Why Is the Picture Carrier Placed 6 dB Down in the IF Response Curve?

[brainbaby]

Hi folks,

My query is that...
As one can see that in the IF response curve the picture carrier is placed 6db down the amplitude curve. Such a placement modifies the output of video detector in such a way that the picture carrier lies at half point or 50 % of the total amplitude.
Now this whole story is about VSB correction. I wonder what exactly goes inside the video detector that making a 6db adjustment of picture carrier in IF response curve shapes the output of video detector this way.

What so special about this 6db..?

Please be specific..

 

[Tom.G]

See: https://subjects.ee.unsw.edu.au/tele3113/lecture_notes/TELE3013_week4.pdf

above found with: https://www.google.com/search?&q=tv+vestigial+sideband+shaping

 

[brainbaby]

The output of the sideband filter gives 50 percent of the gain at picture carrier frequency.
First i want to ask where is the sideband filter is located. either is it in the IF amplifier(IF section) or the video detector..
Let me answer this ...it is located in the video detector as per my present understanding. why i say so is that in my diagram you can see that the picture carrier in the IF output is at 38.9 Mhz where I have marked it 6dB,at this point there the amplitude is not half..Its the video detector that is yielding half the gain at picture carrier.

however your figure makes me feel that the sideband filter is located in the IF section or can say is an IF filter. Please could you clear it out.

 

[Tom.G]

First i want to ask where is the sideband filter is located.

It is at the Transmitter. Since there is identical information in the two sidebands, there is no need to use spectrum space or transmit power to send both of them.

 

[brainbaby]

You mean to say that the entire vestigial side band correction is done in the transmitter not in the receiver nor in the IF section..

 

[Tom.G]

You mean to say that the entire vestigial side band correction is done in the transmitter not in the receiver nor in the IF section..

Not at all. I'm saying that one sideband is not completely suppressed at the transmitter (making it vestigial) to supply a make-up signal to the receiver. Then the receiver, not having a perfectly steep-sided passband, passes the main sideband and the remaining bit of the suppressed sideband to the detector. The detector automatically combines the two sidebands, filling in the main signal that is on the slope of the I.F. passband.

The alternative would be to tune the receiver I.F. to pass all the video frequencies to the detector. But since the passband doesn't have absolutely vertical sides, other interfering signals could get thru, which means the TV channel spacing would have to be wider to avoid interference.

 

[tech99]

Hi folks,

My query is that...
As one can see that in the IF response curve the picture carrier is placed 6db down the amplitude curve. Such a placement modifies the output of video detector in such a way that the picture carrier lies at half point or 50 % of the total amplitude.
Now this whole story is about VSB correction. I wonder what exactly goes inside the video detector that making a 6db adjustment of picture carrier in IF response curve shapes the output of video detector this way.

What so special about this 6db..?

Please be specific..

View attachment 224096

In the VSB system, low modulating frequencies, up to say 1 MHz, are transmitted as Double Sideband, and high modulating frequencies, say up to 5 MHz, are transmitted as Single Sideband. In general, SSB reception using an off-air carrier is subject to a slight distortion called Quadrature Distortion. This arises because the carrier is subject to inadvertent phase modulation as it is not large in comparison to the side frequencies. By using DSB for low modulating frequencies, the sync pulse shapes are preserved and the DC component arising from the carrier, and which controls black-level, is correct.
If you now imagine the transmit filter to achieve VSB, it might be be flat from Fc +5 MHz to Fc -1 MHz. When the DSB portion is demodulated, each pair of side frequencies add in phase, giving double amplitude compared with the SSB portion. So it is necessary to attenuate the DSB portion by 6dB. This can be done at the TX or at the RX, and the latter is the method chosen. It is done by rolling off the filter between Fc +1MHz and Fc -1 MHz such that the carrier is 6dB down. The sloping response causes each pair of DSB side frequencies to add up to an amplitude equal to that of the SSB ones.
The video detector is just a diode envelope detector. I believe It is sometimes the practice to use a flat RF/IF filter and then reduce the amplitude of the low frequencies by 6dB after demodulation.

 

[sophiecentaur]

What so special about this 6db..?

Some of the following points have already been made by @tech99
The 6dB is Vital!
As far as I know, the transmitted VSB TV signal is straight DSB for lower frequency (+-1MHz) sidebands and SSB for the higher frequencies (to 5MHz) with a flat channel filter from -1MHz to 5MHz. The spec for the transmitter response is flat, over the VSB width as far as I remember and the adjacent channel protection levels are based on that. Such a signal is far easier to amplify in a range of transmitting devices and channel combiners etc. If the signal is passed directly to an envelope detector, there will be twice the amplitude of lower demodulated frequencies. This would need some clever video equalisation at the transition around 1MHz. So, before the demodulator, a filter with a slope which is antisymmetrical about the carrier frequency and extends over the DSB portion. The IF filter characteristic is half amplitude at the carrier frequency so that pairs of sidebands (upper and lower) will add (phasors) to unity and the demodulator will produce a flat video response for all video frequencies. The inevitable group delay distortion can be corrected fairly easily.
I have not come across a system with a simple diode detector followed by a fancy video filter. Wouldn't that require a difficult filter characteristic right on the transition? The gentle VSB correcting filter works over the whole 2MHz and that makes group delay distortion a lot easier. TV images are as much (or more) to do with phase than amplitude response.

 

[sophiecentaur]

First i want to ask where is the sideband filter is located.

I just looked at this again. The answer is that there are TWO sideband filters. A flat VSB filter in the transmitter and a receiver IF filter with 'that slope' around the vision carrier frequency.

 

[brainbaby]

Sorry to be a little late...

My conclusion...
The output from the rf tuner consist of a small part of lsb and a prominent usb.Now if the signal at this point is not treated and passed on to the detectot stage then the output of the detector would be constant till the lowest usb frequency and double for the lsb frequencies which would lead to some kind of distortion. Now to prevent this a special treatment to the signal is done at the IF section.This treatment is vsb correction.The vsb correction is achieved by regulating the passband of the if section. The IF section consist of an if filter which pass on the usb frequencies to full extent and begins a rolloff for LSB freq. such that the corresponding amplitude to the lsb frequencies adds up and is equal to the corresponding amplitude for the usb frequencies.Here it has been observed that the picture carrier frequency yields a corresponding amplitude to be half (50%) of the net corresponding amplitude to the usb frequencies. So the entire scheme is to regulate the amplitude of the lsb frequencies such that it would be constant for the usb and that is achieved by a simple roll off of 6 db of the passband of if filter. If the roll off is not kept to 6 db or let's say other value for e.g 8 or 12 db then corresponding amplitude of lsb freq. may not be equal to that of the usb freq. After the vsb correction at the if stage the output of the demodulator or detector is equal to my last figure.

The mistake which i made was that i thought the if filter response to be a some kind of signal which made me think that how would a signal with a sharp 6db rolloff gets demodulated in the video detector to yield a response with 50% of the amplitude at the carrier frequency.

Have I got it right??

Some further queries..
1. What is that "some kind of distortion"..?

2.How does negative LSB frequencies contribute to important signal information as negative frequencies has no real existence..?

 

[sophiecentaur]

The basic “ some kind of’ distortion would be that the video response would have a flat, full amplitude up to say 1MHz and a lower (half) amplitude response at higher video frequencies. Not too bad in itself because it would just make the picture a bit ‘soft’ to look at. But the demodulated high frequency video would have phase error because the single side band component is in quadrature with the carrier. Edges would have fringes along them.
The gentle (not “sharp”) IF slope across the vsb portion gives a slow transition from no quadrature distortion around the carrier frequency to full quadrature distortion from 1MHz upwards. This is easy to correct at the video stage.
Negative frequencies are a perfectly reasonable concept in frequency space even though they may upset intuition. They are negative relative to the carrier frequency.

 

[brainbaby]

But the demodulated high frequency video would have phase error because the single side band component is in quadrature with the carrier. Edges would have fringes along them.

When we try to suppress lsb in vsb transmission what causes the envelop of demodulated lsb to suffer phase distortion...? I am visualising quadrature distortion like that some of the lsb frequency is in phase with the picture carrier and that portion of the lab is called inphase component and that component aka that lsb frequency which is in quadrature with the picture carrier is known as quadrature component. So this quadrature component when passed to the envelop detector while demodulation gives improper detection as shown by the figure.

Now the thing which i am not getting is that at the start of lsb suppression in vsb transmission causes the generation of this quadrature component which further contaminate proper demodulation. Why this quadrature component gets generated..??

text source:
Effect of the Quadrature Component in Single Sideband Transmission by Nyquist, H., Pfleger, K. W.

 

[tech99]

Slight distortion occurs when we demodulate SSB with an off-air carrier, because the carrier is not large compared with a side frequency. An analogy of how this occurs can be found with a reciprocating engine. When the connecting rod is long, the piston follows Simple Harmonic Motion i.e. its position is described by a sine wave. But if the connecting rod is short, the motion of the piston is not perfectly sinusoidal; it lingers near the bottom of the stroke and speeds up near the top.
In this analogy, the crank is analogous to the phasor of a side frequency and the motion of the piston to the resulting envelope as applied to the detector.

 

[Tom.G]

When the connecting rod is long, the piston follows Simple Harmonic Motion i.e. its position is described by a sine wave. But if the connecting rod is short, the motion of the piston is not perfectly sinusoidal; it lingers near the bottom of the stroke and speeds up near the top.

Interesting. Any chance you can supply the supporting math for that? I'm always ready to learn something new! (maybe in a new thread so we don't hijack this one.)

 

[sophiecentaur]

The phasors and carrier of an AM signal (sinusoidal mod to make it easy) are shown in this diagram. The resultant is in phase with the carrier. If you remove the lower sideband and envelope detect the result, the detected amplitude will be the length of a line from the origin to the blue arrow head. This is not the value of the envelope detected full DSB signal and the resulting envelope is out of phase with the detected DSB signal.
The 6dB cutoff VSB filter will produce the full phase and amplitude error for high video frequencies (Only the blue arrow is there) but for video frequencies around zero the green and blue are almost equal (but half the length of the blue phasor for the SSB signal) and without any phase error. A phase slope from zero to more than 45 degrees (depending on mod depth) will give a time error for edges. Very bad for TV images but not much of a problem for sound.

Slight distortion occurs when we demodulate SSB with an off-air carrier,

The TV signal is aimed at the simplest possible detectors and the mod depth is actually low for the picture part of the TV waveform. Things would be better with a phase locked carrier to demodulate the signal but in the 1950s when the (new) system was developed, the sets could not include this. I believe some circuits used an 'Exalted Carrier Demodulation' which filtered the carrier and amplified it. I never saw inside one of those!

 

[Baluncore]

@ brainbaby.
See; Image-Reject and Single-Sideband Mixers.
https://www.rfcafe.com/References/articles/wj-tech-notes/ImageRej_n_SSB_mixers.pdf

 

[sophiecentaur]

@ brainbaby.
See; Image-Reject and Single-Sideband Mixers.
https://www.rfcafe.com/References/articles/wj-tech-notes/ImageRej_n_SSB_mixers.pdf

As with all properly designed comms systems, it involves a phase locked local oscillator, which the old tellys didn't have.