Ever wondered why the intermediate frequency (IF) in radios is fixed at 455kHz? The concept of IF is not only fascinating but highly relevant till date. To truly grasp how wireless communication has evolved and where it is headed, understanding the importance of IF is essential.
Intermediate frequency (IF) is a crucial concept in the field of electronics and communications, especially in the design and operation of radio, television, and radar systems. Let us try to demystify the concept of intermediate frequency and see its use in the household AM radios.
While the radio itself is an age-old concept, the concept of IF is an important field in electronics and telecommunication engineering. So let us also see why we need the IF or how it is helpful in the context of making a good radio receiver. It is interesting to see why the IF is fixed at 455kHz and not any other frequency in our AM radios! Though it is not a comprehensive document concerning everything about radio engineering, it attempts to serve as a good introduction to the fundamentals of the IF concept.
In theory, an audio wave can range from 20Hz to 20,000Hz. For AM (amplitude modulation) based radio communication, a carrier wave is needed to transmit the audio wave. At the radio transmitter side, we take the audio wave, which has a changing frequency and amplitude, and modify the amplitude of the carrier wave according to the audio wave, while keeping the frequency of the carrier wave unchanged. The resulting output, or the modified carrier wave, is called a modulated wave. Note that only the amplitude of the carrier is modified, hence the name amplitude modulation (AM). Essentially, the modulated wave is then sent to the antenna for AM broadcasting.
| How IF works in radio receivers |
| In a typical radio receiver, the incoming radio frequency (RF) signal is mixed with a signal from a local oscillator. This mixing process generates two new frequencies: the sum and difference of the original frequencies. The difference frequency is usually the intermediate frequency (IF). By converting the original RF signal to this IF, it becomes easier to process – filter, amplify, and demodulate. |
Why do we need a carrier wave? It is a good question. Even though it is not very practical, let us just imagine that we were trying to send the audio signal directly for AM broadcasting. How do we then create a circuit that will pick up the radio signal from a particular radio station if there are dozens of them in an area? If all the AM radio stations in a city broadcast the audio directly, we will not be able to distinguish them from each other, because all would be transmitting anywhere and everywhere between 20Hz and 20,000Hz! That would be like everyone in a room full of a hundred people shouting at each other in the loudest way and no one being able to hear clearly what any other person was saying.
To prevent this issue, it was thought in the early days that every AM radio station should be given a fixed frequency to transmit so that any receiver tuning to that frequency would be able to hear only that station. The frequency on which a particular radio station is transmitting is called the carrier frequency of that station.
A radio station just sending a carrier wave of a constant amplitude and frequency is of no use because the receiver will not get any audio out of it. Remember the modulation? Essentially, modifying the amplitude of such carrier wave with an audio signal and then sending that from the radio station will make it useful (see Fig. 2). Yes, it is still at the frequency fixed for the radio station as we modified only the amplitude and not the frequency. Thus, we could send the audio signal by means of an amplitude change of the carrier wave to the receiving radios tuned to that carrier frequency.
For Europe, Africa, and Asia, the medium wave AM band consists of carrier frequencies from 531kHz to 1602kHz, with each station separated by 9kHz. (For a deeper look at the evolution of radio receivers, see the article titled ‘The Transformation of Radio Receivers’ at https://www.electronicsforu.com/market-verticals/radio-receivers-transformation link to understand how the radio receivers evolved.)
Challenges in AM reception
As we now know, each broadcasting station transmits its amplitude modulated carrier wave at its own specified frequency, as fixed by the concerned government authorities and regulations.
This is good. However, we still have a problem.
We now need to have a radio receiver designed with an amplifier for amplifying the received carrier wave. That is also not a problem. The real problem is designing an amplifier that will amplify such high frequency, but the response of the amplifier should also be good for an entire range of possible carrier frequencies covering all the radio stations between 531kHz and 1602kHz!
Having an amplifier that would work well with such high frequencies and for the entire range of frequencies (of various carrier waves) with same quality is very difficult.
Then how about having many amplifiers, each designed for various carrier frequencies? That may work, but it would make the cost and size of such a receiver very high. Remember, even such a solution demands all the amplifiers working for high frequencies as well. Imagine how difficult it could have been in those early days of AM broadcasting and reception where realising high frequency circuits was still a challenge. No wonder, the concept of IF emerged then.
The concept of IF
So, what is the solution for the limitations described above? Well, how about creating an amplifier that is well suited for a single frequency and converting the received modulated carrier wave of any station to the frequency to which the amplifier is well suited for? It is like, let us say we have an amplifier designed to amplify a signal at 100kHz, and we convert the modulated wave received from 1200kHz to 100kHz so that the amplifier can happily amplify it. Similarly, if we need to hear the radio station transmitting the modulated carrier signal at 1400kHz, we simply convert the received modulated wave of 1400kHz to 100kHz and then send it to the amplifier that is designed to work well at 100kHz. So now we are converting any modulated carrier frequency of any station in the entire AM frequency range to a fixed 100kHz, and then such a signal at 100kHz is amplified, and then audio wave is extracted from that 100kHz. This frequency, that is, 100kHz in this example, is called the intermediate frequency or IF.
How IF is achieved
Can it be done? It turns out that we can! How do we do that?
Let us see an even better example.
Imagine a radio transmitter or broadcasting station sending the modulated carrier wave at 300kHz. Let us also imagine that, at the receiver side, we have an oscillator, and let us call this oscillator as ‘local oscillator’ or simply LO (see Fig. 1). Assume the LO would produce any frequency we want. Remember, the LO’s output need not be an amplified signal. Since the received modulated carrier signal, which we are going to mix it with, is weak anyway, and thus the weak output of LO is fine for our work here.
Let us also imagine we have a signal mixer that will produce the addition and subtraction of the two signals we provide. Let us say the received modulated carrier frequency is f1 and output of the local oscillator is f2. The mixer will take the signal of frequency f1 and signal of frequency f2, then it will produce addition and subtraction of f1 and f2 and produce its output as f2+f1, and f2-f1.
We can see an example with various values of f1 and f2 in Table 1.
| Table 1: Various values of f1 and f2 | |||
| Carrier (f1) | LO (f2) | Mixer (f2+f1) | Mixer (f2-f1) |
| 600 | 700 | 1300 | 100 |
| 800 | 900 | 1700 | 100 |
| 1300 | 1400 | 2700 | 100 |
| 1400 | 1500 | 2900 | 100 |
| 1559 | 1659 | 3218 | 100 |
| 1600 | 1700 | 3300 | 100 |
Resulting outputs f2+f1 and f2-f1 of the mixer are result of both f1 and f2, and hence both f2+f1 and f2-f1 contain the information of the modulated carrier wave f1 and LO wave f2.
As you can see, the f2+f1 keeps changing depending on the f1 and f2, while the f2-f1 is always constant! That is, it remains 100kHz in this example. It means, if we apply a filter on the output of the mixer to select only f2-f1 (that is 100kHz in this example) and filter out all other frequencies, including f2+f1, then we will end up only with f2-f1 (without the f2+f1 content).
In other words, to extract any particular station from a bunch of stations, all we need to do is change the LO frequency (f2) such that the f2-f1 is 100kHz!
Thus, for any f1 (of station of our interest), we just need to adjust local oscillator to produce a frequency f2 in such a way that f2-f1 will result in 100kHz. For example, if we want to hear the radio station Akashvani Mumbai, which is transmitting at 558kHz (let us take it as f1), all we now need to do is just adjust the local oscillator to produce 658kHz (let us take it as f2), so that the mixer will produce f2+f1 and f2-f1. Thereafter, when the mixer output is passed through a filter to allow only f2-f1, we will get f2-f1, which is at 100kHz. The Important thing to note here is, that the 100kHz is the intermediate frequency or IF. When this IF (see Fig. 3) is sent to an envelope detector or demodulator (see Fig. 4), it will remove the carrier wave and give the original audio (baseband signal) out. This audio signal can be further amplified suitably and heard through a loudspeaker.
Did you notice the advantage of the IF? As you can see, we just need an audio amplifier that will amplify only the audio signal. One can also have an amplifier before the envelope detector so that the mixer output is amplified a bit more before sending it to the envelope detector, if needed. The main advantage of the IF is, we can convert the incoming carrier frequency (f1) to a smaller frequency (f2-f1), which is the IF, and it becomes lot more convenient to design all the circuitry after the mixer to work with a fixed frequency.
Why IF is fixed to 455kHz in AM radio
There is a very interesting background for this. Following are some reasons for choosing the exact value of 455kHz for IF:
1. Keep it outside the AM band
It was important to ensure that any radio transmission was not exactly like IF. That is, the IF should be outside the AM band of 531kHz to 1602kHz. That means the IF should be either lower than 531kHz or higher than 1602kHz.
2. Be as less as possible
Lower frequency was better as we did not have to use the expensive circuitry such as filters and IF amplifiers for working at high frequencies in the olden days.
When active filters were used, built using vacuum tubes or transistors, they generally had higher gains at lower frequencies. So fewer amplifier stages were required if the frequencies were lower due to high gains at the lower frequencies. Thus, the choice of having IF above the AM band (i.e., above 1602kHz) was ruled out.
3. Be as high as possible.
The higher IF provides a sufficient bandwidth to accommodate the audio signals without distortion, ensuring clear and high-quality sound reproduction. Of course, it should still be lower than the lowest of the AM band (531kHz).
Further, the higher the IF, higher the gap between tuned frequency (f2-f1) and the image frequency. This higher gap helps in filtering out the image frequency more easily. (We will discuss the image frequency soon.)
| Advantages of IF |
| While IF is not mandatory, here are some advantages of using it: Consistent performance. Since the IF is fixed, regardless of the actual broadcast frequency, the filters and amplifiers can be optimised for this specific frequency, ensuring consistent performance. Simplified tuning. It allows the use of a single, fixed-frequency filter and amplifier stages, simplifying the design and tuning of the receiver. In the early days, high frequency and wide-band amplifiers were a lot more expensive than they are now. Improved selectivity and sensitivity. With a stable IF, the receiver can more effectively select the desired signal and reject others, improving both selectivity and sensitivity. |
4. Not be matched by the difference of any two channels.
Two strong stations whose frequencies happen to be separated by the IF could produce IF. This unwanted IF generated from such interference will get through filters, which are designed to allow IF to the next stage, such as detector. In other words, the IF should not be 10kHz or 9kHz, or even the multiples of 10kHz or 9kHz. That means any number that is divisible by 10 or 9 is ruled out.
Taking all these into consideration, many frequencies were popular choices during World War I timeframe, in early 1900’s.
As we can see, dealing with higher frequencies, even in the range of a few kilohertz (kHz), was challenging in those early days due to unavailability of suitable components. This resulted in radio makers working with lower frequencies for IF. However, this resulted in poor audio quality when the audio was extracted from very low IF, such as 30 or 35kHz. Then the competition began to build better systems with higher IF to generate better audio qualities in radios.
Many radio manufacturers started making radios in those days with higher and higher IF, such as 100, 155, 300, 455, 500, or even 700kHz! While higher IF produced better audio quality, it obviously resulted in a higher cost.
After World War I came the Great Depression (1929-1939), just before the World War II. It was a period of economic depression in the United States. Naturally, demand for low-cost systems, components, and interchangeable parts in the radios went up. The 455kHz appeared like a good choice considering the good-quality audio it could produce and the lower cost for components compared to other higher IF choices. So, then 455kHz was adopted as a standard by the industry.
| What is Heterodyning |
| The process of mixing two signals, i.e., modulated carrier wave (f1) and the signal generated by the local oscillator (f2), is called heterodyning. Also called frequency conversion and beat frequency generation, it is very widely used in communications engineering to generate new frequencies and move information from one frequency channel to another. The resulting output of the mixer is two signals: addition and subtraction of the frequencies of the two input signals (f1 and f2), one the sum of the two frequencies (f1+f2), and the other the difference of the two frequencies (f1–f2). The new signal frequencies are called ‘heterodynes.’ Typically, only one of the heterodynes is required and the other signal is filtered out. The word heterodyne is coined from the Greek language ‘hetero’ (different) and ‘dyn’ (power). The heterodyning was first developed in early 1900’s. Super heterodyning or superhet. It is basically same as heterodyne, but all the signals involved, more particularly the IF, are beyond the audible range to prevent any interference with the audio signal (higher than 20,000Hz or 20kHz), hence the name ‘supersonic heterodyning’, which is often referred as ‘super heterodyning’ or ‘superhet.’ |
Image frequency
Let us quickly understand what image frequency is, and why and how it should be eliminated.
Let us say an AM radio receiver’s IF is 455kHz. Suppose we want to tune to a radio station broadcasting at 610kHz. That means we need to get the local oscillator (LO) to produce 1065kHz so that when we send LO output of 1065kHz and broadcast frequency of 610kHz to a mixer, the mixer produces the difference of 1065kHz and 610kHz as 455kHz, which is our desired IF.
We have a problem here.
If there is another radio station broadcasting at 1520kHz, then the same mixer with its local oscillator supplying 1065kHz will produce the difference of 1065kHz and 1520kHz as 455kHz, which is same as 455kHz produced in earlier case!
| Typical IF values |
| The value of the intermediate frequency varies based on the application. For example, in standard AM radios, it’s typically around 455kHz, while in FM radios, it’s around 10.7MHz. In television receivers, the IF might be around 38MHz for video and 33.4MHz for audio. |
This means two stations will be extracted at the same time on 455kHz, one from 610kHz and another from 1520kHz.
In this case 1520kHz signal is known as ‘image of the unwanted signal frequency,’ and many times it is simply referred to as ‘image frequency.’
This image frequency needs to be prevented from entering the mixer.
One way to do that is having a tuning circuit to select the incoming RF signals. The tuning circuit may use a variable capacitor (Fig. 5) to select a desired frequency. However, one must manually change the variable capacitor responsible for frequency selection at the RF front-end according to the frequency of the LO. That is cumbersome.
Therefore, we need a mechanism where the capacitor of the tuning circuit should automatically be modified, thereby allowing a particular RF signal of our desired frequency in relative of the frequency produced by the LO. Therefore, many systems have twin gang variable tuning capacitors, which consist of two separate variable capacitors whose value can be varied together (not independently) mechanically (see Fig. 6). Thus, one capacitor is used in RF tank circuit on the RF front-end, and another is used to control the frequency of the local oscillator (LO). When changed, capacitance of both the capacitors is changed simultaneously.
Thus, when the RF front-end circuit is tuned for selecting the 610kHz with the help of one variable tuning capacitor, for example, the local oscillator (LO) will automatically (being mechanically coupled) produce 1065kHz with the help of the second variable tuning capacitor. Since the RF front-end is tuned to pick 610kHz, it is not picking other frequencies, including the unwanted image frequency 1520kHz!
Thus the image frequency has no chance to enter the mixer.
| Image frequency calculation |
| It is well accepted that: Image frequency = Receiving RF frequency + (2 x intermediate frequency) But how? Let us say we have received an RF broadcasting of an AM signal at frequency R1 with local oscillator frequency L. That means: IF=L–R1 Therefore, L=R1+IF The image frequency is the unwanted image of the signal transmitted at frequency R2. When this R2 is mixed in the mixer with the oscillator frequency L, this should also produce IF. IF=R2-L Therefore, L=R2-IF Thus, R1+IF=R2-IF Solving for R2 (image frequency), we get, R2=R1+2 (IF) |
Future designs
Modern microprocessors allow building of a software-defined radio (SDR), also known as software radio architecture, where the IF processing after the initial IF filter is implemented in software. (See https://www.electronicsforu.com/electronics-projects/software-defined-radio-with-android-smartphones.)
Many low-cost FM radios have now incorporated the SDR architecture. With greater advancements in the capabilities of the analogue-to-digital converters and the digital signal processing (DSP) capabilities of microprocessors or the dedicated DSP processors, the incoming modulated carrier signal is often directly sampled instead of first getting converted into the IF and then sampled.
Regardless of the advancements, the journey and the concept of IF and the image frequency are fascinating. They are fundamental and relevant even in our modern telecommunications, both in audio and video.
Janardhana Swamy served as a Member of Parliament in 15th Lok Sabha (2009-2014). He holds an MSEE degree from IISc, Bengaluru, and has worked in India and the US in various engineering and management positions at Sasken, Cadence, Sun Microsystems, Dell, and Cisco Systems. Swamy believes a good education in India can help the nation in building a stronger knowledge-based society and economy
Neha JS is pursuing Bachelor’s Degree in Electronics Communication Engineering at Reva University in Bengaluru. Her interests include mathematics, physics, electrical, electronics, and computer engineering
