|
SONY DXC-1640P Internals.
|
|
The Sony 1640's advantage (and to some extent also, it's disadvantage) was clever single-tube
colour 'technology'. Above we also see early efforts at accessible camera electronics, as by undoing two screws, both camera sides hinge open. The image tube is inside the copper coloured cylinder, lower left. But actually there is very little that needs adjusting inside this camera.
The front view below shows how both sides of the camera hinge open. The Canon lens has manual
zoom and is a standard cine camera C mount, and I am not sure when Sony (and everyone else) started using unique and incompatible quick-release lens mounts for their bigger lenses.
Tube cameras are very easily damaged by bright objects (such as the Sun), fortunately the lens
automatically closes when the camera is turned off. |
|
Sony's TRINICON was an ingenious modification to the humble black and white VIDECON
tube. Videcons were of course the standard imaging device of all (?) non broadcast television cameras before the CCD came along.
Sony had already introduced their own improvement of the black and white Videcon, by doing
the deflection of the electron beam electrostatically and the focusing electromagnetically. They called this an MF (mixed field) tube, and the early versions of this may be found in their second type 'Rover' camera part no. AVC-3450CE elsewhere on the site. I have an idea that the deflection 'coils' were actually etched conductive 'zig zag' patterns on the external glass envelope of the tube. This would seem to have been developed further for later use in their full broadcast cameras such as the BVP3. I suppose there were two advantages to an MF tube; firstly the deflection 'coils' were much smaller (thinner) and less susceptible to mechanical interference (shock), and the whole tube assembly was more compact as a result. It is also likely that the geometry of the tube would be improved.
The difficult bit was how to make a basic black and white camera tube into a clever colour one.
Usually to do this properly you used 3 separate tubes, each fitted with a colour filer: red, green and blue, and then to have them 'look' through one lens. This always gave the best quality but was hideously expensive because in effect you had 3 complete cameras in one (see the HL97D). With only one tube, the trick was to fool it (and our eyes) that it was also looking through colour filters, and then use some clever electronics to sort things out into a conventional composite colour signal. It was not so easy though to prevent this type of camera from producing some rather strange colour pictures...
With the TRINICON, Sony did the colour filtering on the front of their tube by dividing the
'target' electrode into hundreds (600?) of translucent vertical strips. They were arranged so that the 1st, 3rd, 5th, and 7th etc. strips were each connected together, and the 2nd, 4th, 6th, and 8th etc. were also connected together. Think of a pair of very fine metal combs with very thin fingers meshed together while insulated from another. This pair of microscopic conductive gratings extended across the front of the tube, and the alternately connected rows of vertical electrode strips connected across the secondary of a transformer. By passing an alternating voltage into the primary of this transformer, the two target 'combs' could be made negative and positive to each other, and thus in effect be 'turned on and off'. In front of these odd and even sets of target electrodes were deposited slightly narrower, offset repeating strips of red, green and blue filters. This meant that the first target electrode will 'see' through both red and green stripes (yellow) and the second target electrode will 'see' through both a blue and a green (cyan) stripe. I imagine they must have used integrated circuit fabrication techniques to do all of this. Now it gets interesting; a television picture consists of an interlaced horizontal scanned image of even and odd lines, the first scan is say a scan of 312.5 even (horizontal) lines and the second scan fills in the 312.5 odd lines. So 25 complete images are scanned each second, consisting of 50 interlaced scans of odd and even lines (total 625 in the UK and 525 in the USA). This both decreases the bandwidth needed for a TV picture, and reduces perceived flicker. What Sony did with their TRINICON system was to switch each of their target 'combs' in sync with the odd and even scans, so that one set of target electrodes would look 'through' say the set cyan filters during odd scans, and the other 'comb' would look through the yellow set of filters during even line scans. This gave 25 sets of cyan pictures and 25 sets of yellow pictures every second, which was actually enough information to provide a full colour picture. The rest of the electronics in the camera would combine these cyan and yellow signals by means of a mixture of delay lines and band pass filters, and reconstitute quite a reasonable full colour picture. In fact the pictures from these Sony single tube cameras were rather in advance of most other low end TV cameras, and very much better than those from the very awful domestic colour cameras of the time.
The TRINICON tube was expensive (£600 for a replacement, when I enquired in 1985), but you
only needed one of these, and the processing electronics were comparatively cheap. Oh, and in addition to not having another 2 tubes, no expensive optical splitter was needed. There were drawbacks though: all that extra 'stuff' in front of a fairly insensitive tube meant that lots and lots of light was needed to get a reasonable picture. Indeed in the very old days, it was sometimes cheaper to use expensive (but much more sensitive) B&W plumbicon studio cameras in preference to cheaper videcon ones, because of the higher cost fitting and running an adequate lighting rig for the videcons! The second drawback, was that having an image tube with 2 targets would halve available resolution; the older B&W Sony Rover camera (MF tube but no colour trickery) gave around 500 horizontal lines, but there again for colour, resolution was not very significant considering where the pictures were ending up; Umatic (300 lines resolution) or VHS (240 lines resolution).
Recently looking at the images from this venerable 1640P, I am surprised how natural it's colour
pictures actually are. It's no broadcast camera and it is embarrassingly insensitive (we have become spoilt with CCDs), but when one recalls the very strange colour casts or green highlights, or very odd shadows of many an old domestic colour camera, together with a complete absence of registration artifacts (which you get from badly adjusted 3 tube cameras), this old Sony device seems rather good. The shadows are 'normal', the highlights are white (not green*) and the edges of objects have no 'rainbows', all rather good.
And of course the vertical stripe colour filter system has been used for years in Sony's colour
television picture tubes (TRINITRON).
* Domestic and cheap color video cameras of the 1970's were generally rather nasty. Even
products from very well known manufaturers produced all sorts of quite obvious problems, perhaps most obvious of which was varying amounts of 'Green'. Green highlights, green shadows, poor resolution and low sensitivity were all part of the fun of having a non-broadcast 'colour' camera in those days. It's colour Jim, but not as we know it (these days)! |
|
Below we see a close up of the Trinicon image tube. The copper foil surrounds the
electromagnetic focusing coil. At the front of the tube is a filter wheel for differing lighting conditions (cap, Tungsten, bright outdoors, cloudy), and the central slot towards the rear of the tube is to adjust mechanical 'back' focus. |