Role of the capacitor, coil and choke in this television receiver design

[brainbaby]

TL;DR Summary

This thread has been based on Dr. Philo's Farnsworth first television circuit.

Hi friends,

I am studying this circuit which is the first all electronic television receiver created by Dr. Farnsworth back in 1927. Triode are being employed along with some caps and coils. The design is very simple and clear, basically this receiver is a detector circuit which detects the first carrier wave from the transmitter circuit (not discussed here as for now). The signal that has been radiated by the transmitter contains the light signal generated by the photoelectric cell of the transmitter which modulates the first carrier wave, a second carrier wave and a two analysing potentials. The analysing potentials are basically oscillating currents of different frequencies (10 cycles and 500 cycles/s). The analysing potentials are used for deflection of the electron beam in the oscillograph (or CRT). The circuit consist of three triodes as detectors and an oscillograph.

I have three questions.
1. If 168 is used for passing high frequency currents and blocking low frequency then why 166 is used. The text says that 166 is used for passing high frequency currents, but as 168 is also by-passing high frequency current of the first carrier wave then why 166 is used for the same purpose twice.
2. What is the exact role of triode in detection (Is it provide coupling between stages).
3. Role of choke 169 as it passes the analysing potentials. Analysing potential current contains two frequencies current one is at 10 cycles/s and other is at 500 cycles/s. As we know that the inductor or a coil passes only low frequencies currents thus blocking high frequencies, this coil 169 should block a high freq. current of 500 cycles, but as per the text its passes both high and low-frequency currents?

I have attached the circuit diagram along with the patent pdf (refer page 7 for receiver description) as well a screenshot.

Thanks!

 

[davenn]

Summary: This thread has been based on Dr. Philo's Farnsworth first television circuit.

1. If 168 is used for passing high frequency currents and blocking low frequency then why 166 is used. The text says that 166 is used for passing high frequency currents,

"166" is a DC block. It stops the battery DC that is being fed to the anode of the triode detector from getting to the Grid of the second triode "164"

Summary: This thread has been based on Dr. Philo's Farnsworth first television circuit.

2. What is the exact role of triode in detection (Is it provide coupling between stages).

It's a detector ... just as any diode detector in any receiver

As we know that the inductor or a coil passes only low frequencies currents thus blocking high frequencies, this coil 169 should block a high freq. current of 500 cycles, but as per the text its passes both high and low-frequency currents?

As it stands, this is an incorrect statement, so every stated following it is irrelevant

 

[Tom.G]

There are four signals, all at different frequencies, being processed in the receiver:
500,000Hz the RF (Radio Frequency carrier frequency)

and the signal frequencies:
150,000Hz for the video (Brightness)
500Hz for the Horizontal deflection (500 scan lines per second)
10Hz for Vertical deflection (500/10, or 50 lines for an image)
1. a)The value of Coil 162 is chosen to act as a load for triode 159 at the Brightness frequency.
b) Capacitor 166 would be a relatively small value to pass the Brightness signal to triode 164 and block the two (much lower) deflection frequencies.
c) Cap 168 value is chosen to pass remaining Brightness signal to circuit common (ground), with minimal effect on the deflection frequencies.

2. Triode 159 serves as an RF detector with gain whose output contains the three signal frequencies.

3. The value of coil 169 is chosen to block any remaining high frequency Brightness signal from getting to the deflection circuits, while still passing the lower frequency deflection signals.

Triode 164 is an amplifier to drive the "Light Rotor." Triode 179 is an amplifier to drive the Vertical deflection 134.

Cheers,
Tom

 

[tech99]

Farnsworth seems to use a sub carrier at line frequency, 500Hz, which is itself modulated by the 10Hz frame frequency. It looks as if the video is only above 500Hz.
Triode 159 is a leaky grid detector which will convert the incoming signal to baseband. The baseband is video plus a sub carrier at 500Hz.
164 is a video amplifier controlling a light gate. The grid leak and capacitor might be to provide DC restoration or more likely just a mistake.
Inductors 162 and 169 probably form a LPF to block video and pass 500Hz.
The 500Hz line frequency is developed across inductor 162 and applied to the oscilloscope.
The 500Hz sub carrier is also applied to triode 179, a leaky grid detector, which will detect the 500Hz and provide the 10Hz frame frequency output to the oscilloscope.
I am uncertain whether the blue arrows are correct. It is uncertain whether triode 179 is a detector of 500Hz or an amplifier of 10Hz. If the latter, inductor 169 is a low pass filter for 10Hz.
I can see a number of mistakes in the circuit, for instance no plate supply for triode 159 and the filtering arrangement being not quite right due to the terminating loads, but I think the circuit is intended to be just a concept.
It is interesting that at this time a systems approach to design was not used, and everything was at component level. As there was no test gear beyond a voltmeter (being more complicated than the actual system) this was the only way.
The concept is similar to that of Campbell-Swinton, 1908, which used separate synchronising signals together with cathode ray tubes at both ends.

 

[brainbaby]

There are four signals, all at different frequencies, being processed in the receiver:
500,000Hz the RF (Radio Frequency carrier frequency)

and the signal frequencies:
150,000Hz for the video (Brightness)
500Hz for the Horizontal deflection (500 scan lines per second)
10Hz for Vertical deflection (500/10, or 50 lines for an image)

Got it!

Discussing 1st and 2nd points for now.

It seems brightness frequency is a band of signal comprising of low frequency component as well as high frequency component.

b) Capacitor 166 would be a relatively small value to pass the Brightness signal to triode 164 and block the two (much lower) deflection frequencies.

If the value of capacitance of 166 would be higher then it might block the low frequency component of the brightness signal and that would flow across with 162 and then to 169 which would rather interfere with the deflection currents. So there is no need for low frequency component of brightness signal to pass through 162.

c) Cap 168 value is chosen to pass remaining Brightness signal to circuit common (ground), with minimal effect on the deflection frequencies.

But how can the remaining high frequency current will flow in this branch as it would have been blocked by 162 as an inductor imposes resistance to high frequency.

I am thinking it this way, both 162 and 166 is acting as a load to 159. It is a kind a parallel LC circuit which is resonant at the first carrier wave.

For a parallel LC circuit impedance is infinite at resonance, hence we can safely assume that since no current should flow between x and y the potential at x and y should be equal(as for current to flow there should be a potential difference).

There would be a certain potential at node y which is making a possibility of left over high frequency current to pass through 168.

But here I feel again confused.

Would resonance be established between 166 and 162 since it contains both high and low freq. components that might prevent resonance to be established at a static frequency.??

2. Triode 159 serves as an RF detector with gain whose output contains the three signal frequencies.

Shouldn't the output contain RF carrier as well?

 

[tech99]

A 500kHz RF carrier component is present but will be too high for the oscilloscope to respond to. It will also be reduced by the stray capacitances in the system.
I also tend to think now that Farnsworth intended the 500Hz line and 10Hz frame frequencies to be carried as two "pilot tones", and not by modulating the 500Hz with the 10Hz which is how I would have done it!
Regarding the low video frequencies, below 500Hz, it is true that these frequencies are present but it is possible to remove frequencies below line frequency if we use good black level clamping each line. As Farnsworth probably did not do this, it is possible there would be shading across the picture after an object. In practice, to avoid the shading effect, we need a video response down to about ten lines, ie 50Hz. This defect would perhaps be the least of his worries.

 

[Baluncore]

I can see a number of mistakes in the circuit, for instance no plate supply for triode 159 and the filtering arrangement being not quite right due to the terminating loads, but I think the circuit is intended to be just a concept.

It looks real to me.
VT159 is operated with negative = -HT on the cathode so the plate is near ground potential, where it drives the grid of VT164, a cathode follower, with plate connected to +HT.

 

[brainbaby]

Farnsworth seems to use a sub carrier at line frequency, 500Hz, which is itself modulated by the 10Hz frame frequency. It looks as if the video is only above 500Hz.

The transmitted output contains a package of frequencies which consist of the modulated first carrier wave in which brightness signal modulates the first carrier wave, this modulated first carrier wave in turn modulates the second carrier wave and the analysis potential.
So the transmitted output contains and image signal first and second carrier wave and two analysing oscillating waves, all modulated.158 and 160 is a tuned circuit which is resonant with the sub carrier wave.

Question: If 158 and 160 is resonant with the sub carrier then all the other frequencies i.e line freq. frame freq. second carrier is blocked before this tuned circuit, only subcarrier modulated with light signal is allowed to pass, but according to the text it is not the case, as the package tends to pass ahead, then what is the use of resonating the sub carrier with 158 and 160.
162 and 166 is also a tuned circuit which resonate with the sub carrier wave.The input to 164 is the light signal modulated upon the sub carrier. The output of 164 is the detected light signal.

LPF to block video and pass 500Hz.

Right, I agree.
Since the input to 164 is the subcarrier modulated on light signal and we are talking in terms of voltage (potential) then the LC resonant circuit (162 and 166) is have a high resistance to modulated sub carrier wave.where as for the non resonant frequencies i.e line freq. frame freq. are passed by 162.

he 500Hz sub carrier is also applied to triode 179, a leaky grid detector, which will detect the 500Hz and provide the 10Hz frame frequency output to the oscilloscope.

It hard for me to agree with the statement here because there is no chance for the sub carrier to be applied to 179 as the LC resonant circuit (162 and 166) is resonant with the sub carrier itself, rather sub carrier along with light signal is applied as an input to 164.
179 is acting as a detector for frame freq. as 169 passes both the frame and line freq. but then what is the use of 168?? since in text its said that it is used to bypass the high frequency of the first carrier wave, which is a bit vague to understand, rather it should be said that 168 is used to pass any other high-frequency components which is not resonant with 162 and 166. Also as per text its said that 168 is acting to block the line and frame freq. to prevent them to pass to any other part of circuit except 179 and oscillograph.

 

[Tom.G]

If the value of capacitance of 166 would be higher then it might block the low frequency component of the brightness signal and that would flow across with 162 and then to 169 which would rather interfere with the deflection currents. So there is no need for low frequency component of brightness signal to pass through 162.

Yes, if C166 were a higher value it could pass more low frequency signal. However the input impedance of VT164 is extremely high, on the order of 100k to a few MegOhms. This would not be a significant load on VT159 regardless of C166 value. So there would still be enough low frequency signal available for the rest of the circuit to use.

c) Cap 168 value is chosen to pass remaining Brightness signal to circuit common (ground), with minimal effect on the deflection frequencies.

But how can the remaining high frequency current will flow in this branch as it would have been blocked by 162 as an inductor imposes resistance to high frequency.

But it is not an infinite resistance. If you recall, the impedance of an inductor varies directly with frequency, so it can not perfectly block any frequency... some signal will always pass thru it. C168 helps to suppress (bypass) this pass-thru of L162.

Would resonance be established between 166 and 162 since it contains both high and low freq. components that might prevent resonance to be established at a static frequency.??

I'm not really sure what you are trying to ask here, but any series resonance of C166 and L162 would not be noticeable because of the extremely low Q. Note that the input impedance of VT164 is in series with C166 and L162, making that loop very high impedance at all frequencies.

Shouldn't the output contain RF carrier as well?

Yes. Likely at a high enough frequency that the rest of the circuit can not respond to it. For instance you won't get those mirrors in the light path to move half a million times per second!

Cheers,
Tom

 

[tech99]

I think Farnsworth drew the circuit using separate batteries in order to show the concept. There were no good DC amplifiers at that time and I think he was possibly trying to do use amplifiers with a response down to DC. I have also re-drawn the circuit and I don't think it uses a cathode follower, which was not invented until later. I have myself built a mechanical TV system using 32 line, 12.5 frames per sec. This has similar scan frequencies to Farnsworth. If you apply any sort of filter at line frequency it will completely destroy the picture due to ringing and is not workable. Further, if you allow a pilot carrier at line frequency to contaminate the video, it will also destroy the picture.

 

[sophiecentaur]

There are four signals, all at different frequencies, being processed in the receiver:
500,000Hz the RF (Radio Frequency carrier frequency)

and the signal frequencies:
150,000Hz for the video (Brightness)
500Hz for the Horizontal deflection (500 scan lines per second)
10Hz for Vertical deflection (500/10, or 50 lines for an image)

I'm wondering about that way of looking at the basics of TV.
An interesting bit of history and the date shows just how fast progress was made towards a working broadcast TV system.
That old patent is written in a (not surprisingly) strange way because more or less every basic aspect of TV had to be explained to the Patent Inspectors. It was hard going and I have to say, I scanned through the text fairly fast - trying to interpret what is said in modern terms.

The TV signal is usually (these days, at least) looked on as a time domain signal (the TV waveform, rather than spectrum makes much more sense in most instances) and the field and line sync signals are actually wide band pulses, occurring during the gaps at the ends of lines and not really separate CW signals. The sync pulses are essentially part of the video signal. That's the only way to ensure that the sync signals do not produce crosstalk onto the luminance signal. Field and line sync pulses were / are separated by level and time, rather than just with a simple filter. The pulses are fed either a ringing circuit or a phase locked oscillator. I guess these original circuits were doing their best with only a few fairly low performance components. It would be great to see the actual quality of the pictures they got.

 

[davenn]

b) Capacitor 166 would be a relatively small value to pass the Brightness signal to triode 164 and block the two (much lower) deflection frequencies.

I disagree, it is primarily a DC block. The cap and parallel inductor following 166 is the frequency filter

Plate voltage from the previous stage MUST be stopped from reaching the following grid of stage two

The tuned circuit of the C and L, in parallel, do not achieve that

 

[brainbaby]

But it is not an infinite resistance. If you recall, the impedance of an inductor varies directly with frequency, so it can not perfectly block any frequency... some signal will always pass thru it. C168 helps to suppress (bypass) this pass-thru of L162.

A layman trying to visualise how the picture signal modulates the carrier wave.

Case 1: Since the picture signal is not an individual frequency signal, rather its a group of frequencies, now each of these frequencies of the picture signal modulates the carrier wave.

This is not technically possible as the amplitude of the carrier wave cannot be congruent with different amplitudes of the picture signal at the same time.

Case 2: Only a particular frequency of the picture signal modulates the carrier wave.

If case 2 is right then apart from the picture signal frequency that modulates the carrier and resonates with 162 and 166, every other image frequency will pass down through 162 and 169 and 168 and to the oscillograph. If this is the fact then aren't we losing important picture details ?. Then this won't be an effective circuit design as a efficient receiver should always isolates the picture signal from the deflection circuits.

Also, if case 2 is right, then how can the other video frequencies traveled to receiver circuit from the transmitting end, as each of these picture frequencies might needs a strong carrier wave to get transmitted over a distance.

So we can't assume case 2 to be right either.

I guess that Farnsworth is not thinking the picture signal as a group of frequencies rather he assumes the signal to be a single light current having amplitude/voltage vs time relationship.

C168 helps to suppress (bypass) this pass-thru of L162

The value of 168 would have been such that it should block the frame and line freq. at all cost.

The sync pulses are essentially part of the video signal. That's the only way to ensure that the sync signals do not produce crosstalk onto the luminance signal

Yes absolutely, can 162 and 166 can also be thought of also as a sync separator here??

 

[Tom.G]

I disagree, it is primarily a DC block. The cap and parallel inductor following 166 is the frequency filter

I agree that C166 functions as a DC blocking cap. Additionally as a high-pass:
Quoting from the PDF of the patent:
Pg 7, LNs 49-51 "...166 indicates a condenser for passing the high frequency and blocking the low frequency currents..."

Pg 7, LNs 37-51 Indicate that L162 is (self?) resonant with the first carrier, and Pg 4, LN 118 et seq put 'first carrier' at 500kHz

Guess we will just have to wait until someone finds one of these TVs in a museum.
(edit: or takes time to read and digest all 13 pages of the patent!)

Here is one possibility if anyone cares to dig that deeply:
https://www.museum.tv/

 

[sophiecentaur]

Yes absolutely, can 162 and 166 can also be thought of also as a sync separator here??

Yes, I think so. Of course, I was forgetting the way old 405 line sets used to work. The vertical hold was very dodgy; it was more of an oscillator with adjustable frequency which would manage to lock itself in phase - almost by chance- once its frequency had been set by the user ('vertical hold' knob). I would think that the original receiver circuit would need constant attention to keep the picture stable.
That document really demonstrates the need for functional block diagrams (which are missing). Even with the advantage of hindsight, the system is pretty hard to understand. Imagine the problems that original readers would have had without being able to identify Sync Separator, Scanning waveform, Black level, Peak white etc. etc.. "It'll never work" would have been the reaction in many cases.

Then this won't be an effective circuit design as a efficient receiver should always isolates the picture signal from the deflection circuits.

Treating the signal in the frequency domain makes it very hard to think about blanking signals, which turn off the video drive during flyback; the time domain description instantly shows what is needed.

 

[sophiecentaur]

I disagree, it is primarily a DC block.

DC block or HP filter - it's the same thing. But the triode is self biasing which I think would perform a (required) rudimentary clamping function (?).
It is worthwhile remembering that the frequency components of the Luminance signal extend below the field rate frequency. To avoid the brightness of a part of the image varying with the content of the rest of the image then the response must extend below the field rate. Some non-linearity is needed to avoid the effects of the AC coupling (hence the clever clamping that's used these days).
Basically, it would be necessary to do some in depth study of the paper to find exactly just how sophisticated the circuit is. Some actual component values would be essential because of the spectra of the video and sync waveforms.

 

[Baluncore]

37 years later, in 1964, TV is still B&W with vacuum tubes. This is an interesting read.
https://ia800700.us.archive.org/5/items/DualStandard625LineTvReceivers/King-DualStandard625LineTvReceivers_text.pdf

 

[brainbaby]

Is it possible to simulate this circuit in ltspice.

 

[sophiecentaur]

Is it possible to simulate this circuit in ltspice.

Ha - be my guest. Meanwhile I will light my gas stove by rubbing two sticks together and get my shopping on my donkey cart.
I reckon these things are best viewed through a fuzzy screen (a bit like those old TV images). How well are those ancient components actually modeled on modern simulations? Waxed paper capacitors could have behaved very oddly at times.
Otoh, there may be some wheels from that time that have been re-invented severals times since.

 

[Baluncore]

Is it possible to simulate this circuit in ltspice.

Yes it is possible. But exactly what circuit do you simulate?
A patent is called “letters patent” because it is supposed to be described clearly.
But this patent is far from clear and is best abandoned.
What can possibly be gained by simulating the drafting errors?

 

[sophiecentaur]

Also a Patent is not guaranteed to work. It may have required a lot of extra effort to actually make it work if, indeed it ever did. In the absence of component values, a simulation would be doomed, I think.

 

[brainbaby]

Yes it is possible. But exactly what circuit do you simulate?

As per an advice from a prominent figure of PF states that just by memorising complex television circuit will bring no good to me, rather I should know the 'legacy requirements' before studying any circuit.

I now assume that rather inspecting complex design of an AM television it is better to start-off with very simple primitive designs first and then slowly moving on to newer designs along with time hierarchy and increasing complexity.

I assume that there will be certain benefits of this approach as-

1. It will provide an intuition about the design style of the designer (why, when and how to place circuit components in order to achieve desired specifications and what changes will it bring to the circuit response by adding/removing and modifying component values.

2. Enhance analytical circuit design skills, by knowing what to expect and how to deliver.

However, I am not eager to simulate this circuit as I was just thinking of to reap benefits of point 1.

I agree with sophiecentaur, that without knowing component values we cannot correctly simulate this circuit, also knowing the circuit values helps to know the exact function of the component which in this case cannot be framed.

Also we are unaware of the output response of the circuit for an image signal (also unknown), which kills our last hope to reverse engineer the circuit and then to declare the component use.

 

[tech99]

I suggest looking at the designs of early monochrome TV equipment for what was then called High Definition - 525 line in USA and 405 in the UK. Maybe 1950s. By then the system was well understood.

 

[sophiecentaur]

It is worth pointing out that analogue TV is definitely on the way out (if not already totally dead). It is a terrible use of spectrum space and was ditched as soon as the appropriate digital technology came available.
The development of the PAL Colour TV system was truly inspired and the system is unbelievably clever at getting the very most out of a 7MHz wide AM transmission channel. A fascinating bit of history (of which I was a small part).
Sadly (perhaps) all of that is 'water under the bridge' and the coding, transmission and storage of TV programme material is now part of the greater world of Digital Data. The basics of image coding that were used in the development of Analogue TV have provided a great base to work from. But I would question of the value of any study of TV technology that's based on the early theories (almost a century old). Interesting, of course, but on a very superficial level.

 

[tech99]

You might like to see a 32 line 12.5 fps image taken with my home made camera about 30 years ago. In the picture, you can see the rectangular camera, which uses a photo multiplier tube and a 9 inch scanning disc. The picture can be displayed on a neon tube/LED and disc or by converting it to 625 lines as I show here. The converter uses RAM in a hardware only circuit. I found that the limitation of these early systems was largely the display device. The actual picture was 312 line 50 fields non interlaced. The sampling was at about 3 times Nyquist, about 54 kbit/s.

 

[sophiecentaur]

But I would question of the value of any study of TV technology that's based on the early theories (almost a century old). Interesting, of course, but on a very superficial level.

You might like to see a 32 line 12.5 fps image taken with my home made camera about 30 years ago.

That was a brilliant project and I am really impressed. It sort of goes against my comment, above, but I would suspect that you did not initially learn about TV by building the project. Rather, you used your existing knowledge of TV systems to help you in the 'historic re-enactment'.

 

[tech99]

I was experienced in TV engineering, but interested to know what the pictures had looked like in earlier times. I did learn about new areas such as optics, low frequency requirements for TV, phase locking a mechanical scanner and digital electronics.
I think the clock was 54kHz and the samples were 8 bit. The clock was the source of every frequency in the system by using dividers ie 54kHz sampling freq, 312 line freq, 50Hz field freq, 400Hz line freq, 12.5 Hz frame freq. The clock was locked to the 400Hz from the camera.