RF cavities and related devices

[artis]

Your last post was unclear to me at first but then on a second though are you trying to tell me by the water analogy that at RF and microwave frequencies the cavity is nothing but a "tank" whose geometry and parameters allows for the wave to "slosh" aka resonate and that just as one cannot identify a spring and a mass in a resonating water tank one cannot identify the individual parts in a cavity unlike a traditional LC circuit formed from inductors and capacitors ?

Would I be better off as viewing the cavity as a special sort of waveguide whose geometry is such that the wave is like a standing wave just resonating back and forth between its two possible states (E field and B field) and due to the geometry the E field forms between the capacitor like plates while the B field forms when the charge runs back and forth through the torus in between the E field cycles?

So a modified waveguide would achieve the same resonance even at the same frequency as a RF cavity just the difference is that in a typical waveguide it would not be possible to isolate out the E field as homogeneous in an area big enough to do any useful work on passing charges ?
I hope I'm shooting in the correct direction.

 

[sophiecentaur]

Would I be better off as viewing the cavity as a special sort of waveguide whose geometry is such that the wave is like a standing wave just resonating back and forth between its two possible states (E field and B field) and due to the geometry the E field forms between the capacitor like plates while the B field forms when the charge runs back and forth through the torus in between the E field cycles?

I would say you have stated it the wrong way round. A Waveguide is a particular form of Cavity which happens to be of the right dimensions to support a traveling wave with no reflections . A cavity can also be designed so that it has a high E field at some local place in it and that is good for accelerating charged particles. There are many examples in Science where it is better not to take Classification so far that it begins to drive understanding. It can be useful to classify things, of course but . . . .

 

[artis]

What is the average electron density in the electron bunch that forms before the output cavity of a medium power klystron due to velocity modulation at the input cavity for example?

Also what would happen to these electron bunches if one removed the output cavity of the klystron and simple made the tube end into a vacuum chamber where there also wouldn't be any confining magnetic field , I mean how far the electron bunches would travel with their gained kinetic energy before they spread out alot? Sort of like imagine klystron being turned into an electron gun but instead of a typical electron gun where it shoots a stream of electrons this one would shoot bunches of them with some distance between each bunch. I simply wonder how far such bunches would be able to travel before significant divergence would happen and also what factors would determine this?

 

[sophiecentaur]

Also what would happen to these electron bunches if one removed the output cavity of the klystron

The electrons would just end up on the collector, as they do in the normal course of events. I wound;t imagine that passage through the output cavity would have much defocussing effect on the beam.
You could calculate the electron density by looking up the beam current for the klystron in question and seeing what the diameter of the beam tube is (for a minimum estimate), I would think. The distance between peaks of density would be the beam velocity X frequency of the RF. All these figures would depend on the particular klystron and application but ball park figures would not be hard to calculate. If you want better than ball park figures you would need to burrow down to a deeper level of design criteria.

 

[Klystron]

This document from SLAC (Stanford Linear Accelerator) provides klystron output data at high energies. Equations do not begin until ~page 5.
http://www.slac.stanford.edu/cgi-wrap/getdoc/slac-pub-9557.pdf

FTR capital letter Z represents impedance in the calculations, a measure that (roughly) combines aspects of capacitance (which you requested), inductance, and resistance. Small letter "e" or function exp(x) and inverse function ln() refer to logarithms. Note the grave danger operating klystrons without correct shields and protection.

 

[artis]

well sophie I was thinking not just removing the output cavity I was also thinking in moving the collector further away and then having an estimate at how far (in cm for example) the electron bunches would go (in free space outside the drift tube without a "steering" B field) before they start to spread out significantly. I was asking here because I lack the sophisticated math skills to arrive at a number myself.

by the way here is why I was wondering , I though about a scenario where you put two modified klystrons 180 degrees oppositely to each other , keep the electron source but the output cavity is not driven but rather it works as an accelerator cavity being fed from another powerful RF source (klystron) the two oppositely faced klystron cavities then synchronized in such a way as when one accelerates electron bunches toward the other the other can then "catch" them and push them back in the opposite direction and so these bunches of electrons oscillate back and forth between the two cavities. while the traditional klystron gun and input cavity makes up for lost electrons from the oscillating bunches

Maybe I'm into Sci-Fi but in IEC they have the problem of keeping the electrons (the potential well) confined in the middle due to the charge and low mass of the electron so I was wondering what would the the result of instead forming a traveling potential well on a vertical axis.

 

[sophiecentaur]

FTR capital letter Z represents impedance in the calculations

Impedance of what? It's confusing me. Is it the Impedance that the electron beam loads the cavity with? (Or perhaps the source impedance of the modulated beam?)
The "Z + 1.2" bit is interesting. It implies that the "impedance" must be fairly low, for 1.2 to be significant. Perhaps the impedance to the alternating currents in a beam of charged particles is, in fact, very low.

 

[sophiecentaur]

well sophie I was thinking not just removing the output cavity I was also thinking in moving the collector further away and then having an estimate at how far (in cm for example) the electron bunches would go (in free space outside the drift tube without a "steering" B field) before they start to spread out significantly. I was asking here because I lack the sophisticated math skills to arrive at a number myself.

I found this link which has some of the ideas that you have been looking at. There is an optimum length of amplifying section of a klystron. For a given beam purveyance (look it up), there is a maximum possible gain. The bunching process carries on along the length of a drift tube but the density modulation is soon non-sinusoidal and, eventually, fast electrons overtake slow electrons (over-bunching). The waveform starts by looking a bit like ocean waves, that are peaky and eventually can 'break'. Harmonics, by then, can be higher level than the fundamental - not what's wanted.
You say "I lack the sophisticated math skills`' so there is little point in trying to extrapolate into 'better' designs based on arm waving ideas. The first thing that will hit you is Bessel Functions, which describe the spectrum of the signal carried by the beam. Look through the link I posted. It gives some practical information about beam devices in general.

 

[artis]

I assume you meant perveance not purveyance? because searching purveyance google talks about crown princes etc

So per Wiki article on perveance in my own words, it is the ratio of how much current can be attained from a fixed PD across a fixed length and diameter due to electrons repelling each other and wanting to spread out (space charge effect) , does this sound right?
So in order not only to increase electron current but also stay at the same level of perveance, a larger Cathode-anode voltage is required?

 

[sophiecentaur]

I assume you meant perveance not purveyance?

That's auto spell checking for you. lol.
Apart from the anthropomorphic "Want to spread out", I would go along with that.

 

[Klystron]

Impedance of what? It's confusing me. Is it the Impedance that the electron beam loads the cavity with? (Or perhaps the source impedance of the modulated beam?)
The "Z + 1.2" bit is interesting. It implies that the "impedance" must be fairly low, for 1.2 to be significant. Perhaps the impedance to the alternating currents in a beam of charged particles is, in fact, very low.

@sophiecentaur, artis seems stuck on capacitance as a measure in his high-energy RF configuration while documents describe distributed impedance including the safety documents I included. Slowly, but surely, artis is describing a lab setup that would produce at least x-rays.

If artis or 'RF engineer friends' apply power to the rig s(he) describes with the shields removed so they can adjust the internal components of the dual "basketball player sized" drift tubes without calculation and, of course, after disabling safety interlocks most likely it will arc a/o burn out their RF sources. I can envision two fixes but cannot condone anyone conducting high-energy experiments without any math skills or understanding basic electronics such as impedance. Safety first.

Apologies if I have misjudged but any real seeker of electronic knowledge would at least study basics before designing high-E RF cavities, no?
@artis, please describe the klystrons you are working with and to what purpose. Thanks.

 

[artis]

I am not working with or designing a klystron, those that I posted where theoretical ideas as it is easier for me to learn through ideas to which I can then find solutions like the kid who puts the square in the square hole as then he learns how to fix certain things. I have repaired some old CRT's and vacuum tube amps and I think I have a basic understanding of how an ordinary vacuum tube works so that is as close as I'm standing to electron beams for now.
Sure klystrons non-linear way of operation and the RF cavity in itself is a whole different animal species which I only recently began to appreciate so take it slow with me.

So is such a setup theoretically possible where there are two oppositely positioned klystrons with externally pumped output cavities so instead of working like a RF source they would work like an electron accelerator and could the bunches be made such that they could oscillate back and forth between the cavities if timing is correct etc or would there be a large divergence of the electron bunches after they pass each cavity that only few of them would return while the rest would impact the opposite cavity walls and other surroundings? I already said why I'm asking this, just a curious idea.

 

[sophiecentaur]

So is such a setup theoretically possible where there are two oppositely positioned klystrons with externally pumped output cavities so instead of working like a RF source they would work like an electron accelerator and could the bunches be made such that they could oscillate back and forth between the cavities if timing is correct etc or would there be a large divergence of the electron bunches after they pass each cavity that only few of them would return while the rest would impact the opposite cavity walls and other surroundings?

That sounds to me like the daftest idea I have come across in many years. What are you planning to do with the DC components of the beam currents? How would the 'bunches' serve to augment each other? Just draw yourself a diagram of a simple two cavity klystron and figure out where its beam would go so as to merge with the bunches from a second klystron in such a way as to induce the same E field in a cavity. Where the beams cross, what are the electrons doing and, quite importantly, what job could they do for you? Would it be a 'just for fun' experiment?
You seem to want to treat this sort of equipment a bit like lego models.

 

[artis]

Well from what I read and which wasn't told to me here nor is loudly said in other places is that even though the electrons bunch up and form clusters by the end of the drift tube at the output cavity which is why there can be a high frequency AC field in it in the first place does not mean that there are no electrons in between the bunches. I saw the diagrams of the peaks and averages of the electron beam in a graph and then I understood that it seems there is a fair amount of steady static electron beam current and the peaks sort of ride on top of that.

Sure if my experiment was to have any merit the klystrons would have to be heavily modified there could be no steady electron current and the operation would have to be fully pulse like , the timing would have to be precise etc etc which is why it probably is a bad idea and not attainable in reality as you seem to suggest.But ok, suppose I have a RF device of high power and I can indeed create these discrete pulses which are precisely timed with an electron gun that replenishes for the lost electrons could I then have my electron bunch oscillate back and forth between two RF cavities in a free (vacuum) space inbetween them without the bunch spreading out so much as it is completely lost, sure it probably has to do with the strength of the accelerating field and distance of the "free space" between the cavities , or would this distance be very very small if no confining B field and drift tube is present?

 

[sophiecentaur]

But ok, suppose I have a RF device of high power and I can indeed create these discrete pulses

That's just a supposition. Do you not realize how difficult it is to produce pulses of high energy electrons? The purpose of a Klystron is RF Amplification and not to produce pulses of electrons. To do that, you wouldn't include an output cavity, would you? An output cavity will tend to spread the pulses out again, once the electrons have passed because it extracts RF Power carried by the beam. To produce pulses the thing to use is a control grid. The spectrum of the RF signal carried by narrow pulses is wide band and not what's needed from a klystron amplifier.
I really don't think there's much point in continuing this thread because you don't seem to have read about klystrons; you seem to have made up your own story about them.

 

[artis]

No I wasn't thinking about using the output cavity as an output cavity as it is normally used in a klystron instead use it as an accelerator cavity being fed by high power RF. Maybe to make it more clear say just an accelerator which instead of continuously accelerating electron/proton bunches forward oscillates them back and forth between two such cavities , that is all I wanted to know and this in no way ruins my knowledge or learning about klystrons, why be so pessimistic all the time?Also one more thing. In high frequency power supplies, the transformer for example is wound with litz wire , due to skin effect in order to make a larger surface area for the same cross area of wire I can understand this.
An RF cavity seems to be fabricated out of ordinary solid copper sheet metal that is bent etc but not made from multiple isolated copper parts, the same goes for waveguides , So please explain why here seemingly skin effect doesn't exist anymore?

 

[sophiecentaur]

use it as an accelerator

The output cavity would need a supply of RF power for this. You are talking about the techniques that are used in particle accelerators now.

So please explain why here seemingly skin effect doesn't exist anymore?

A wound component can usefully be made of litz wire when there is a long electrical path (as in an MF inductor in a receiver tuned circuit ). Litz wire is not used for physically large components- even the coils in mf transmitters. In a waveguide, which direction would you lay the strips of wire when the current could be flowing in many different directions as the wave passes? Instead of asking a question like that, why not read about it and get yourself informed before asking? Don't expect to get educated by an endless process of Q and A on PF. That will not generate good will here.

 

[Baluncore]

@artis.
Perhaps you should go back to some early RADAR engineer training references.
The “MIT Radiation Laboratory Series” covers equipment used in the first quarter century of RADAR.
You might start with Volume 7. Klystrons and Microwave Triodes.
The series archive can be found at the Jefferson Lab site; http://www.jlab.org/ir//MITSeries.html

 

[artis]

Thank you for the link, I downloaded most of the PDF files ,began reading about klystrons and microwave triodes

 

[artis]

the books in the link are rather old so some things have changed so just to sum up in modern times we use semiconductors like gunn diodes and transistors for low power RF and microwave amplification while we (due to technical reasons) resort to using klystrons, TWT etc for medium to high power RF and microwave amplification, now that being said I suppose that an average cell phone radio tower station doesn't use klystrons or TWT but semicoductors for feeding the antennas? I wonder what about tv, from what power level for the antenna usually semiconductors are abandoned?One more thing, due to the nature of, for example, klystron's velocity modulation which later translates into electron bunching "down the road" I feel like a klystron is best suited in amplifying sinusoidal waves but how good would it be with a different type of wave structure? I guess what I'm asking is doesn't the noise to signal ratio at output increase dramatically if other than sinusoidal signal is attempted to work with in a klystron, how about TWT , I read traveling wave tubes also work on the velocity modulation principle.?

 

[sophiecentaur]

I feel like a klystron is best suited in amplifying sinusoidal waves but how good would it be with a different type of wave

How about an analogue TV waveform or a DTV multiplex? Both signals occupy a total bandwidth of 7MHz at UHF frequencies of hundreds of MHz? Linearity limits the maximum signal level but there is no substitute for a 20kW UHF Klystron that I know of. TWTs can handle higher bandwidths but the powers tend to be less. But what do you actually want from this device and why?
Engineering is all about identifying a need and satisfying it as cheaply and as efficiently as possible. It is not about modifying a device and then looking for something to do with it.
"Signal to Noise" ratio at the output of a transmitting amplifier is Enormous. The signal to noise ratio of a single CW wave can also be enormous if you don't modulate it and are prepared to have a 0.01Hz bandwidth receiver but what would it be used for? You seem to be just using word / idea salad.

 

[Baluncore]

I feel like a klystron is best suited in amplifying sinusoidal waves but how good would it be with a different type of wave structure?

sophiecentaur is right. If the wave structure of microwave transmissions were not sine waves they would interfere with other transmissions. The limitation is on active element efficiency where bandwidth or Q is determined by the coupling networks employed around the active element.

Voyages of exploration are always more exciting and fascinating when less information is available, which tends to be the way of the freestyle inventor. Going around in unproductive circles can be fun for a while, a bit like doing cryptic crossword puzzles, but when an engineer needs to make something more reliable they go to the library and survey the literature. The advantage of old books such as the MIT Rad Lab series is that they contain surveys of historical fields that help us avoid wasting time and energy reinventing the wheel. They provide a solid foundation on which to understand and build.

There will always be advances that make older generations of technology redundant. The line between semiconductors and thermionic devices is dynamic and blurred in time, so it depends on application and your viewpoint. The gyrotron has extended power levels above 10 GHz, well beyond any semiconductors. https://en.wikipedia.org/wiki/Gyrotron

 

[artis]

thank you for the answers but I think to myself why are you assuming on my part and cornering me a little in your posts , sure I'm not saying my approach to learning is the best one it's just the one I have. Since I am polite and respectful I can't see why someone would be bothered with my questions.

I was simply asking about the types of waveforms these devices can best amplify because I thought about the possible limitations of the process that does the signal amplifying which is the bunching of electrons due to velocity modulation.
So sophie, you've said you worked with television etc back in the day , I tried googling the waveforms did not found much, so for NTSC signal or analog Tv in general it was both AM and FM modulated? So from the viewpoint of a klystron that has to "pump" an antenna I can understand the AM part where you take a fixed high frequency and by increasing or decreasing it's amplitude create a lower frequency waveform on "top" of it, that can be extracted by appropriate filters but how then the FM part comes in? At some specific AM peak in the right moment a burst of low amplitude but high frequency FM is put in? At least judging by the waveforms I saw the FM part seemed higher in frequency than the AM carrier?
The whole information into waveform coding is quite complicated so I'm not trying to learn television signals for now I am just curious of how a klystron can effectively with somewhat decent S-N ratio amplify all these minor ups and downs of an analog tv signal or does it do almost ok given that both the AM and FM parts are sinusoidal in nature ?
Although I might be giving electrons less credit than they deserve in terms of how fast they can react (velocity change wrt time) in order for these changes to be noticeable in the output cavity

 

[Baluncore]

@artis.
I think you are missing a key point. An RF transmission is a sinewave because that is all that will fit through the power amplifier (PA) or in the band. For a 500MHz carrier signal, the 2nd harmonic at 1GHz would be in another band. The important thing is that the power amplifier and the channel have sufficient bandwidth to pass the modulated carrier. For a 5MHz wide signal at 500MHz, the bandwidth need be only be 1% of the carrier frequency, so the Q of the amplifier and tank circuit will be about 100. None of the carrier harmonics are needed, so there is no point generating or radiating them. Any carrier harmonics generated in a radio transmitter PA must be attenuated before they are radiated.

An AM sinewave carrier will vary slowly in amplitude as it is modulated. An AM power amplifier needs to be linear and be able to generate an output amplitude that is proportional to input amplitude.

For FM it is the period of the sinewave that is important. The power amplifier only needs to produce a fixed amplitude sinewave with frequency varying slowly in time with the input signal. The PA does not need to be linear, it can be class C that kicks in time with the input frequency. Any carrier harmonics that result will be attenuated by the narrow bandwidth of the PA output circuit and antenna.

 

[sophiecentaur]

Voyages of exploration are always more exciting and fascinating when less information is available, which tends to be the way of the freestyle inventor.

And it accounts for the shelves full of non functioning projects that can be found in the sheds of many 'inventors'. It can become a habit to start a project and then abandon it before fruition. I have a feeling that your definition of 'freestyle' is probably not mine. Edit:[ You are not the kind of person who would launch out on a project without a good level of knowledge, I'm sure.]

, I tried googling the waveforms did not found much,

Analogue TV is not going to appear on the first page of many Google searches because it is out of date. If you look for the waveform of NTSC or PAL TV vision signal, you sill see the basic shape. The line synchronising pulses are very obvious and they correspond to maximum RF signal power because 'negative modulation' is used.
Your use of "Signal; to Noise" is still not very appropriate and I have no idea what you mean by "minor ups and downs". They are what constitutes the signal. A CW signal on its own is pretty useless for carrying information. CW is used in other applications (for instance, particle accelerators).
You really should settle down a bit and address one topic at a time. It's the only way you will actually learn anything. So far, in this thread, you seem to have addressed the major part of what I was involved in over decades of work and I wonder what you have actually taken on board. Scattergun Q and A is soooo inefficient.

 

[artis]

By minor ups and downs I meant the FM part of the analogue tv signal , I saw a bunch of captured scope images of the signal and between the AM modulated main signal one can see a much smaller in amplitude but higher in frequency signal "riding" the AM modulated signal at certain spots.
Yes i could identify the synchronizing pulses , having the highest peaks.

As for what I've learned from this topic, actually alot, now I have a basic visual and technical understanding of how a klystron and a TWT works, where they are mostly used etc I also learned AM radio and FM some time ago but with TV even though the technique is similar everything feels more complicated, also this helps me understand other fields that I'm pursuing at the moment better as EM physics is all sort of "interlaced" just like the analogue tv signal frames.

As to what Baluncore said ok I understand RF signals are just AM or FM modulated sinewaves both the carrier and the "carried" signal itself.
I have always had some problem with understanding bandwidth, I do understand the basic concept but ok here's a question, if we use a carrier frequency that is optimized for best Q in a particular klystron, (when you talked about power amplifiers I assume you referred to klystrons) would it then matter for the efficiency of the klystron whether it amplifies a 10Hz AM modulated signal onto a high frequency carrier or 20khz modulated one as in both cases the klystron beam is velocity modulated at some high frequency just the amplitude changes differ wrt to time?
When you said an "AM carrier will vary slowly in amplitude as it is modulated" then I have to ask what is the frequency spectrum of analogue tv both picture and sound if used without the high frequency carrier or in other words what is the actual transmitted frequency for a single channel for example?
One more thing, so digital TV over air transmitted through the same devices as analogue doesn't have the luxury of having square wave PWM signals passing down parts of a PCB so I assume they use the same sinewave RF signal with FM modulation? Is the digital info then transmitted with the help of FM where it is later (at the receiving end)decoded into a PWM signal (in a satellite decoder box or one attached to an antenna)? like higher frequency parts translates into longer square pulses while lower frequency parts into shorter pulses with more "dead" time and these pulses are then interpreted into 1 and 0 in order to make a useful signal at the end ?
I know this covers many topics but while we are at it I want to know, thanks:)

 

[sophiecentaur]

Endless rambling threads with multiple topics are not PF style. If you have a specific question then ask it in a new thread. You have a better chance of an answer if the thread title relates to your question.

 

[artis]

agreed too many questions in one thread make a mess but too many threads for simple matters also make a mess , hard to make it right.
can you at least please then go on and answer this last post of mine here and any other related questions I will then ask in new thread.

Thanks.

 

[sophiecentaur]

threads for simple matters

None of these matters are "simple". Each one is worthy of several threads in itself.
Your last post has a half a dozen questions that do not have one word answers.
Whatever gave you the idea that this is simple stuff?

 

[Baluncore]

@artis.
You need to thoroughly understand amplitude modulation and channel bandwidth before you try to reason further about RF transmission systems. To do that you need to understand what Fourier analysis and synthesis are, and you need to be able to think about the same signals in both the time domain and the frequency domain.

If you do not have the basic concepts necessary to discuss the situation, the exploratory path you are on will be frustrating and inefficient for all. You need to be interested in and study signal processing to acquire the concepts needed to reason about signals. You cannot ignore, deny or skip the subject. Nor can we deliver those concepts as footnotes in this one thread.

 

[Tom.G]

I have always had some problem with understanding bandwidth,

Bandwidth (needed): The highest frequency of the information-carrying signal.

carrier frequency that is optimized for best Q would it then matter for the efficiency of the klystron whether it amplifies a 10Hz AM modulated signal onto a high frequency carrier or 20khz

In a practical sense for this discussion, No. Using examples that were introduced above, a 500MHz carrier and a Q of 100, the available bandwidth would be 5MHz.

what is the frequency spectrum of analogue tv both picture and sound

The video signal is about 3MHz bandwidth. Tha audio channel is 15kHz, the same as FM radio. Since the audio is FM modulated on the TV signal, it is actually an additional carrier (sub-carrier) transmitted as part of the overall TV signal. This Audio Subcarrier is 4.5MHz and is frequency modulated to ±25kHz deviation from from 4.5MHz.

In an analog TV receiver, the 4.5MHz audio signal has it's own IF amplifier chain separate from the video. The audio chain feeds a conventional FM detector just as you find in an FM radio, and emits the audio signal.

Now, please get back to getting an understanding of the basics and limiting threads to one or a very few questions.

Hope this helps some.

Cheers,
Tom

p.s. I had the same problem you seem to be having trying to reach an understanding. I got into electronics as a teenager living in a rural town. There was not anyone around to answer my questions, so I turned to books and experimenting. When I ran across something new I would find a book or magazine article that gave a more basic level of the subject; this was my approach to tracking down the 'How' and the 'Why' of difficult subjects. If something was just too complicated, it would be put aside and revisited occasionally over several months... eventually becoming understandable as I learned more of the underlying stuff. Perhaps not the 'Ideal' approach but it worked with what was available.

You seem to be trying to approach PF more as a personal tutor than we here are comfortable with. Doing your own research to answer your questions will lead you to a much better understanding of the field. An alternative is taking some formal instructional courses either on-line or in person, somehow I don't see you as embracing this option though.

So start digging out answers/understanding yourself and when you have a particular rough spot or hard time about something, post your specific question on PF; after all, that's what we do here!

 

[sophiecentaur]

Bandwidth (needed): The highest frequency of the information-carrying signal.

It's actually the Range of frequencies [Edit: occupied] by the signal. A 700MHz TV signal will have a bandwidth of around 7MHz, the same as if the signal is mixed down to an IF frequency (say 45MHz) or up on a microwave carrier of many GHz. "Bandwidth" would often mean "Channel Bandwidth" in practice. The "Baseband Bandwidth' would refer to the highest frequency that the Video circuits would handle - it would be the trace on a 'scope of the signal that the Receiver extracts from the RF signal. The term is also used for the rate of data in a digital signal. Bit rate would perhaps be a better term but Bandwidth is the popular but misleading term that's used.

The video signal is about 3MHz bandwidth. Tha audio channel is 15kHz, the same as FM radio. Since the audio is FM modulated on the TV signal, it is actually an additional carrier (sub-carrier) transmitted as part of the overall TV signal. This Audio Subcarrier is 4.5MHz and is frequency modulated to ±25kHz deviation from from 4.5MHz.

For NTSC and PAL colour TV, the frequencies involved are a bit higher than this. A 6MHz sound carrier is used, with similar FM characteristics to Sound FM Radio and the colour (Sub-carrier) is at around 4.43 MHz on PAL.

@artis Read Read Read is the only way to get into this subject. It is vast!

 

[artis]

Well I actually understood AM modulation as done in AM radio for some years now. To put it simply, you take a high frequency CW carrier of sinusoidal shape (for efficiency transmitting over large distances) you take ordinary audible spectrum "voice" from a microphone and you make this low frequency audible waveform control the amplitude of the high frequency CW carrier , feed the result into a transmitting antenna then pick it up with a receiving one add a filter and you get back your low frequency modulated wave without the high frequency carrier.As for bandwidth the part I don't seem to get is this I think, say that you AM modulate your signal with some high frequency carrier , audible frequencies range up to about 16-20kHz but for this question assume they are up to 100 Khz , I still could modulate the whole spectrum from 0 to 100Khz on some high (Mhz region signal) right? Even though modulating it would require a broader bandwidth because the frequency spectrum is larger?
So what I am asking is probably this, if my carrier frequency is say 100Mhz (round numbers for simplicity) then why my bandwidth would only be a fraction of those 100Mhz instead of me being able to modulate everything from low audible few Khz up to signals that come close to the carrier frequency itself?

Wiki says that "Bandwidth is the difference between the upper and lower band of frequencies in a continuous band of frequencies"
So why can't the difference be all the way from zero to close to the carrier frequency itself, now I have a feeling it has something to do with harmonics and transmitter and receiver circuits having limited ability to "latch" on to certain frequencies without having to physically alter the circuit rather than the physical limitation of a signal over air itself ? Also I think that it might have to do with the fact that certain electronics are designed to be more efficient at their design frequencies (because one can't make a device "for everything") so in theory a klystron and a transmitter could AM modulate everything from zero to the carrier frequency but most of the band would be very inefficient and there would be "noise" problems at the receiver etc , am I getting this right?

Trust me I read its just that the way I formulate my questions I can't find answers easily to them without probably reading few dozen books but please don't judge.

 

[sophiecentaur]

So why can't the difference be all the way from zero to close to the carrier frequency itself,

Because that would take up the whole of the spectrum so you could send only one signal over a particular wire or radio connection. If you "understood" the modulation used for AM radio, wouldn't that be clear to you? You have heard about Radio Channels, no doubt. Each channel will be a few kHz wide and uses a different central carrier frequency.
Read all of the Wiki article about AM and try to avoid skimming between highlights. You can't just extrapolate at random from Sound Radio and assume that comments translate straight to the operation of a klystron. A Klystron cavity is tuned to a narrow band of, say 1% of the carrier frequency. How would you expect it to transfer RF power onto or off an electron beam? Have you ever heard of an MF Klystron? Why would that be?

am I getting this right?

More aimless Q and A here. It just doesn't work this way. Ask a specific question on a separate thread and you may get a useful answer.

 

[artis]

at this point your simply angry. Ofcourse I know it would take up the whole "channel" I wasn't asking the question with respect to how commercial radio stations do their business I was curious from the point of physics, but instead of a simple answer you are teaching me a "lesson" again
different carrier frequencies for each channel is I think so that the radio could tune and latch onto each different station with the help of heterodyne.

I understand that the physical size of a rf cavity determines the frequency range it is suitable for but in AM modulation of a single carrier wave for example the frequency the cavity operates is fixed isn't it? just the amplitude changes so as long as there could be this one mystical radio station why couldn't they use the whole spectrum from zero up to carrier frequency the klystron should be fine with it?

So from a klystron point of view (since this is a thread about them) the klystrons that amplified the lower frequency AM channels had to be physically bit different (size, diameters etc) than the ones amplifying the high end of the AM? how about Tv?
Also the lower frequency would require more power put into the antenna for the same signal strength at distance x away?I will make new thread but honestly I have no guarantee that in this new thread you will show more empathy towards my curiosity.