Receiver Circuit Question -- What role does this antenna capacitor play?

[The Tortoise-Man]

Xc, is the coupling capacitor reactance, which is in series with the antenna.
ZL = Ra + ( Xa + Xc ) j; the effective load impedance.
The magnitude of the load vector is then compared to the reactance of one component in the tuned circuit.

Right I, see, and here since we assume that the whole circuit is running at the resonance frequency of the tuned circle, the (amounts of) reactances of the coil and capacitor of the the tuned circuit coincide, so it doen't matter to which one we compare the magnitude of the load impedance, right?

 

[Baluncore]

Right I, see, and here since we assume that the whole circuit is running at the resonance frequency of the tuned circle, the (amounts of) reactances of the coil and capacitor of the the tuned circuit coincide, so it doen't matter to which one we compare the magnitude of the load impedance, right?

Right.
At resonance, XL = -XC ; They are equal and opposite, so the sum is zero.
I use XL when I can, because it is positive.

 

[The Tortoise-Man]

I have another question closely related to previous discussion about the usage of the
word Q factor in this context.

Say we are dealing generally with a receiver or transmitting circuit and
are going to discuss about it's matching issues. I read several times
in online texts on constructions of hobby radio receivers that one talks in this
context about THE Q factor of the receiver/transmitter.

Now by definition the factor Q in general can be associated pure formally to any
partial subsystem of the circuit, ie e.g. to the antenna, to the tuned circuit
(like in previous discussion), but even to any capacitor or coil.

My question is if generally the author talks in text dealing with such receiver/transmitting
circuit about THE Q factor of the receiver/transmitter without
explicitely mentioning to WHICH partial component of the receiver/transmitter circuit
this Q is associated, may it conventionaly be assumed that implicitely it is in almost all cases
adressed to the Q factor of the tuned circuit of the receiver/transmitter like in your answer?

The point is that in a lot of sites dealing with designing receiver/transmitter
networks I saw several times the authors talking about the Q factor without
explicitely explaining to which part of the circuit the Q factor is
associated to.

So this leads me to conjecture (based as well on our previous discussion where you used to impose the matching condition only Q factor of the tuned ciruit and not that one of the antenna) that by convention when one talks about THE Q factor of a receiver/transmitter network one means the Q factor of the tuned circuit. Is this correct?

 

[Baluncore]

The point is that in a lot of sites dealing with designing receiver/transmitter networks I saw several times the authors talking about the Q factor without explicitely explaining to which part of the circuit the Q factor is associated to.

Can you give a link to an example of that usage.

Have you considered the possibility that it might refer to the Q of the entire signal path ?
A cascade of identical tuned circuit transfer functions will raise the resultant Q by the power n.

 

[tech99]

The Q factor can apply to any component, or a group of components such as a tuned circuit. It relates stored and dissipated energy, or reactance and resistance.

 

[The Tortoise-Man]

Can you give a link to an example of that usage.

Have you considered the possibility that it might refer to the Q of the entire signal path ?
A cascade of identical tuned circuit transfer functions will raise the resultant Q by the power n.

Yes, eg here (https://physics.stackexchange.com/q...tity-for-electrical-oscillations-transmitting) (V.F.'s answer) or here (https://www.toppr.com/ask/en-gb/question/conceptual-questionsa-explain-how-the-quality-factor-is-related-to-theresponse-characteristics-of-a-radio/)

Yes, to consider the Q of the complete circuit seems reasonable. But the point which still confuses me is that in previous discussion about tunning of the simplified receiver circuit from #19 the only Q factor that there mattered was the Q of the tuned circuit (the "LC tank"), see the imposed conditions in #25 which the desired Q have to fullfill.

There you only focussed only the Q of the LC tank as patial component, not that one of of the total circuit (ie of antenna + coupling cap + LC tank). How are these reasoning are related to considerations above?

If one is interested on the improvement of the receiver's selectivity, should one focus on the Q of the total receiver circuit (="the entire signal path") or do is it suffice to deal with the Q of the LC tank of the circuit and "ignore" the other parts (seemingly, that's what we did in case of circuit 19)?

 

[Baluncore]

If one is interested on the improvement of the receiver's selectivity, should one focus on the Q of the total receiver circuit (="the entire signal path") or do is it suffice to deal with the Q of the LC tank of the circuit and "ignore" the other parts (seemingly, that's what we did in case of circuit 19)?

We focused earlier on the tank Q only, because that was the only resonant module present in the simple receiver system. We needed to reduce the dampening of that oscillation to improve the Q, or we would hear all the channels at once, like many discussions in a crowded room.

Q is a bandwidth concept. It can be applied to any tuned energy storage module, or to the system as a whole. Q is a measure of transfer function bandwidth. The transfer function of the system is the product of the transfer functions of all stages in a cascade.
https://en.wikipedia.org/wiki/Transfer_function

Frequency tuning of radio, “syntony”, was invented in about 1897 by Prof Oliver Lodge. That made it possible to have several radio transmitters operating at one time in the same region. Resonance also improved the sensitivity of the receivers, and so made longer range communications possible.

 

[The Tortoise-Man]

We focused earlier on the tank Q only, because that was the only resonant module present in the simple receiver system. We needed to reduce the dampening of that oscillation to improve the Q, or hear all the channels at once, like many discussions in a crowded room.

But the antenna itself is strictly speaking a resonant module of this receiver system as well. Why nevertheless in the discussion before the Q_A of the antenna not really matters and so can be swept under the carpet?

 

[DaveE]

Be careful when talking about Q with RF guys. While they (nearly) all understand the basic concept, they live in a world where that term is used in analogous or sloppy ways. Like "the Q of an inductor" or Q as a measure of bandwidth in a systems that are more complicated than a single resonance. These aren't bad ideas, but they also aren't the strict definition of Q, which, in its many forms, is only a measure of the energy loss of the natural response of a single resonator (a simple harmonic oscillator, to physicists). This often can be confusing if you don't know the assumptions behind their simplifications. If the circuit under question can't be modeled as a SHO, like an LCR circuit, then Q doesn't strictly apply, and assumptions have been made along the lines of "similar to a SHO", "with the impedance at the resonant frequency", "when used as part of a SHO", etc.

 

[Baluncore]

But the antenna itself is strictly speaking a resonant module of this receiver system as well.

For the broadcast band, the antenna is not resonant and so has a very poor Q.
For receive the antenna must work over the whole band without tuning.
The Q of the tank is therefore critical to selectivity.
The two stages, antenna and tank, must not be tightly coupled, or they will share low Q.
That is why there is a very small coupling cap between stages.
Your OP title was "Receiver Circuit Question -- What role does this antenna capacitor play".

 

[Baluncore]

Be careful when talking about Q with RF guys. While they (nearly) all understand the basic concept, they live in a world where that term is used in analogous or sloppy ways. Like "the Q of an inductor" or Q as a measure of bandwidth in a systems that are more complicated than a single resonance.

At a specified frequency, an inductive component will have a reactance and a series resistance. From that the maximum Q of any circuit employing that inductor at that frequency can be determined. The inductor does not have to be used in an LC tank circuit. Recognising when high Q inductors are needed is essential to designing RF circuits.

The rise-time of a signal in a receiver is an inverse function of bandwidth, or more generally Q. That is true of just one stage, or a cascade of very many stages.

That is not a sloppy use of Q, but an understanding that the concept of Q is applicable to all levels of the universe, not just simple resonators.

 

[DaveE]

At a specified frequency, an inductive component will have a reactance and a series resistance. From that the maximum Q of any circuit employing that inductor at that frequency can be determined. The inductor does not have to be used in an LC tank circuit. Recognising when high Q inductors are needed is essential to designing RF circuits.

Yes, this exactly illustrates my point. You can put a bound on the maximum Q, with inductor losses. Provided you have specified the resonant frequency (which of course determines the associated capacitance). But resonators also have a capacitor which can have losses that will in part determine the actual Q. Granted inductor losses typically dominate in RF circuits. But I see no reason, in a pedantic sense, that you couldn't say the same thing about capacitors and their losses. This is exactly the sort of underlying assumption that is common in the RF community.

As I said, y'all aren't wrong, you're sloppy. In a normal resonant circuit the concepts of Q, Z_o, ω_o aren't that complicated and all are important and necessary. But if your design assumes 50Ω, or a specific band, you have fixed some of those and it's convenient to speak of "high Q" inductors and such. Most others would be happy with R and L, instead of L, Q, and ω_o.

the concept of Q is applicable to all levels of the universe, not just simple resonators.

Applicable or useful? I think we'll have to agree to disagree on this point. But ultimately it's a question of how you define things.

If a cat walks across your piano keyboard exciting many strings, you could define the piano Q with the concepts of energy stored vs. energy lost. But I'm not sure that's a useful concept and I am pretty sure that you would be the only one that would do that. You wouldn't be wrong, but you also wouldn't facilitate efficient communication with others without explaining that you aren't using the common meaning.

The rise-time of a signal in a receiver is an inverse function of bandwidth, or more generally Q. That is true of just one stage, or a cascade of very many stages.

I can see that you would think that's useful. In my world if the system dynamics aren't described with a Laplace transform like $\frac{1}{(1+\frac{1}{Q} (\frac{s}{\omega_o}) + (\frac{s}{\omega_o})^2)}$, an SHO, then you are using a definition that is complex enough to require elaboration. If you have a 2 stage high Q BPF, you can pretend that it has a single value for Q, that wouldn't be a bad model, but it would actually have two values for Q, one for each quadratic pole pair. If you doubt that, write out the equations, the math is clear in this case.

Again, I'm not saying that RF EE jargon is wrong. It wouldn't be so common if it was. But it's not that easy to understand from the outside until you understand the assumptions. It isn't something you'll ever learn as a physical science undergrad, at least not at the schools I went to. It isn't always understandable from basic principles.

My point was simple and, I think well demonstrated by these last few posts. Watch out if you are speaking with RF EEs about Q. It's not the simple version the rest of us normally use.

 

[Baluncore]

If a cat walks across your piano keyboard exciting many strings, you could define the piano Q with the concepts of energy stored vs. energy lost.

I would not do that, I am not an idiot.

 

[DaveE]

I would not do that, I am not an idiot.

Sorry. Hyperbole. I know your not an idiot. I don't respond to idiots. The longer my rebuttal, the more I respect whatever you said that I thought wasn't quite right. I this case, I don't even think your wrong, just using a different language than some of us.

 

[alan123hk]

Traditionally and customarily, it seems that the Q factor is mainly used for a single second-order LC tuning circuit because the Q factor has a simple mathematical relationship with the components of the single tuning circuit.

It seems that relatively few people will say what the Q factor of a very complicated circuit is, for example, what is the overall Q of a multi-stage tuned amplifier. I believe it is because the situation of a multi-stage tuned amplifier is more complicated. Its overall frequency response is related to the center frequency and bandwidth of each amplifier stage, but these parameters may vary for each amplifier stage. Although we can still define the overall Q factor of this multi-stage tuned amplifier based on overall the center frequency and bandwidth, etc., a question of interest may be whether it contributes to the actual engineering design process.

Of course, I am definitely not saying that certain things should not be or are not so correct, because the method of engineering design is defined and continuously innovated by different people according to the needs of the application.

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

 

[Baluncore]

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

 

[The Tortoise-Man]

Be careful when talking about Q with RF guys. While they (nearly) all understand the basic concept, they live in a world where that term is used in analogous or sloppy ways. Like "the Q of an inductor" or Q as a measure of bandwidth in a systems that are more complicated than a single resonance. These aren't bad ideas, but they also aren't the strict definition of Q, which, in its many forms, is only a measure of the energy loss of the natural response of a single resonator (a simple harmonic oscillator, to physicists). This often can be confusing if you don't know the assumptions behind their simplifications. If the circuit under question can't be modeled as a SHO, like an LCR circuit, then Q doesn't strictly apply, and assumptions have been made along the lines of "similar to a SHO", "with the impedance at the resonant frequency", "when used as part of a SHO", etc.

Firstly let me try to clarify your concern on the terminology & usage of Q:

Let consider an abstract circuit consisting of several components
(like resistors, coils, caps, etc).

My understanding of Q is that it can be associated to every "elementary"
component (by "elementary" I mean a component of the circuit which cannot
be property decomposed in more elementary components; eg a single capacitor; on the
other hand, eg a LC-tank, can of course be properly decomposed in more
elementary components).

But of course the Q can also be associated to EVERY submodule of the circuit as
well, like eg the total circuit itself or LC-tank or antenna as subsystem,
formally that's always possible. And it's clear to me that since
in general of course a circuit may contain non self-oscillating
components (eg a real resistor), Q should in most general setting a priori
NOT interpreted as something what has to do with bandwidth, in most
general setting Q of a component or submodule should be defined
as quotient of stored energy and dissipated energy per cycle, as you already said.

As I understand it correctly, the interpretation of Q as the reciprocal of the bandwidth
can clearly only applied to Q associated to submodules which
if we keep it sloppy/informal have the properties of a "swinging" system, so which
posses the potential ability to resonate. (of course "everything" can swing,
but let focus on modules which swinging properties are commonly
studied in electrical engineeing.) So the later is a "additional interpretation in special cases", not the main def.

As you noticed with his cat example this sloppy formulation can be
taken to absurdity, but that's a "pathology", let's next focus again on
usual receiver circuits and
the circuit from #19 as a "toy model", where as the common sense "swinging" submodules
should be regarded the antenna, the LC-tank, but the total receiver circuit as well, so
in that case we only limit ourselves to these components the concept
of "bandwidth" is well defined.

 

[The Tortoise-Man]

For the broadcast band, the antenna is not resonant and so has a very poor Q.
For receive the antenna must work over the whole band without tuning.
The Q of the tank is therefore critical to selectivity.
The two stages, antenna and tank, must not be tightly coupled, or they will share low Q.
That is why there is a very small coupling cap between stages.
Your OP title was "Receiver Circuit Question -- What role does this antenna capacitor play".

May I try again to reprase and extract your main point?

If we have a general receiver and we are interested in improving it, then
naively we can always decomope is in "antenna" and "rest part" including
eg tuned LC-tank, demodulator, filter etc:

And of course one important part of the tuning procedure is to
adapt the right receiving properties on our purposes. And these mathematically
are rougly characterized by Q, so we want modify the receiver circuit
by modifying the reveriver's Q, right?

Naively, as in case #19, we have different Q's (which allow
in reasonable sense this "bandwidth interpretation"): the Q_A of the antenna,
the Q_T of the total receiver system (ie with antenna),
the Q_r of the "rest part" (receiver without the antenna), but
we can also of course as I said associate a Q to internal submodules
of the rest part, ie to the LC-tank like in #19.

So do I understand it correctly that theoretically it is
"recommendable" to work with the Q_T of the total receiver system, but
in practice in order to adapt the right bandwidth of the receiver
the considerations on antenna can be neglected, so instead of working
with Q_T, one tends to working with Q_r of the "rest system" simply
"ignoring" the antenna part completely?

So the question is about effort and benefit: how much better our
receiver system with become, if we try to adapt our Q_T with respect
the total system or if it already suffice to with with Q_r for our purposes,
assuming that in practice it's easier to work with rest system's Q_r.

In other words the question is how much better the receiver system would
become if we would try to improve Q_T, instead of Q_r. If the proper improvement
is low, then it suffice to work with Q_r.

Is this what you essentially wanted to say in #45? So that the considerations
on antenna's Q are formally reasonble, but in practice dispensable (probably just because it usually requires some additional effort, if we as you said additionally try to adapt the Q of the antenna and the rest circuit together ), so it suffice to focus on the Q of the rest part of the receiver ingnoring the antenna?

Did I interpret you correctly?

 

[DaveE]

My understanding of Q is that it can be associated to every "elementary"
component (by "elementary" I mean a component of the circuit which cannot
be property decomposed in more elementary components; eg a single capacitor; on the
other hand, eg a LC-tank, can of course be properly decomposed in more
elementary components).

You may want to rethink or rephrase this. From a modelling point of view, the elementary components you speak of are the ideal R, L, C. i.e. a capacitor defined by $i = C \frac{dv}{dt}$. Q makes no sense at all for these ideal and most elementary components by themselves.

For real world components, they really act more like networks, often resonant in some places. So a real capacitor is a resonant network sometimes with its parasitic ESL and ESR. In this case Q makes perfect sense and for simple models of the network, or at specific resonances, is well defined and useful.

Q only makes sense when associated with a resonant structure of some sort.*

* although mathematically 2 well separated real poles can be described with very low Q. In this case the poles will be at $Q \omega_o$ and $\frac{\omega_o}{Q}$

 

[Baluncore]

Q only makes sense when associated with a resonant structure of some sort.

To you maybe, but this is an engineering forum, discussing a 100 year old radio receiver circuitry. In this field, high-Q and low-Q components are identifiable real components. Here, every component has a loss resistance, so you can never have a pure reference inductance. If you want to enforce discussion to mythical perfect components, then you should get out of the way of the engineering, and restrict that discussion to the physics and mathematics sub-forums.

 

[Baluncore]

May I try again to reprase and extract your main point?

My main point is that there must be very little coupling between the antenna and the tuned circuit. Your model of a resonant antenna, tightly coupled to the tuned circuit, is simply not applicable here.

There is only the one tuned circuit, but it is placed inside a feedback loop. The Q of a regenerative receiver is not fixed, it is decided by the degree of regeneration provided. The feedback is adjusted to compensate for almost all the resistive losses in the tuned circuit. That makes the circuit the original Q-multiplier.

Obviously, without feedback the circuit will have poor sensitivity and selectivity. The receiver will be deaf if the tuned circuit cannot be isolated from the antenna.

Feedback regeneration is equivalent to raising the transfer function of the tuned circuit to a greater numerical exponent. Too much feedback and you have a transmitter.

 

[The Tortoise-Man]

In case of the regenerative receiver that's exactly the purpose
of a Q-multiplier, this is an additional mechanism to rase the Q of the
"bare" receiver, that might be a tuned circuit in most simple case
or surely something more complicated, but the effect of this feedback loop
is as far as I understand always the same here.

If we consider
the selectivities of this regenerative receiver or the toy circuit
in #19, we observed that in both cases the Q factor of the
antenna almost not matters for the selectivity of these receiver
circuits, only Q factor of other components matter.
This leads naturally to following general question
towards Q factor of the antenna:

Is in a general receiver the Q of the antenna nearly irrelevant
for receiver's selectivity issues? The reason why I'm aking
this is that on several sites (like this one:
https://turbofuture.com/industrial/...r-What-Affects-It-and-How-Do-You-Calculate-It)
the Q factor of an antenna is discussed in full details, but
on the other hand following this discussion I come to conclusion
that in receiver circuits seemingly the Q factor of the antenna alone
plays does not play any significant role. Is that correct?

Or in what types of receiver circuits the Q factor of the antenna
conveys an important contribution to receiving quality?

 

[DaveE]

To you maybe, but this is an engineering forum, discussing a 100 year old radio receiver circuitry. In this field, high-Q and low-Q components are identifiable real components. Here, every component has a loss resistance, so you can never have a pure reference inductance. If you want to enforce discussion to mythical perfect components, then you should get out of the way of the engineering, and restrict that discussion to the physics and mathematics sub-forums.

Sigh... Did you actually read that post? I was also saying that ideal components aren't really applicable when discussing "the Q of an inductor". I made a point of discussing that real components are best modeled as networks of ideal components.

Anyway, I've hijacked this thread for too long now. I think I'm done. If you can't understand, or don't want to consider my point about subject matter jargon, then I don't think more discussion from me will help.

However, I will repeat, ONE MORE TIME, I am not attacking your methods or insulting your intelligence. I know you know this subject. I was simply pointing out that some of us speak of these concepts differently, and that readers need to be careful to translate properly. I'm sorry if you feel the need to defend yourself, that was not my intent.

 

[Baluncore]

In any case your reply was good, and my post was off target, confusing. Notation soup isn't helpful, but immittance can be left out of that recipe too.

I remember earlier, how you wanted to censor the one indirect mention of the term "immittance". You will just not let it go. As a self appointed censor, you have caused sufficient chaos in this thread.

However, I will repeat, ONE MORE TIME, I am not attacking your methods or insulting your intelligence. I know you know this subject. I was simply pointing out that some of us speak of these concepts differently, and that readers need to be careful to translate properly. I'm sorry if you feel the need to defend yourself, that was not my intent.

Censorship of language or pronunciation is a common bullying technique, often used by bureaucrats to disempower a representative member on a committee. I am not saying that is your intention, but I have to wonder why you persistently insist on only the puritan language of physics, on an engineering forum.

Anyway, I've hijacked this thread for too long now. I think I'm done.

Good move.

 

[jsgruszynski]

Which role plays an additional capacity in a receiver circuit between the antenna and the matching-box part like in this example (found in https://www.frostburg.edu/personal/latta/ee/twinplex/schematic/twinplexschematic.html):

View attachment 290649

Is it necessary for this receiver circuit or just optional? What is it's proper usage?

It's for tuning the antenna to its resonance.

 

[Baluncore]

It's for tuning the antenna to its resonance.

Then can you explain why the capacitor is not adjustable, it is fixed in value, yet it works with any antenna connected.

 

[tech99]

It is true that some antennas can be brought to resonance with a series capacitor, which effectively shortens the electrical length, but in this case it is not the purpose of the capacitor. It is undesirable to have a resonant antenna with a regerative detector, because it causes large variations in the amount of feedback required as the tuning is varied. It can even result in complete suck-out frequencies where oscillation cannot be obtained.

 

[The Tortoise-Man]

It's for tuning the antenna to its resonance.

But for the tuning in the linked circuit there is the LC-tank (ie coil L_1 + cap 10-140 pf)
responsible, or not? Or do you mean that the intention for adding of this coupling
capacitor 2-18 pf a just a jazzing up the tuning part of the circuit? So not really
neccessary, but makes it more powerful? Nothing more?

 

[tech99]

The capacitor is essential because without it the tuned circuit is heavily damped by the antenna and oscillation cannot be obtained.

 

[Baluncore]

The key parameter is the ratio of the antenna coupling capacitor to the tuning capacitor. That ratio decides (in part) the proportion of the resonant current that is diverted from the tuned circuit to the antenna, and so the Q of the tuned circuit.

If a variable capacitor is used to tune the resonant circuit, then the Q will change over the band as the tuning capacitor changes.

In effect, the changing voltage of the tuned circuit, pumps current through the two capacitors in proportion to their capacitance values. C = Q / V ; Q = I·t ; ∴ i = C · dv/dt.

Because the antenna impedance is in series with the antenna coupling capacitor, the antenna impedance may reduce the proportion of current lost to the antenna, and so increase the Q of the tuned circuit.

 

[The Tortoise-Man]

The key parameter is the ratio of the antenna coupling capacitor to the tuning capacitor. That ratio decides (in part) the proportion of the resonant current that is diverted from the tuned circuit to the antenna, and so the Q of the tuned circuit.

Maybe here is a reason for my confusion:
You are talking about the "Q of the tuned circuit".
Firstly, what do precisely mean by
the "tuned circuit" in the image from
https://www.frostburg.edu/personal/latta/ee/twinplex/schematic/twinplexschematic.html?

I thought that the part of this receiver which is called
the "tuned circuit" is only the LC-tank given
by the parallel L_1 and capacitor with 10-140 pf below on the
right?

If my last statement is true, then I not understand
why as you said the Q of the tuned circuit (this LC-tank)
depends also on the coupling capacitor.

Indeed, the Q of a parallel LC-unit has a concrete formula
(see https://en.wikipedia.org/wiki/Q_factor#RLC_circuits);
note that of course the formula requires that the
LC-tank should also have a resistor; I guess that
in the linked image it is implicitely assumed to be there
(eg real coil has always a additional real resistance) but not
explicitely drawn).

And due to this formula it seems that according to it there
is no dependence of the tuned circuit's Q to the coupling capacitor.

So what do you mean by that the ration of the
antenna coupling capacitor to the tuning capacitor
(clearly it depends on the coupling capacitor)
decides (indirectly) the Q of the tuned circuit?

Maybe when we are talking about "tuned circuit" we are thinking
of different parts of the total unit? What is the
"tuned circuit" for you there?

 

[Baluncore]

If my last statement is true, then I not understand why as you said the Q of the tuned circuit (this LC-tank) depends also on the coupling capacitor.

Because the resonant LC tuned circuit has many different ways to lose energy, and the most significant one of those is to the antenna, via the coupling capacitor. Another is the resistance of the L and C components, then there is eddy current loss in the nearby chassis from the inductor's magnetic field, and some radiation from the inductor. A small amount of energy goes to drive the VT grid. In the real world, Q is not just the simple LCR textbook equation, it is one corner of the universe.

There are also two sources of energy that maintain the LC resonance. One is the signal from the antenna, the other is the regenerative feedback that makes up for many of the losses.

 

[The Tortoise-Man]

Because the resonant LC tuned circuit has many different ways to lose energy, and the most significant one of those is to the antenna, via the coupling capacitor. Another is the resistance of the L and C components, then there is eddy current loss in the nearby chassis from the inductor's magnetic field, and some radiation from the inductor. A small amount of energy goes to drive the VT grid. In the real world, Q is not just the simple LCR textbook equation, it is one corner of the universe.

There are also two sources of energy that maintain the LC resonance. One is the signal from the antenna, the other is the regenerative feedback that makes up for many of the losses.

Interesting. So what you mean is that the Q factor
of the tuned circuit considered as the red encircled
thing in image below highly depends on what stuff is in the
magic box on the left, right?

But finally what about this quoted formula from
https://en.wikipedia.org/wiki/Q_factor#RLC_circuits
which suggests that the Q of parallel LC-tank depends
only on components L, C, R? So it's finally just
a poor's man approximation tool for some special
cases? Or, let me say, under
which assumptions on the total network this formula
gives "nearly" the correct Q of the LC-tank?

 

[Baluncore]

Your irrational line breaks make it hard to understand your reasoning.

If the only losses were the series resistance of the inductor, or the dielectric losses of the capacitor, then the wikipedia simplicity would be correct.

But in the real world there are other networks in parallel with L and C that lower the Q. There is no advantage in having an isolated tuned circuit if it does not have energy inputs and energy outputs.

 

[DaveE]

Interesting. So what you mean is that the Q factor
of the tuned circuit considered as the red encircled
thing in image below highly depends on what stuff is in the
magic box on the left, right?

View attachment 292703

But finally what about this quoted formula from
https://en.wikipedia.org/wiki/Q_factor#RLC_circuits
which suggests that the Q of parallel LC-tank depends
only on components L, C, R? So it's finally just
a poor's man approximation tool for some special
cases? Or, let me say, under
which assumptions on the total network this formula
gives "nearly" the correct Q of the LC-tank?

Personally, I don't see much value in dealing with the Q of a resonator if you haven't included all of the loss elements. I suppose you could; it's all just definitions, I guess. But I would be careful about drawing arbitrary boundaries. The way the tank works depends on all of the "stuff", not just what's nearby in the schematic.