View Single Post
Old 27th Apr 2017, 11:26 pm   #1
Studio263
Octode
 
Studio263's Avatar
 
Join Date: Apr 2007
Location: Hampshire, UK.
Posts: 1,574
Default How the Sony non-PAL decoder works

The Sony non-PAL decoder was used in the UK models KV-1320UB, KV-1320UB MkII, KV-1300UB, KV-1310UB, KV-1330UB and KV-1800UB. It decodes the standard UK PAL I transmissions but does not use any techniques covered by the original Telefunken PAL patents. When these sets were made, Telefunken were reluctant to award PAL licences to Japanese television manufacturers because of fears of high quality, highly reliable Japanese sets flooding the European market. By the time Sony’s original KV-1320UB was ready Hitachi had acquired a PAL licence, but the Sony sets made Telefunken realise that it was futile to attempt to resist the unstoppable force of Japanese industry. After all, Sony could have easily made its technique available on licence terms, cutting Telefunken out of the loop completely.

The Sony decoder isn’t a modified NTSC circuit, nor does it convey any less colour information than a conventional PAL D decoder does. Its only disadvantage is that it does not automatically cancel phase errors so a user ‘hue’ control is necessary. It is most akin to a highly developed PAL S (simple PAL) circuit, which I shall explain later. Sadly, the exact workings of this arrangement were never made particularly clear and in service the decoders were very reliable, meaning that engineers never really had to get that involved with their basic principles. In what follows I hope to correct this omission and reveal one of the great masterpieces of television engineering in the early colour era.

PAL Patents

The PAL system has two key features which differentiate it from the NTSC technique on which it is based and which are covered by Telefunken’s patents. In simple terms these are:

1) The phase of the V signal (e.g. that from which the R-Y information is recovered) changes by 180 Deg. on each subsequent line. The U signal (from which the B-Y information is recovered – think of blUe to help remember which is which) – remains in constant phase. In a PAL D receiver, these phase changes are used with a 1 line electronic delay to cancel any phase errors that occur in the broadcast chain. Extreme errors cause nothing more disturbing than a slight loss of saturation. In a PAL S receiver, lines of chroma with opposite phase errors are placed next to each other on the screen so that when viewed at a distance the colours blend together in the viewer’s eye to yield an average (and correct) hue. This is why PAL S is only considered suitable for small screen sets. Therefore, both PAL D and PAL S receivers both average the chroma signal over two scanning lines, the former electronically and the latter optically. This is of no consequence as the chroma signal is of limited bandwidth and is not used to carry information about picture detail.

2) The colour burst, which is a 10 cycle snapshot of the suppressed 4.43MHz subcarrier, swings in phase by +/- 45 Deg. on each subsequent line. In the receiver, the automatic phase control (APC) loop of the subcarrier regenerator has far too long a time constant to respond to these changes and the subcarrier which it creates retains the mean phase of the two values, e.g. 180 Deg. in respect to the U signal. However, the ripple which is created in the DC control loop of the APC system is used to identify which lines have which phase of the V signal, thus allowing the receiver to rapidly automatically synchronise with the broadcast signal. Some sets also use the presence of this ripple to identify the presence of a colour broadcast and to automatically turn the colour circuits on.

Clearly, the Sony non-PAL circuit cannot use either of these features of the PAL signal in order to reproduce the picture.

Basic Principles

The Sony non-PAL decoder has at its heart two electronic switches that change state at line rate. They are synchronised to each other and to the line flyback period by a pulse from the line output transformer. Like a conventional decoder the Sony circuit has a chroma channel and a reference channel. It uses a pair of synchronous demodulators which recover B-Y and R-Y and a matrix to create an RGB signal for the Trinitron tube (the Trinitron requires RGB drive since in its smaller versions only the cathodes are available as separate connections, the grids are part of the common single-gun structure). These circuits are too all familiar and conventional.

Reference Channel

Like a conventional decoder, the Sony non-PAL circuit requires two 4.43MHz signals, 90 Deg. apart and with a fixed phase relationship to the average timing of the swinging burst, to drive the two synchronous demodulators. In a conventional PAL set (D or S) these would come from a single subcarrier regenerator which is phase locked to the burst. A 90 Deg. phase shift is inserted in the feed to the V channel, along with a means of further switching the phase by 0 or 180 Deg. in step with the phase alternation of the V signal. In the Sony circuit, there are two reference channels. Both are largely identical and use a 4.43MHz crystal in each (there are two crystals in these sets) as a filter (not as the basis for an oscillator) to separate out the fundamental frequency of the previously gated bursts. The output of the crystal excites a resonant circuit which forms the oscillator. This generates a continuous 4.43MHz signal to drive the relevant synchronous demodulator. The ‘Q’ of the crystal filter and the resonant circuit are high, so the effect of the swinging burst phase is rejected, leaving the oscillator phase at the average of the two values (U signal – 180 Deg.). This is all there is to the U channel, which also provides a control voltage for the automatic colour control (ACC) system and the colour killer. Meanwhile, in the V channel a clever technique is use to generate the required 90 Deg. phase shift. On alternate lines, one of the two line-locked electronic switches (see above) is used to alternately invert the chroma feed to the burst gate of the V channel only. Since the burst is sinusoidal, inversion is the same as a 180 Deg. phase shift, so in addition to the 45 Deg. swinging burst, the burst phase in the V channel now swings by a further 180 Deg. Just as the limited bandwidth of the circuit averages out the effect of the swinging burst to 0 Deg, it also averages the 0 / 180 Deg. phase changes to 90 Deg. Thus, a constant 90 Deg. phase shift is created in respect to the U channel, as is required to correctly deconstruct the quadrature modulation of the PAL signal. Note that there are no APC loops in this design and the information contained in the swinging burst is neither decoded nor used. This is the first part of how the need for a PAL licence is avoided.

Chroma channel

The chroma channel of the Sony non-PAL decoder is surprisingly simple. At its input however is a 64uS (one line) delay line whose output is exchanged with the direct signal line by line by the second electronic switch. Therefore, the chroma channel receives only the “even” line numbers, first directly and then as a delayed repetition whilst the “odd” number line is transmitted. No chroma information from the odd lines ever reaches the synchronous demodulators or the screen, it is simply ignored. Even lines contain only the phase inverted version of the V signal and so line by line the relative phase of the -V signal which is being processed is constant. Since the electronic switch that makes this selection is synchronised to the one that creates the 90 Deg. phase shift in the V reference channel the two are permanently locked together and the reference signal is always of the correct phase to correctly decode the selected -V signal. This is why the 90 Deg. phase shift in the reference channel is created the way it is, a fixed 90 Deg. shift circuit could not guarantee that the phase shift was always correct for the selected lines.

There is no particular reason why the –V signal is used, the circuit would function in an identical fashion if the non-inverted V signal which occurs on “odd” line numbers were chosen so long as the reference channel worked the opposite way round as well (which as we have seen, it automatically would). However, it is important that once chosen the same lines (even or odd) must always be used, more on this later.

By using pairs of the same line, the Sony non-PAL decoder removes the phase alternation of the PAL system. This clears another patent hurdle, but it also removes the system’s ability to automatically cancel phase errors. Because of this, a manual phase shift has to be included in the chroma channel so that the viewer can correct the reproduced colours; this is what the ‘hue’ control does. It works by altering the phase response of the chroma amplifier which follows the electronic switch. The output of a tuned transformer is applied across an R/C circuit of which the hue control is one of the R elements. When the decoder is set up, mean phase is set by receiving a standard signal, setting the hue control to its mid point (where there is a little mechanical “click”) and tuning the transformer for correct colour reproduction. With the control set like this it would not matter if the electronic switch selected the odd or the even lines to repeat, however as soon as the hue control is used to correct for transmission errors the line selection becomes important. This is because the action of the control would be reversed by shifting from one set of lines to the other, leading to potential appearance of colour errors (and the need to readjust the hue control) when the set was first switched on or if the signal was lost for whatever reason. For this purpose a novel ident system is used, more on that later.

After the hue circuit, the chroma signal, which contains both the U and the V elements in quadrature, is applied in identical form to both synchronous demodulators. This is the same as standard PAL S practice, making the Sony non-PAL decoder a variant on the PAL S design. PAL D sets use an electronic delay line separate the U and V signals by a process of addition and subtraction of the current and previous lines, hence the two synchronous demodulators receive different signals. Note that repeating the same line of chroma information twice carries an equivalent amount of information as each line of chroma being the average of two subsequent lines. Therefore there is no loss of picture detail with the Sony arrangement when compared to PAL D, since both retain the full luminance bandwidth. The Sony system does not use optical averaging on the screen like basic PAL S does, so it can be used for full sized sets as well. The largest set to use the non-PAL decoder in the UK was the 18” KV-1800UB.

Ident system

As previously described, the Sony non-PAL decoder requires an ident system in order that the hue control works consistently. This is a different requirement to the ident system of a conventional PAL D or PAL S decoder, which requires it to ensure that the PAL switch works in the correct phase and so operates the 0 / 180 Deg inversion in the V channel correctly. Using the swinging burst (+/- 45 Deg.) would be technically feasible, but this would violate the PAL patents so this information cannot be used. Instead, the set relies on a refinement of the transmitted signal which ensures that the burst phase is always the same just before and just after the blanking period. If this is not done an uneven number of each of the two burst phases would be received by the subcarrier regenerator (conventional PAL D and PAL S decoders), causing lock-in problems which manifest themselves as incorrect colours at the top of the picture. Instead of changing the phases, the bursts are blanked for a number of fields in a sequence of four possibilities which occur in a fixed rotating sequence. This way the burst always resumes in an orderly manner and PAL decoders lock quickly and accurately at the start of each field.

The Sony non-PAL decoder makes use of this sequence in order to identify which line is which as the fields start. To do this, a vertical sync pulse is required which occurs at the very start of the field. When this is received, the action of the electronic switches is blocked by a long pulse from a monostable circuit. At the same time, amplitude normalised bursts are received by another part of the ident system. Action of the electronic switches can only resume when the long monostable pulse (which is adjustable in length) has cleared and the bursts have resumed. Since this condition only occurs in the broadcast signal at a definite burst phase it follows that chroma on that line must also have a consistent phase. The circuit is set up so that it is the even (-V) lines that are used. Sony called this the “burst detector clock pulse check”, and in simple terms it makes the electronic switch ignore the line transformer pulses until the signal conditions are correct. The information contained in the swinging nature of the burst phase is not used in any way by this arrangement.

Conclusion

This is only a brief description of how the Sony non-PAL decoder works. In fact there are many further details in the practical implementations of the idea, which partly explains why after 40 or more years there are still so many of them working despite receiving little or no attention in that time. For further reading, I can recommend the original Sony manuals which give the complete circuits and setting up instructions, along with Volume 1 of the Newnes Colour TV Servicing Manual (the red one) by Gordon King, which has a brief supplement on the Sony non-PAL decoder in the back. There is also an excellent article in the October 1971 issue of ‘Television’ called “Secrets of the Sony Colour Receiver” by K. Royal which I cannot recommend highly enough.
Studio263 is offline   Reply With Quote