Generating TV spectrum in LTSpice XVII...?

[brainbaby]

Can some one tell me how to generate following signal on LTspice XVII..

Thanks..

 

[sophiecentaur]

That is, in fact, not a "spectrum". It is a template to show what parts of the spectrum can be occupied over some time. If you want a representative spectrum of a coded TV picture (PAL or NTSC) plus sound, you need to take a suitable picture, amplitude a carrier and filter it with a vestigial sideband filter. You can then add a 6MHz carrier which is modulated (a bit more straightforwardly) with, say a high amplitude tone. remember, the TV picture consists of fields at 50 or 60 Hz, all of which may be different.
If you want a more simple signal to test a channel with then you need to have an idea what aspect of the channel you want to test.
This link shows a typical spectrum of an NTSC signal modulated on a carrier. You may like to do a search using terms like "typical TV signal spectrum". I found that one fairly easily but you may need to spend some time getting a better one.

 

[brainbaby]

I correct myself...its a single channel rather than a spectrum...

If you want a representative spectrum of a coded TV picture (PAL or NTSC) plus sound, you need to take a suitable picture, amplitude a carrier and filter it with a vestigial sideband filter.

Below is a more feasible diagram

I am looking a .asc & .plt files for the above signal...?

 

[Baluncore]

It would be difficult to program a fake TV signal generator by building the signal up from all it's spectral parts.
It is hard to work out how to approach a solution without knowing why you need to do it.

The file.asc will contain the circuit under test, with a source of excitation.
Once you have a circuit defined you can excite it with a current or voltage source having a value of AC 1.
You then do an AC analysis and plot the circuit response across frequency with a command like .AC oct 1000 50MEG 72MEG.

The file.plt specifies which variables you display. You can invoke an FFT of a plotted variable from the plot pull down menu once you have a time trace in the plot window.

Why do you want that signal?
Can you attach the circuit you are wanting to test?

If you have an LTspice.asc circuit, you can rename it as LTspice.asc.txt and attach it to your next post.

 

[sophiecentaur]

Below is a more feasible diagram

But it is not a spectrum diagram. It is a template.
As Baluncore says, you need to specify what you ( the OP) actually wants and why you (think you) want it. It is not hard to produce a sample TV spectrum by finding the base band spectrum of an actual 525 (/625) signal. That spectrum is, of course, dynamically varying over the TV field and line waveforms.
I am always wary of the 'simulation' approach to these things because they don't involve the sort of reality check that you get with real equipment and real signals. You have to be very careful about drawing any conclusions from the numbers from a simulation. (Garbage in garbage out)
NTSC and PAL are definitely 'quart in a pint pot' systems that only work because of the redundancy of the information in most scenes. They were specifically designed to 'get away with' a very limited channel bandwidth. The colour information is very fragile and its presence can cause horrible patterning effects, even in a good linear channel.
PS Google "NTSC test signals" to widen your knowledge and to confuse yourself further.

 

[brainbaby]

Why do you want that signal?
Can you attach the circuit you are wanting to test?

Hi Baluncore,

You may remember that we have discussed the circuit in my previous post https://www.physicsforums.com/threads/vhf-tuner-circuit.893615/ ...There I was trying to find out that by varying the value of L4 the bandwidth varies i.e larger the value of L4 larger is the bandwidth as an output.
So I thought that it would be great if I test it out practically, so I made a circuit in LTspice XVII and tried to simulate it by changing the value of L4 at each attempt...but unfortunately I didn't found any noticeable change in the output of the circuit and that led me to think that may be its the input signal which is causing this dissimilarity as the text is strictly according to the input signal which is mentioned in post 3 of this thread...
So here I am probably looking for help...

 

[sophiecentaur]

@brainbaby: you have really confused me now. The thread you reference is about an FM sound circuit. Why would you need to include the whole TV signal. If you want to get the FM sound carrier from the TV (i.f.) signal then the most popular way is to use 'inter carrier demodulation' which puts the whole TV signal through a non linearity and filters out the 5.5MHz component (difference between sound and vision carriers).
You really must be more forthcoming about what you actually want to do here. We are still pretty much totally in the dark. Do you actually understand the comments I (and others) have been making about your original diagram and what it really represents.

I made a circuit in LTspice XVII and tried to simulate it by changing the value of L4 at each attempt...but unfortunately I didn't found any noticeable change in the output of the circuit

What change did you expect? FM demodulation is not an intuitive topic.

 

[Baluncore]

So I thought that it would be great if I test it out practically, so I made a circuit in LTspice XVII and tried to simulate it by changing the value of L4 at each attempt...but unfortunately I didn't found any noticeable change in the output of the circuit and that led me to think that may be its the input signal which is causing this dissimilarity as the text is strictly according to the input signal which is mentioned in post 3 of this thread...

The circuit under discussion comes from fig 22.17 of “Monochrome and Colour Television”. by R.R. Gulati. 2009. ISBN 8122416071. I would not expect the value of L4 to have a critical value in that wideband VHF TV tuner.

With Spice you simulate circuits, not signals. An AC analysis is a simulated frequency sweep of a circuit that is then presented as the circuit's small signal transfer function.

The problem with simulation of a VHF turret tuner is in estimating the stray capacitance, inductance and the coupling coefficients of the switched inductors. It is also hard to tell from the schematic provided which inductors are being selected in the turret for each channel in the VHF band. I suspect that different L2, L3, L4 and L5 elements are selected for each channel. Those inductors will have different values for each channel. That is why specific values for those components are not specified on the schematic diagram.

If you want to simulate general "band-pass coupling" you should be doing it over only one channel, not over the entire VHF band. You should also restrict your model to only the 10 components involved from the collector of Q1 = an AC 1 current source, to the base of Q2 = an output load resistor.

Different inductance and coupling values should be computed for each channel studied. Because the band-pass coupling is double tuned, the elements will be staggered so as to give a wider flat response for the channel selected.

 

[brainbaby]

If you want to simulate general "band-pass coupling" you should be doing it over only one channel, not over the entire VHF band. You should also restrict your model to only the 10 components involved from the collector of Q1 = an AC 1 current source, to the base of Q2 = an output load resistor.

I' ve designed the circuit but I am not sure about the values of L1,L2 and L3 for a start...

Also how close does .ac oct 1000 50MEG 72MEG corresponds to a single channel (as in post #1), I mean 50 to 70 MEG is ok for a frequency range of a VHF channel, but what should I do for rise and falls at 1.25,4,.5 Mhz...

 

[brainbaby]

You really must be more forthcoming about what you actually want to do here. We are still pretty much totally in the dark.

Not making things complex...the value of L2 above will govern the extent of band pass in the circuit...means the more the value of L2 the more will be the bandwidth i.e the more will be the band pass...
This is what I want to prove experimently using LT spice..

 

[berkeman]

Not making things complex...the value of L2 above will govern the extent of band pass in the circuit...means the more the value of L2 the more will be the bandwidth i.e the more will be the band pass...
This is what I want to prove experimently using LT spice..

Then just do an AC Analysis in LTSpice and look at the plots. Step the value of L2 to see the comparison of the passbands.

 

[Baluncore]

I mean 50 to 70 MEG is ok for a frequency range of a VHF channel, but what should I do for rise and falls at 1.25,4,.5 Mhz...

The dips and hollows represent less used parts of the spectrum that contain low energy, as such they are not important and so are irrelevant to the simulation. It is important to realize that the timing of NTSC or PAL analogue TV signals is very important, the spectrum usage is less critical.

For TV a flat frequency response is needed. With a double tuned BPF that can only be as good as the flat top of a stagger tuned BPF. I think the design was discussed in the previous thread.

The attached LTspice files show how as L2 becomes larger, the value of L2 becomes less critical and the transfer function becomes wider and flatter. Note that I have adjusted the AC current source amplitude to match the input impedance and so bring things near to 0 dB.

 

[sophiecentaur]

Also how close does .ac oct 1000 50MEG 72MEG corresponds to a single channel (as in post #1), I mean 50 to 70 MEG is ok for a frequency range of a VHF channel, but what should I do for rise and falls at 1.25,4,.5 Mhz...

Why would you even consider using a TV demodulator for receiving FM sound signals? You still haven't explained that for me. Just what is the application?
A broadcast VHF FM channel width is about 200kHz, which is pretty similar to the sound carrier on TV. What do you actually want to do, apart from using SPICE for some experience?

 

[Baluncore]

Why would you even consider using a TV demodulator for receiving FM sound signals? You still haven't explained that for me. Just what is the application?

The aim is I believe, not to build, but to study the double tuned band pass coupling network.
The circuit provided is for a VHF TV turret tuner that includes FM sound.
@sophiecentaur; where does @brainbaby specify only FM sound?

 

[sophiecentaur]

The aim is I believe, not to build, but to study the double tuned band pass coupling network.
The circuit provided is for a VHF TV turret tuner that includes FM sound.
@sophiecentaur; where does @brainbaby specify only FM sound?

He quotes (post 6) a previous thread of his involving a VHF tuner filter and he has not produced a frequency plot of that circuit's response. I'm a bit flummoxed by the mention of a "turret tuner". I never came across PAL on a VHF channel. That must surely be an historical spec (That quoted book is fairly recent, though). A proper superhet design would be suitable for all frequencies (VHF UHF ) and not a clunk clunk control with only a few channels. CATV may be a different environment. In the UK, it's all digital, afaik.
If that amplifier has to deal with the composite signal, it would be much more important to get amplitude and group delay at the colour subcarrier. The sound carrier can take its chances. I don't think the spec is very tight.
I am still confused by what's actually going on here. Is that filter definitely a real vsb IF filter? Even the necessarily crude channel filter, used at UHF, has four or five sections to get it right. The vsb receive filter is a difficult design and must have more than just a couple of squeeze coils. it has to have a symmetrical slope across the region of the vestigial sideband. You could never do that with just a couple of resonances, I am sure.
But the whole scenario here has not been specified enough for a proper answer and the main premise of producing a 'test' signal like the template at there top is really nonsense.
I think some of us are being over generous towards the OP and not helping him at all by inventing possible questions that he may be asking. He and we have wasted a lot of time on this which could have probably been spent in helping the OP come up with a situation that can actually be analysed.

 

[brainbaby]

As from the simulation it can be clearly seen that with the increase in the value of L2 the response becomes broader (i.e bandwidth increases) and now its a clear evident fact.
A special thanks to Baluncore for correctly judging my intent.

The aim is I believe, not to build, but to study the double tuned band pass coupling network.

Absolutely true.

The circuit provided is for a VHF TV turret tuner that includes FM sound.

@sophiecentaur Yup, as depicted by the figure in my post 1, the sound carrier is mixed with the AM modulated picture carrier in a television station.

 

[brainbaby]

@brainbaby: you have really confused me now. The thread you reference is about an FM sound circuit. Why would you need to include the whole TV signal. If you want to get the FM sound carrier from the TV (i.f.) signal then the most popular way is to use 'inter carrier demodulation' which puts the whole TV signal through a non linearity and filters out the 5.5MHz component (difference between sound and vision carriers).

My reference thread was not about an FM sound circuit. It was about a rf tuner circuit which clearly composed of a rf amplifier and a mixer circuit. There and here also I have discussed about a part of the Rf amplifier which is the coupling network.

It seems that you are more focussed in discussing the later part of the whole TV reciever circuit, which is the demodulater (post video detector stage) where both video and sound signal gets separated and each one of them goes to respective subsections of the circuit for processing.

If that amplifier has to deal with the composite signal, it would be much more important to get amplitude and group delay at the colour subcarrier.

Its a monochrome circuit.

But the whole scenario here has not been specified enough for a proper answer

Now here the resistance R is common(shared) between two circuit loops and by manipulating its value we can affect the magnitude of current flowing in loop 2. The same goes with inductances as well (like L1,L2,L3)..as manipulating value of L2 will effect transfer in the rest next of the circuit. This is known as shared impedance coupling and that exactly was what I tried proving using a simulation in LT spice.

 

[sophiecentaur]

I am seeing some light at the end of the tunnel here but it's not all clear yet. You originally asked for a 'test signal' but, from what I can read here and in the other thread, the amplifier you want is just a gentle channel filter. (There is also mention of a wider band response to cover more than one channel) I am also confused by the first two diagrams which seem to portray channels with different sound carrier frequencies. But the mention of turret tuners really amazed me because CATV means Digital in the UK

It seems that you are more focussed in discussing the later part of the whole TV reciever circuit, which is the demodulater (post video detector stage)

No really. The VSB filtering is before the AM demod, in the IF filtering.

Its a monochrome circuit.

If it's a monochrome circuit then why are you choosing a spectrum template that refers to a colour channel? I only know about monochrome tv in the context of the old 405 (uk) line system. Is there a monochrome TV service anywhere in the world, these days? I used the terms NTSC and PAL right at the top of the thread but no one picked me up. Which TV system letter are we discussing? I cannot find any mono systems that are current. This is even more confusing, now. Where is your interest in that niche in TV?

Now here the resistance R is common(shared) between two circuit loops and by manipulating its value we can affect the magnitude of current flowing in loop 2. The same goes with inductances as well (like L1,L2,L3)..as manipulating value of L2 will effect transfer in the rest next of the circuit. This is known as shared impedance coupling and that exactly was what I tried proving using a simulation in LT spice.

Now all this makes total sense but has nothing to do with the test signal you were wanting in the OP.
Why not define the channel bandwidth you need and the details of the band pass needs and check any designs with a range of spot frequencies? A graph would certainly have helped and you could give some actual background and context. Why did you choose this particular scenario for what seems to be a (very worthwhile) exercise in tuned RF amplifier design? I see you have had some interesting interchanges about the basic circuit ideas; there is a fair bit of experience here on PF.

 

[brainbaby]

A little further doubt friends...

I am trying to frame out certain values of frequencies along with amplitude for different values of Lm…but the problem is that for this I need a relation aka formula which establish a relation between these three variables i.e Lm , fx and amplitude.

So can I have it??

Second problem is that the value of these three variables are different at an instance of time. So to calculate relation between any two variable say (Lm and fx) the third deciding factor (here amplitude) should be kept constant but here it can't be possible because at a certain time the value of all three varies. (like at time instance T1, value of L2 is Lm' and frequency is fx1 and so on for other time instances as well)..

 

[Baluncore]

I am trying to frame out certain values of frequencies along with amplitude for different values of Lm…but the problem is that for this I need a relation aka formula which establish a relation between these three variables i.e Lm , fx and amplitude.

So can I have it??

The references you need are here.
http://www.qsl.net/va3iul/Files/Old_Radio_Frequency_Books.htm

Download a copy of the Radiotron Designers Handbook. 4th Edn.(1953)
http://www.tubebooks.org/Books/RDH4.pdf

Then look at pages 412 to 427, and 1025 to 1049.

 

[brainbaby]

Ya thanks I'll check it out