I have been wondering how these things work and I've been asking myself questions and trying to figure out answers, but I don't wanna research it since I understand things better if I think on my own. Here they are, and there is some little responses I have came up with

1. How is the frequency of the demodulator wave set and what type of wave is the demodulator wave?
I've figured out that the modulator wave frequency is set by the tuning on the FM radio, and I'm not sure what type of wave the modulator wave is, since this is analog electronics I'm going to assume (please tell me I'm not wrong!) that it is a sine wave.

2. How does the demodulator work?!
I'm not sure how the circuitry is able to extract the carrier wave (music, voice etc.) and output it to the speaker, does anyone know?!?

I also do not understand how you have the air right, which is one medium, and you have like dozens if not hundreds of transmitters operating in the FM radio range, yet by setting the demodulator one can focus on one transmitter, cancelling the rest of it out!?! How is this possible?!!?!

Boy, I wanna be an EE so bad :D
Posted on 2004-05-12 16:52:16 by x86asm
I hope your *good* at trigonometry. If not I suggest your refresh yourself on Sin and Cosin identities, and not just the simple Sin @ = y/r stuff ;)

I will give you some math to chew on in a bit (running out the door at the moment).
Posted on 2004-05-12 17:06:11 by NaN

1. How is the frequency of the demodulator wave set and what type of wave is the demodulator wave?

I never heard of a 'demodulator wave'. Do you really mean the output from the demodulator? Your questions seem to indicate you need to get some background before you inquire about this very large and complex subject. Look at these links. Ratch

http://www.tiscali.co.uk/reference/encyclopaedia/hutchinson/m0030280.html DEFINITION
http://contact.tm.agilent.com/data/static/downloads/eng/Notes/interactive/an-150-1/hp-am-fm.pdf EXCELLENT!
Posted on 2004-05-13 09:33:27 by Ratch

I suggest you visit the links posted by Ratch. You should get a pretty good idea.
Here are my humble dissertations:

1. The frequency of the station is that of a VCO, or voltage controlled oscillator, which indeed produces a sine wave. Its frequency changes according to the a control voltage - the modulating signal (music, speech, etc.). The amplitude of the modulating signal gives the frequency change from center frequency, or frequency deviation. The speed at which the modulating signal changes (its frequency) makes the frequency of the VCO change faster or slower from the center frequency. With no modulating signal, the VCO outputs the center frequency.

The tuning control in the radio is part of an LC tank. Its resonant frequency changes when you move the tuning control (a variable capacitor). At resonance, the LC circuit allows only the signal with a frequency equal to its resonant frequency to pass, attenuating all the others (this is a little more complicated than that, a single LC tank is not enough, so more are cascaded, although they work at another frequency, but this is a topic for another post, maybe).
This is how you select between stations, since each station has its own operating frequency and maximum frequency deviation set (by its operating licence, stations don't just broadcast in any frequency band, these are strictly regulated, this is related to "air right").

2. The signal you need to recover in the radio receiver is the modulating signal (music, speech). The demodulator can be a PLL circuit, as described in one of Ratch's links, or it can be a frequency discriminator.
In essence this is again an LC tank operating around its resonant frequency. When the FM signal is applied to this LC tank, the phase of the output signal changes proportionally with the frequency deviation. This resulting signal is now fed to a phase detector (a couple of diodes and RC filters), whose output voltage is proportional to the phase change, hence it is proportional to the initial frequency deviation. This is the recovered signal or demodulated signal. Amplified and applied to the speakers it produces the sound.

I hope I have shed some light. Math is rather complicated for these things, so I left it out.
Posted on 2004-05-13 11:40:30 by VVV
It is demodulated with a discriminator. It's basically a small phase detector that turns then transmitters frequency variations back into amplitude variations.
You are right that the signal is modulated in the tuning. If we take a free running VFO (variable frequency oscillator) for simplicity and vary the frequency above and below the center or Osc. free running freq. than we have modulation of the true FM type. More commonly you will see a VCO (voltage controlled oscillator) as part of a PLL loop for digital frequency control and audio will be coupled directly through a small RC network to the Varactor in the VCO which creates PM (phase modulation) but is detected by a typical FM receiver the same.
To answer your last question on spectrum or bandwidth then using last example the VCO, the level of audio fed to the varactor will determine the band width in typical audio transmission such as vioce which in voice narrow band operation is limited by FCC to 5Khz but usually is only 3.5Khz. Techniques such as Compression on the transmitter side and Expansion on the receiver side have improved intelligence in communications. Compander is the name of the system (Compressor/Expander).
FM stereo brandcast is 75KHz deviation or 150Khz bandwidth or Swing. Deviation refers to how far the frequency swings either side of center frequency. Compression is absolutely necessary with FM stereo braodcast because typically in order to transmit full fidelity of say a symphony orchestra it's understood that atleast 75KHz deviation is required. This is the amount of intellegence to be transmitted. Sometimes a few extreme high's or extra loud notes may cause the transmitter to over deviate. Consequently on the receiver there is a circuit in the descriminator known as the Limiter. The limiter's job is to ensure the incoming signal being detected does not exceed the specified 75KHZ. If it does than the limitor being a diode/resistor arrangement will exceed it's biasing level and clipping of the signal will occure. Trust me, very unpleasing to the ear.
Selecting the particuliar frequency is done at the intermediate frequency(IF) level. Today much of it is done with DSP(digital signal processing) which is Nyquist theorem digitally imulated filters. Let's stick to basics for now. A good typical FM receiver will have atleast two conversions. That is the incoming frequency will be converted twice. When the RF front end or rf small signal amplifier receives the signal than again you are right that it will receive a whole band of frequencies depending on the filtering between the antenna and the RF amp. The signal will when then be fed to a mixer stage wich is injected with a frequency from the Local Oscillator. The output of the mixer will be the sum and diference of the incoming signal and the local osc frequency plus the original two frequencies. Here is a good place to select the frequency you wish to listen to. For simplicity we will insert a Crystal filter with a bandwidth of the modulated wave. Now everything else is rejected except that which can pass through the filter. We can change that frequency by changing the local oscillator frequency thereby producing new sum and difference frequencies and feeding them into the crystal filter. For more selectivity we often convert the frequency down one more time to say 455KHz by mixing the output signal from the crystal filter with another local oscillator that is set to 455KHz above or below the 1st IF frequency. Usually we will use a cheap ceramic filter at this mixer output which is 455KHz. The key to the sensetivity of the reciever is in the IF amplifiers following the first mixer which produces the first (IF) intermediate frequency.
Posted on 2004-05-16 10:18:23 by mrgone
I just want to point out a few things here.

The 5kHz bandwith actually refers to the permissible bandwidth of the modulating signal (voice, for instance) for AM broadcast. The resulting bandwidth of the modulated radio signal is then 10kHz, since in AM the maximum frequency of the output signal (upper sideband) is the oscillator frequency (note that this is a fixed-frequency oscillator, not a VCO) plus the maximum frequency of the modulating signal The lower sideband (minimum frequency) is the difference of the two.
Thus, the modulated radio signal has a maximum bandwidth of 10kHz, regulated by FCC.

The limiting action in an FM receiver is actually a desirable process that helps eliminate parasitic AM modulation in the signal, since it carries no information. It is the frequency change of the signal that carries the useful information. Because of this limiting action, FM signals are much more immune to noise, since the noise mostly affects the aplitude of the signal, not its frequency. By amplifying the signal to the point its amplitude is limited, any noise-induced AM modulation is removed (try producing sparks by discharging a capacitor close to an AM receiver and then an FM one and check which one "pops" louder).

The typical intermediate frequency for FM modulation is 10.7MHz. 455kHz is a typical IF for AM.
Increased selectivity is obtained by the use of the IF amplifier, due to the fact that it contains many tuned amplifier stages (typically 3 for AM, 4 for FM; each stage introduces noise, but the larger number of stages in FM is possible because of the noise rejection described above).
However, there is generally a tuned stage at the input, ganged-tuned together with the local oscillator, such that the difference is always equal to 10.7MHz (the intermediate frequency). This is a great advantage, since all IF stages are tuned amplifiers but work at a FIXED frequency. Otherwise, you would need to gang-tune them all.
Even better selectivity is obtained by using ceramic filters in the IF amplifier.
Posted on 2004-05-17 11:46:16 by VVV
Broad band amplifiers are more desireable now days. They have no tuning but are excellent impeadance matching devices. They use ferrite in the shape of a toroid becauce of it's self sheilding properties. A good source for them is Amidon in California. Transformer are wound on to these ferrite toriodal forms. Yes 10.7 MHZ is a common 1st IF frequency thow I have seen others. I respectfully disagree with you that 455KHz is only AM. Not true at all.
A typical FM receiver is usually dual conversion superhetrodyne meaning there are to IF frequencies. Usually 10.7MHz 1st IF and almost always 455KHZ 2nd IF. Try transmitting on either one of these frequencies and see if the FCC is beating on your door promptly.
In applications where a high degree of sensetivity is needed such as satelite receivers it is common to have several IF frequencies/stages. Actually it is not a good idea to use LC filters in IF stages because they are so very prone to feedback. This is why when several stages of amplification are needed they are converted and then amplified so there is no danger of feedback to the input of a previous stage.
The actual selectivety of a receiver is performed in the mixer. This is a whole subject or book in it's self. One of the most important components of a high degree of selectivity is a strong local oscillator with a clean signal, free of harmonic content, feeding into the mixer. The mixer must have good intercept points and be resistive to blocking and IMD (intermodulation distortion), You know, Joe alligator station down the block...lol.
Oh one more thing Narrow band 5KHz is used quite alot. I would dare to guess right off the top of my head that all cell phones are probably 5KHz narrow band FM service. The police, fire and rescue,taxies are all 5KHz narrow band FM.
In all fairness VVV your messing with the wrong guy here. I am licensed advanced class ham over 30 years. My call is WA4BJO. I have designed and sold radios across the united states. In fact my tranmitters where rated cleanest on the market in their service class and yes they were PLL controlled with phase sidebands down over 120 DB. One of my favorite challenges is well of course reciever but also RF Power Amplifiers. There are quite a few tricks to designing power amps. Especial Linear power amplifiers which are not required in FM but are an absolute must in SSB (single side band).
Posted on 2004-05-17 13:21:06 by mrgone

I am not trying to mess with anyone. I am just trying to shed some light on a subject.

The original question was about FM radio, in a general form. At least that is the way I understand it.
That would be broadcast FM radio. In these receivers there is rarely a second conversion and the IF is 10.7MHz. It is high because the frequency deviation is rather large (150kHz).
For narrow band FM you can have a second, lower IF, such as 455kHz. That is essential, NARROW BAND, because with such a low IF the bandwidth would have to be really wide to accommodate a large (150kHz) frequency deviation.

Sorry, the selectivity of a receiver is due mostly to the filters in the IF amplifier. Whether these filters are LC or ceramic, it is irrelevant. In part, the selectivity is due to the input filters and the tuned RF amplifier (if it exists).

The role of the mixer is to provide a fixed IF (conversion also adds gain). This allows the IF stages to work at a fixed frequency and be optimized for that frequency. The filters are also designed to work at that fixed frequency, which simplifies things enormously, since they do not need be tuned when you change stations and so their parameters are stable, no matter what station you tune in to.

These are the advantages of the superheterodye.
Posted on 2004-05-18 11:43:16 by VVV
You and I VVV could never have a flame war but only intellectual conversation :grin: I will say this. You would be surprized how much a mixer does have to do with selectivety even though it is gonverned by a crystal filter or the like. Research the subject "Dynamic Range" of a mixer and you will see that it is a very in depth topic. You see if the mixer circuit is not right than you will be subject to harmonics and IMD and even , odd and even order harmonics etc. These unwanted harmonics and spurious emmisions will still pass through the filter because frequency wize they are supposed to be there.
Posted on 2004-05-18 12:33:29 by mrgone
I absolutely agree with the fact that the mixer is very important, as is the local oscillator, as far as harmonics are concerned.
Any harmonics that get into the mixer will produce unwanted signals, some of which may fall into the bandwidth of the IF amplifier and so they cannot be distinguished from the desired signal. They will be amplified just like the useful signal.
Posted on 2004-05-19 11:19:23 by VVV
Maybe I was going a little into depth for the original question...lol. Anyway I would like to compare notes with you if you've bee envolved in rf design. I don't care what they think about spectrum/realestate. I say there is still a good sized frontier out there in the areas of new types of modulation techniques and also and this is a biggy, BETTER DESIGN. I am convinced that alot of these cell phones are thrown together with cordless telephone ICs and sloppy engineering and this is the reason for so much crap on the air waves.
I'm currently that is in my spare time designing some new PLL techniques for finer resolution with greatly decreased phase noise. Obviously the more energy Xmitted in phase sidebands,harmonics and spurious emmisions greatly reduce transmitter effeciency and clutter up the air waves for much needed realestate. Not to mention that the propagation of a spectrally pure note is far superior to Xmitted garbage. I think the cell phone industry needs to back up a bit and take a breather and do it right this time rather than the usual "quick! produce and make a buck, the CEO doesn't have a billion dollars yet" if you catch my drift. Since these phones are battery operated than possibly you could appeal to the big corporations by showing them higher efficiency and longer lasting battery capacity.
If you ever watch America's Most Wanted then you probably have seen them crying about the cop frequencies being redered useless more and more and they think the only answer is realestate/spectrum. I think we know otherwize and I have contacted them and proposed a contract that could be paid for in law suits. I need the equipment to pinpoint the colprits of this spurious garbage but I know I'm right because to put it simply. Cell phones should in no wize be interfering out of frequency like this and I'm sure the FCC has been taking a few pay offs.
What's your opinion on all this and do you see a market there? After all if your like me we do work for a living.
Posted on 2004-05-19 13:25:05 by mrgone
WOW this is an interesting topic, scary for me since I seem to understand digital circuitry much better than analog, but this is an interesting topic nonetheless, I will definitely do some research in it and hopefully you guys can verify what I have picked up is correct, thanks all for your help!
Posted on 2004-05-19 15:48:07 by x86asm
Your best resource to understanding analog electronics is your Calculus book.

Study up 1st and 2nd order derivative equations (homogenous and non-homogeneous). Once you get it and start to see the mathematical 'pattern' that seems to repeat itself from the text examples, you can turn around and apply it to analog electronics in the sense that each 'part' in the patterns are very closely mirrored by an electrical component (ie. Inductors, Capasitors, and Resistors). << Note: 2nd order differentials have three terms, and I just pointed out three unique principals (wink wink, nudge nudge) :alright:

I realy like analog electronics, so if you have questions, ask :P


PS: I was going to post math for you to absorbe, however Ratch's reply with the link marked "EXCELLENT" it pretty well what I was going to post. I suggest you review that one closely!
Posted on 2004-05-19 23:10:38 by NaN

The 5kHz bandwith actually refers to the permissible bandwidth of the modulating signal (voice, for instance) for AM broadcast. The resulting bandwidth of the modulated radio signal is then 10kHz, since in AM the maximum frequency of the output signal (upper sideband) is the oscillator frequency (note that this is a fixed-frequency oscillator, not a VCO) plus the maximum frequency of the modulating signal The lower sideband (minimum frequency) is the difference of the two.
Thus, the modulated radio signal has a maximum bandwidth of 10kHz, regulated by FCC.

Actually the bandwith of a signal as rated by FCC is the width of the signal being radiated from the antenna. In AM it is composed of the carrier (center or oscillator freq) and two sidebands (upper and lower). The intellegence is in the two sidebands. So the bandwidth is all three of these components. The width of the two sidebands will depend on the percent of modulation. Either the upper or lower sideband can be used to recieve intellegence. In fact WARC (world amatuer radio commision) part of the UN is working with FCC to enforce all AM broadcast stations to eventually go to single side band reduced carrier since only one side band is needed. This will require receivers to have the appropriate circuitry to recieve this signal.
A typical AM xmission will be approx 6KHz wide with 100% modulation. 2Khz for carrier and 2Khz for each sideband. Since the sidebands are what contains the intelligence the carrier is not necessary so years ago hams come up with the ballanced modulator circuit that could remove the carrier and leave you with pure double sideband. Later through phasing techniques and then ultimately sharp filters such as a Xtal filter they were able to remove one of the sidebands, upper or lower it doesn't matter as long as the receiver is tuned for upper or lower reception in conjuction with the particuliar transmitter. The reciever actually reproduces the carrier on it's end by what is known as the BFO (beat frequency oscillator). This a hetrodyne process to recreate the original signal and be detected like normal AM. SSB is an AM signal.
Posted on 2004-05-20 15:21:02 by mrgone