume that AC signal can be frozen in any time and observed as DC signal, so the resulting signal would be (V1+V2)/2 if the circuit is like the one on the picture (green-op amp output, red and blue - signal generators).
I seem to be missing something here so I was hoping for some explanation. Thank you in advance.
Also I would be very gratefull if you could point me to some literature regarding design of radio circuits which includes explanation of building blocks, I cant seem to find any on the internet, mostly bare schematics.

It may be worth considering a dual gate MOSFET - it seems to be a very popular method.