Wednesday, April 8, 2009

FM Radio Systems by Lee Bond, N7KC

Educational Radio Net, PSRG 46th Session

This week let’s take a look at the typical FM radio system from both the transmit and receive standpoint.

The first step is to understand just what FM or frequency modulation really means. Here is the situation… we have some information that we would like to pass along to others. Assume that this, so called, ‘baseband’ information is voice generated and that the frequency content extends from about 300 hertz to 3000 hertz. In principle we could transmit baseband information directly however the antenna presents problems that are huge given the very long wavelengths associated with audio frequencies and the enormous physical size of any wire array that could do the job. A better scheme is to take advantage of smaller antennas associated with higher frequency ‘carrier’ waves and impress the baseband information onto the carrier in some fashion.

There are three ways to modify a carrier wave using baseband information. We could change the instantaneous amplitude keeping the frequency constant using a process called AM or amplitude modulation, we could change the instantaneous frequency keeping the amplitude constant using a process called FM or frequency modulation, or we could change the instantaneous phase with respect to some reference phase keeping the amplitude constant using a process called PM or phase modulation.

For this discussion lets choose just AM and FM and contrast the differences. Amplitude modulating the carrier as in AM produces sum and difference frequencies on a one to one basis with the baseband information. Assuming a carrier frequency of 10 MHz we would see lower sideband energy extending downward from 9.999700 MHz to 9.997000, the 10 MHz carrier, and upper sideband energy extending upward from 10.000300 MHz to 10.003000 MHz. The required bandwidth is 6000 hertz. The carrier amplitude never changes in AM so, to 100% modulate the carrier, we must add audio power equal to ½ the carrier power and ½ of this audio power ends up in the USB and the other ½ ends up in the LSB. SSB as in single sideband suppressed carrier is a subset of AM and the bandwidth is reduced by slightly more than ½ to 2700 hertz.

The bandwidth required when using FM is somewhat different and is related to a couple of ideas called Carson’s Rule and modulation index. Modulation index is defined as the maximum carrier deviation from nominal resting frequency divided by the highest modulating frequency and Carson’s Rule is a rule of thumb used to estimate the RF bandwidth required to transmit baseband information at a given deviation. In fact frequency modulating a carrier generates an infinite number of sidebands but about 98% of the radiated energy is contained within the limits of Carson’s Rule. As a practical matter, if the modulation index is less than 1 and in the .7 to .9 region then the FM sidebands look very much like what would be observed in an AM signal and the required bandwidth is about twice the highest modulating frequency. On the other hand, if the modulation index is larger than one then Carson’s Rule takes over and the bandwidth is about twice the sum of the deviation and the highest modulating frequency.

Amateur FM transmitters are deviation limited at 5 KHz and if the highest modulating frequency is about 3 KHz then the modulation index is 5/3 or 1.67. Carson’s estimate of the bandwidth required would be 2(5+3) or 16 KHz. This example shows that FM with a modulation index of 1.67 requires almost 2.7 times as much bandwidth as AM transmitting the same baseband information. FM is pretty much relegated to the VHF bands and above where adequate spectral space is available. A more interesting example is given by commercial FM where the highest modulating frequency is 15 KHz and the maximum modulation index is 5. In this case Carson’s Rule shows that the required bandwidth is 2(5*15+15) or 180 KHz.

Wideband FM enjoys an enormous signal to noise ratio which is the primary advantage over AM. FM also allows for transmitting DC baseband information as well, and was widely used in data telemetry before digital techniques came to fore. Old timers will remember the Ampex multi-track FM tape recorders used to collect data down to DC.

The deviation produced by a baseband audio signal is linearly related to the amplitude of this baseband signal. If you want the maximum audio ‘punch’ from your FM transceiver(s) then make sure that you talk loud enough to deviate your equipment to the maximum. All modern transceivers are deviation limited so there is an upper limit to the effects of shouting. If normal use of your equipment yields low audio at the receiving end and the receiver is known to be properly set up then you might want to change the deviation sensitivity setting in your transceiver.

To summarize operation of the FM transmitter… we know that the RF generating oscillator is running at constant amplitude but its instantaneous frequency is linearly related to the amplitude of the modulating signal. For all modulation indices, especially those greater than one, there are an infinite number of sideband energies produced but about 98% of the sideband energy is within the bandwidth defined by Carson’s Rule. Unlike AM the carrier power in FM is redistributed into the sidebands and, in fact, the carrier power can go to zero (called a Bessel zero) for certain deviation ratios. Deviation ratio and modulation index are defined the same except modulation index is based on maximums and deviation ratio is not.

Now let’s look at the typical FM receiver to see how it functions and make some contrast between AM and FM receivers.

Both AM and FM receivers would typically be superheterodyne type instruments. This means that the incoming signal is mixed with a local oscillator to produce a difference frequency which is called the intermediate frequency or IF for short. Unfortunately single conversion receivers suffer from image problems so one solution is to use two mixers and two IF’s and select the 2nd IF frequency such that the input image is out of the 2nd IF passband. These receivers are called dual conversion. The beauty is that these IF amplifier strips run at a single tuned frequency and offer sufficient bandwidth to accommodate the requirements of reconstructing the original baseband signal. Audio leveling in the AM receiver is accomplished by controlling the gain in the IF stages by the use of AGC as in Automatic Gain Control. AGC is a feedback signal developed in the receiver detector and then applied to the earlier IF stages to control overall receiver gain. AGC is very likely associated with your S meter display since the two go hand in hand. Detection in the AM receiver is very straightforward and can be as simple as passing the signal through a rectifying diode. Single Sideband Band suppressed carrier systems, a subset of AM, use a more complex detector since one must restore the missing carrier to make the audio intelligible.

Unlike the AM receiver, the FM receiver has no RF gain control… automatic or otherwise in the IF sections. The FM front end runs wide open with maximum gain at all times. The characteristic and very loud rushing sound produced by FM receivers when the squelch is opened is a result of front end receiver noise being amplified in the absence of any incoming received signal. The typical FM detector (known as a discriminator) is sensitive to both amplitude and frequency changes so, when no ‘on frequency’ RF is present, the detected noise has a strong varying amplitude component. The underlying principle is to amplify any incoming signal to the point of saturation so that amplitude variations are leveled or limited and thus will not produce any output from the FM detector or discriminator. This is the origin of the term ‘quieting’ which is so often mentioned when describing incoming signal amplitude.

Quieting sensitivity was once the defining measure of receiver sensitivity. ‘On frequency’ RF energy was piped into the antenna port and increased in amplitude until the open squelch noise was reduced (quieted) by 20 db. In voltage terms 20 db is a factor of 10 so 20 db quieting is the same as reducing the noise to 1/10 of its original value or a 90% reduction. At this point one noted the microvolts of RF to produce this noise reduction and tagged the receiver with this value which was typically in the sub microvolt region. Note that 20 db quieting is not the same as full quieting. A noise reduction of 40 db or a factor of 100 would be more in the region of "full" quieting. Quieting sensitivity has been replaced by a scheme to determine the number of microvolts required to produce a 12db SINAD relationship between noise and distortion in FM radio receivers.

There are several schemes for recovering baseband audio from frequency modulated RF signals. The Foster-Seely discriminator and ratio detector circuits appeared early on and are very effective in converting frequency changes into analog audio. Slope detection by using a miss-tuned AM detector is possible as well. The phase locked loop or PLL detector is an excellent example of the modern approach to audio recovery.

Post detection circuits in both AM and FM receivers use standard linear audio stages with appropriate filtering and gain controls. State of the art receivers are now using digital filtering, called DSP for Digital Signal Processing, in the audio stages and sometimes in the IF amplifier stages as well. The squelch circuit associated with the FM receiver is activated by receiver noise and the threshold where the squelch switch is activated is controlled by a manually operated pot or potentiometer in most cases.

This concludes the set up discussion for FM radio systems. Are there any questions or comments with regard to tonight's discussion?

This is N7KC for the Wednesday night Educational Radio Net

No comments: