VDU-K Interak 1 VDU Interface Section 4: Appendices
This characters shown on this page and the page following are an example of one set of characters that has been defined as a standard character generator set. They are largely based on the printable ASCII characters, but characters have also been allocated to the non-printable codes, such as carriage return, line feed, etc. Some further points to note are given at the end of the table on the next page. Purchasers of the VDU-K card as a "bare board" can obtain a pre-programmed character generator EPROM from their supplier at modest cost.
| Hex. | ASCII | Displays | Hex. ASCII Displays | Hex. | ASCII | Displays | Hex. | ASCII | Displays |
00 01 02 03 04 05 06 07 |
NUL SOH STX ETX EOT ENQ ACK BEL |
¿ Graphics Space ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics |
20 SP ASCII Space 21 ! ! 22 " " 23 24 $ $ 25 26 & & 27 |
40 41 42 43 44 45 46 47 |
SP A B C D E F G |
A B C D E F G |
60 61 62 63 64 65 66 67 |
a b c d e f g |
a b c d e f g |
08 09 0A 0B 0C 0D 0E 0F |
BS HT LF VT FF CR SO SI |
¿ Graphics Space ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics ¿ Block Graphics |
28 29 2A * * 2B + + 2C 2D 2E 2F |
48 49 4A 4B 4C 4D 4E 4F |
H I J K L M N 0 |
H I J K L M N 0 |
68 69 6A 6B 6C 6D 6E 6F |
h i j k l m n o |
h i j k l m n o |
10 11 12 13 14 15 16 17 |
DLE DC1 DC2 DC3 DC4 NAK SYN ETB |
¿ Line Drawing Char. ¿ Line Drawing Char. ¿ Line Drawing Char. ¿ Line Drawing Char. ¿ Line Drawing Char. ¿ Line Drawing Char. ¿ Line Drawing Char. ¿ Line Drawing Char. |
30 0 0 31 1 1 32 2 2 33 3 3 34 4 4 35 5 5 36 6 6 37 7 7 |
50 51 52 53 54 55 56 57 |
P Q R S T U V W |
P Q R S T U V W |
70 71 72 73 74 75 76 77 |
p q r s t u v w |
p q r s t u v w |
18 19 1A 1B 1C 1D 1E 1F |
CAN EM SUB ESC FS GS RS US |
¿ Line Drawing Char. ¿ Line Drawing Char. ¿ Line Drawing Char. ¿ Alien 0 ¿ Alien 1 ¿ Alien 2 ¿ Vehicle L.H. ¿ Vehicle R.H. |
38 8 8 39 9 9 3A 3B 3C 3D 3E 3F ? ? |
58 59 5A 5B 5C 5D 5E 5F |
X Y Z [ ] |
X Y Z [ ] |
78 79 7A 7B 7C 7D 7E 7F |
x
y
z
{
}
DEL
|
x
y
z
{
}
DEL
|
Some points to note on the tables above and the previous page:
The characters for 00 and 20 both display as a blank space; 00 can be used as a "graphics" space, and 20 can be used as an "alphanumerics" space. The character for code 5F (ASCII back arrow) displays as the underline character "_". The character for code 60 (ASCII \) displays as "6", code 7C (|) displays as "÷", and code 7F (DEL),displays as a chequerboard pattern (alternating black and white dots). These changes are to comply with convention in some cases, and special requests in others.
Note that if the reverse video option is selected there are a further 128 characters which are simply the reverse video versions of those listed. Their code is found by adding hex. 80 to the codes listed, so for example 2A is the "*" character, AA (=80+2A) is the same in reverse video.
Section 4.1 of this Manual has given an example of one set of characters for what might be termed a standard character set, which is likely to be adopted by most users of the VDU-K. (A programmed EPROM can be obtained at modest cost by VDU-K purchasers from the supplier of the VDU-K card.)
However, many users may wish to make minor detail changes to the characters, or of course may wish to devise their own, this being one of the main attractions of the card. (It is often a great benefit to be able to have a preprogrammed set of special characters, since this saves the software overhead suffered by ostensibly more sophisticated systems which have a RAM based set, and thus have to burden the main program with the task of storing the required character set within itself.)
The arrangement of the various data within the character generator EPROM has not been decided lightly. As each bit set to a "1" in the EPROM corresponds to a lighted dot in the finished character (displayed in normal, not reversed, video), it might not be thought important which particular addresses are used, as long as they are defined. The arrangement which has been adopted in the VDU-K design has been carefully chosen to make the design and programming of a custom set of "applications" characters as easy as possible.
For example the first byte of the data to compose the character "A" (ASCII code 41H, the "H” suffix meaning hexadecimal notation) is to be found at the EPROM address 410H, i.e. 41H x 10H, similarly the data for code 62H (ASCII letter "b"), are found at addresses beginning 620H.
As each character cell is eight dots wide, an eight bit byte is ideally suited to the purpose of storing the pattern of dots. The convention used in this design is that the most significant bit is the one which appears at the left-hand side of the character cell, and the least significant bit is the one on the right. This makes it quite easy to visualise characters, since if the byte is written out in binary format the 1's will represent lit dots, and the 0’s will represent blank dots. It is perhaps even easier to imagine the appearance of the displayed character if say "*" is used for a "1", and "-" for a "0”.
For example if the contents of 410H in the EPROM were 0FH (actually in the standard character generator they aren't, but say they were), then the top line of the character coded 41H would appear as:
----**** (since 0FH = 00001111)
if they were F0H the top line would be
****---- (since F0H = 11110000)
Similarly for the following data:
Address Data Display Binary (Hex.) 410 00 -------- 00000000 411 10 ---*---- 00010000 412 28 --*-*--- 00101000 413 44 -*---*-- 01000100 414 44 -*---*-- 01000100 415 7C -*****-- 01111100 416 44 -*---*-- 01000100 417 44 -*---*-- 01000100 418 00 -------- 00000000 419 00 -------- 00000000
The example above this time is the actual sequence of ten consecutive bytes which are found in the standard character generator EPROM starting at address 410H, and it may be seen that it represents the letter "A", ASCII code 41H.
Ten consecutive bytes are used to make up the character cell, which is a 8 x 10 matrix of dots. The alphanumeric characters do not extend to the sides of the cell in all directions, as for good legibility a blank space is required between adjacent characters, horizontally, and vertically.
The convention which has been followed in the standard character generator EPROM is to compose all the upper-case (capital) letters on a 5 x 7 matrix, located within the 8 x 10 cell as shown in the example above. For each of the alphanumeric characters the top row of dots in the cell is blank, as is the left column and the two right columns. The lower-case letters are mostly composed on a 5 x 5 matrix which begins two rows lower than the tops of the upper-case letters. Some of the lower-case letters e.g. g, j, p, etc. have "descenders", i.e. the character has parts which are lower than the bottom row of the upper-case letters. Lower-case letters with descenders are composed on a 5 x 7 matrix like the upper-case letters, but displaced downwards by two rows.
If it is required that adjacent characters "join up" then no blank rows or columns should be left. An example of one such character in the standard character generator EPROM is given below; it is the "pixel" character 09H, which is located in the EPROM at address 090H:
Address Data Display Binary (Hex.) 090 F0 ****---- 11110000 091 F0 ****---- 11110000 092 F0 ****---- 11110000 093 F0 ****---- 11110000 094 F0 ****---- 11110000 095 0F ----**** 00001111 096 0F ----**** 00001111 097 0F ----**** 00001111 098 0F ----**** 00001111 099 0F ----**** 00001111
It should be noted that the size of the character cell displayed on the standard VDU-K card is very close to being square. This is of arguable benefit in the case of alphanumeric characters, but it makes the design of graphic displays particularly easy.
The amount of space in the EPROM devoted to each character is, as discussed above, ten consecutive bytes. There are in fact sixteen successive bytes available in the EPROM for each character but the last six are not displayed. They can be any code, for example "FFH", the natural state of an unprogrammed EPR0M, or "00H", the code for blank dots, or indeed anything, as these end groups of six bytes are not displayed.
For certain special applications of the VDU-K, some users may substitute a binary counter type 74LS293 for U10 in place of the normal decade counter type 74LS290, since these two devices are pin compatible. If this is done the screen display can no longer be 24 rows of characters, about 16 rows will be the usable maximum, this being a consequence of the fact that each character cell will be increased to display a 8 x 16 matrix of dots. In this case sixteen successive locations in the character generator EPROM will be used for each character instead of the previous 10, plus 6 spare.
If the application involves a repeating set, such as say a set of chess men then the design of each individual character should be carried out on graph paper. A chess men EPROM has already been designed, and it proved very convenient to make each piece occupy a matrix of 16 x 20 dots, i.e. four character cells, two horizontally, two vertically. Other applications may be best satisfied by a different arrangement of character cells, e.g. representations of say sheep, pigs, and cows (for say kindergarten use, teaching children to read, or perhaps for some serious use down on the farm), could be made in a matrix of say 32 x 20, i.e. eight character cells each, four horizontally, two vertically.
Once the shape and format has been designed the various dots can be divided into their eight bit bytes, and programmed into the Applications EPROM at convenient locations.
An alternative type of screen display is one which is not repetitive. An example of this would be a representation of a section of the British Isles for say Meteorological use in preparing a weather map. In such an application there would be fixed features, and repetitive ones (e.g. sun and cloud symbols).
At first sight there might be thought to be insufficient character cells available, since at least one is needed for each section of coastline. This latter statement is true, but there will still be plenty of characters left over when it is realised that the British Isles is mostly land, surrounded by mostly sea. Although the sea for example might take the greatest area on the screen it could possibly be represented by just a single "sea" character (ASCII code "43" perhaps, ho ho), leaving plenty for the other details.
As before graph paper should be used to design the screen layout, but before the EPROM is programmed it should be decided what character codes will be used in the area of interest. For totally patternless areas such as geographical displays, it might be most convenient just to fill that area of screen with ascending character codes, e.g. 80H, 81H, 82H, 83H, etc., and program the features required into the appropriate codes, using perhaps a co-ordinate system. Once written the character codes in the video RAM would not change, unless it was desired to replace for example a piece of Wales with a piece of Scotland.
The TV sync. pulses and other timing pulses are produced by the on-board Z80A-CPU, running a program which is one of those contained in the timing generator EPROM U2. The programs are very simple in concept, but quite complicated in execution.
In this Section of the Manual only an indication of the methods used will be given, for two purposes:
Firstly so that the interested user can say with confidence that he knows the precise purpose of every component part of his Interak 1 System, and secondly to give a certain amount of guidance to those users who are wishing to use the technique to construct their own timing EPROM for some special modification to the VDU-K card, or perhaps even some similar design of their own.
The Z80A-CPU and the EPROM together make one of the very smallest computer systems imaginable. No external RAM is used as the CPU chip itself contains sufficient for the purpose. The principle of operation is very simple and consists of writing chosen data to a storage latch. The outputs of the latch are the required timing signals. In essence, if a "1" is written to a "bit" of the latch its output goes high, and stays high until a "0" is written, at which point it goes low. Since the clock input to the CPU is crystal controlled, the signals obtained from the outputs of the latch are very precisely controlled, and are absolutely stable, and repeatable from VDU-K card to VDU-K card.
The various time intervals required for a TV sync. pulse generator are made up from such short intervals as 2.33 microseconds. (See Section 5.2 of this manual for the Timing Diagrams). Intractable figures such as this become fairly easy to work with when a 3.0 MHz clock is used for the CPU: With a 3.0 MHz clock 2.33 microseconds is 7 "T" states (a "T" state being the smallest unit of time considered in the operating cycle of a Z80A microprocessor and is the lencth of time for one cycle of the clock). Reference to the precise time of the instructions available in the Z80A's repertoire shows that many of them (e.g. LD (HL),A) are exactly 7 "T" states long. The length of a whole TV line (64 microseconds) is a whole number of ”T" states (actually 192), and similarly the time duration for the visible part of the display (42.66 microseconds) is exactly 128 "T" states long.
It should be stressed that when time delays as little as 7 "T" states are required such circuits as a Z80A-CTC (Counter Timer Circuit) provide no benefit. Even with the Z80A-CPU's fast interrupt response the time taken would be much too long, to say nothing of the fact that the extra chips required would remove the outstanding benefit of low cost and small chip count. It is for similar reasons that some of the counting which could have been done in software has in fact been carried out by hardware (e g. the cost of two four bit counters in the shape of a 74LS393 is not so very different to that of an eight bit latch, especially when extra decoding is added to enable the latch, and as a bonus the hardware counter occupies less board space).
In this design the only thing which is read by the CPU is the EPROM, and the only thing which is written to is the latch, and so there is no need for any decoding or differentiation between I/O space and Memory space; the !RD (read) line from the CPU enables the EPROM directly, and the !WR (write) line clocks the latch.
Although the VDU-K card requires only four timing lines (shown on the diagram as V, H, R, S), a six bit latch has been provided to allow for unspecified future applications (for example to provide separate line and frame sync. pulses for TV monitors which cannot accept both line and frame sync. combined).
A novice might be concerned that the action of changing the output of one of the timing lines might disturb one of the others, since all bits in the latch are written to at the same time. There is no need to worry about this as there is in fact no problem. The secret is to remember that if a line is to remain unchanged at a particular point in the timing cycle, it is perfectly acceptable to write data to the appropriate "bits" of the latch, provided they include the same data for those "bits", i.e. continuously writing "1's" on top of "1’s" will result in constant unchanging "1‘s" as outputs, and similarly for "0's" written repeatedly on top of "0's".
In the case of a non-interlaced display each consecutive TV frame has 312 picture lines, and the control program continuously loops round a sequence of instructions which take 312 x 64 microseconds to complete, i.e. 59,904 "T" states, no more, no less. An interlaced display has a sequence which is exactly 625 lines long (ideally suited to a 625-line TV set), i.e. exactly 120,000 "T" states. This latter sequence is divided into two consecutive TV frames, one having 312 picture lines, and the next having 313 (312 + 313 = 625). In the case of an interlaced display the commencement of the frame sync pulse alternates between starting at the beginning of a TV line, and half way through it. The effect of this on the TV screen is to cause adjacent frames of an interlaced display to have lines which are displaced by half the distance between them. This gives a superior display for TV broadcast pictures viewed from a distance of several metres, but the technique is of dubious benefit when computer pictures are viewed close-up. So that all schools of thought may be accommodated, the standard timing generator EPROM supplied includes programs for each format.
There are some diagrams in Section 5.2 of this Manual which illustrate the description in the previous paragraph, but as the authors of the manual prefer to think of themselves as computer boffins rather than TV boffins, the reader is referred to any book on the principles of television for a fuller and more lucid description.
It is hoped that a full listing of the program will be made available for an appropriate fee, in order to aid any users who wish to write their own, or modify that provided. Such a listing does not exist at the time of writing, since much of the required timing was derived empirically. Part of the reasons for this will be appreciated if the following caution is understood:
Any users who are writing their own timing program should not fall into the trap of thinking that a given output will persist for the duration of the instruction which caused it. For example it might be thought that an instruction 7 "T" states in long which writes a "1" to the output latch (assumed to be previously "0"), followed by an instruction 10 "T" states long which writes a "0", would result in a positive pulse of duration 7 "T" states. This is not normally the case; in the example above the pulse would have a duration nearer the 10 "T" state mark. The reason for this is that the effect of a write instruction only occurs towards the end of that instruction, the early part being taken up with the op-code fetch, the refresh cycle, and so on. In other words, for the majority of the 7 "T” state instruction above the output was "0", and remains "0". It goes to "1", somewhere near the end, and then remains at "1" for most of the 10 "T" state instruction which follows it. After about 10 "T" states the instruction takes effect and the "1" output returns to "0". To summarise, in this example a 7 "T" state instruction to produce a positive pulse results in a pulse which in fact turns out to be 10 "T" states long.
It can be appreciated that the task of writing a program which depends for its operation on precise selection of instructions having the desired number of "T" states, is a task which is made doubly difficult when the timing must be taken not from the instruction under consideration, but from the instruction which follows it.
The diagrams in Section 5.2 illustrate on example of the results of such labour. The program to produce the desired waveforms is at first extremely difficult to understand because it follows none of the normal logical rules of programming; most of the instructions are concerned solely with filling in time (counted out with meticulous accuracy to the individual "T" state). When the time comes for a decision in the program, for example after 240 visible picture lines have been displayed, and it is time to move into the blank bottom margin, then normal relative conditional jumps cannot be employed, since they take a different number of "T" states according to whether or not the condition has been met. In the same way care has to be taken at the point when the whole program loops back to begin again, to ensure that an extra 10 "T" states are not inserted inadvertently into one of the final picture lines when the jump is made. In the case of the very tiny time interval 2.66 microseconds (i.e. 7 "T" states) the time taken by an absolute conditional jump (10 "T" states) is too long. This happens when the five "half-line equalisation pulses" are to be generated, and it proves expedient to produce these five pulses by writing the code out five times in a row rather than to set up a loop counter, which would be the conventional way in a "normal" program.
Even with this necessary wastefulness of memory space the program to do the whole job occupies only about 0.5K, and there is space for four such programs in the standard type 2716 EPROM. Two of the four programs are switch selectable, and are generally the same format, but for interlaced and non-interlaced displays. The other two are link-selectable (by breaking a small piece of track), and could be some other format, for example a 16-line display instead of 24-lines, or perhaps to suit a 525-line 60Hz TV set instead of the usual 625-line 50 Hz type.
There are 32 x 24 = 768 locations on the VDU screen, each of which corresponds to a unique memory location in the memory map of the computer. The actual addresses will of course depend on the settings of the switches S2a-f, which are under the control of the user of the VDU. However, if they are set to suit ZYMON 2 (i.e. S2a-f ’on', with S2g-h "off"), then the address of the top left-hand corner of the VDU will be F000H, and the addresses will follow consecutively until the bottom right-hand corner of the VDU is reached at address F2FFH.
The table which follows gives the start and end addresses for each row of characters which can be displayed on the VDU screen (the addresses are given in hexadecimal notation, the "H" suffix being omitted to make them a little easier to read).
| Row | 0 ... 32 Columns ... 31 | |||||||||||||||||||||||||||||||||
| 0 | F000 | F01F | ||||||||||||||||||||||||||||||||
| 1 | F020 | F03F | ||||||||||||||||||||||||||||||||
| 2 | F040 | F05F | ||||||||||||||||||||||||||||||||
| 3 | F060 | F07F | ||||||||||||||||||||||||||||||||
| 4 | F080 | F09F | ||||||||||||||||||||||||||||||||
| 5 | F0A0 | F0BF | ||||||||||||||||||||||||||||||||
| 6 | F0C0 | F0DF | ||||||||||||||||||||||||||||||||
| 7 | F0E0 | F0FF | ||||||||||||||||||||||||||||||||
| 8 | F100 | F11F | ||||||||||||||||||||||||||||||||
| 9 | F120 | F13F | ||||||||||||||||||||||||||||||||
| 10 | F140 | F15F | ||||||||||||||||||||||||||||||||
| 11 | F160 | F17F | ||||||||||||||||||||||||||||||||
| 12 | F180 | F19F | ||||||||||||||||||||||||||||||||
| 13 | F1A0 | F1BF | ||||||||||||||||||||||||||||||||
| 14 | F1C0 | F1DF | ||||||||||||||||||||||||||||||||
| 15 | F1E0 | F1FF | ||||||||||||||||||||||||||||||||
| 16 | F200 | F21F | ||||||||||||||||||||||||||||||||
| 17 | F220 | F23F | ||||||||||||||||||||||||||||||||
| 18 | F240 | F25F | ||||||||||||||||||||||||||||||||
| 19 | F260 | F27F | ||||||||||||||||||||||||||||||||
| 20 | F280 | F29F | ||||||||||||||||||||||||||||||||
| 21 | F2A0 | F2BF | ||||||||||||||||||||||||||||||||
| 22 | F2C0 | F2DF | ||||||||||||||||||||||||||||||||
| 23 | F2E0 | F2FF | ||||||||||||||||||||||||||||||||
The purpose of providing the above table is to give a guide for users who are designing a screen display, and therefore need to know where each line of characters begins and ends.
The difference between an interlaced and non-interlaced display has been mentioned at appropriate places in this manual, in the general description, the detailed description, switch settings, timing generator EPROM description, and so on. In each section the subject of interlaced and non-interlaced display has been approached from the point of view of the needs of the user at the time he is reading that section of the manual.
In this appendix the subject is covered again simply to provide a handy reference point, to save the casual reader from having to wade through the whole manual to find the information.
Commonly the type of TV receiver for which the VDU-K is designed is described as "625-line". There are not many people who have tried to count the number of lines on a TV picture and have remained sane, but those who have will be able to confirm that not all are displayed; there are a number missing.
The missing lines occur at the end of the picture when the electron beam which forms the picture is returning to the top of the screen. They are blanked in the TV receiver, because they would spoil the display as they spot is travelling upwards much faster than it travelled down. The blanked lines are known as "frame flyback" lines, and they are only visible on badly adjusted TV receivers.
Thus there are less than 625 lines displayed on a 625 line TV receiver. In fact the 625 lines mentioned above are achieved by an illusion, only present in an "interlaced" display. In an interlaced display consecutive "frames" are alternately "odd" and "even". (A "frame" is a complete scan of the whole screen, which takes place in 20 milliseconds, i.e. one fiftieth of a second, as there are fifty frames displayed per second.) Since there are two types of frame, "odd" and "even" it can be seen that there are twenty-five of each type to make up the total of fifty. Including the blank lines, there are 312.5 lines to each frame in an interlaced display. The consecutive odd and even frames are displaced vertically by one half the distance of separation of adjacent lines by an ingenious method of timing the frame synchronisation pulse, and due to the effect known as "persistence of vision" the viewer sees both sets simultaneously (excluding of course the invisible blanked lines) and thus observes 312.5 + 312.5 = 625 lines, Q.E.D.
The ingenious method referred to above was devised by the television engineers of long ago and is to arrange that the frame flyback starts half way along the the last line of an odd field, and at the end of the last line of an even field. Successive fields therefore start at a time which is different to the preceding field to the extent of half a line, which results in a vertical difference of one half the spacing of adjacent scan lines.
At the end of this section of the manual there is a diagram which attempts to clarify the above description, and demonstrate visually the difference between an interlaced and a non-interlaced TV display.
It will be seen in the case of the interlaced display that the number of lines displayed for each frame is a whole number, which may seem to conflict with the description above referring to a fractional number (312.5). This is fairly easily resolved when it is appreciated that the time taken for the frame flyback includes the missing half line.
By comparison, the non-interlaced display is quite simple. In a non-interlaced display all fields are the same "hybrid" type i.e. they start as "odd" fields and finish as "even". The extra half line which was used to cause the alternation of frames to give the interlaced display is removed, and each frame consists of 312 lines. Consequently the time taken to complete fifty frames is reduced slightly, and the frame frequency is thus increased from 50.00 Hz to 50.08 Hz. This is entirely insignificant in this application. (It would however be important if the signals from this card were being blended into an existing TV display, e.g. to insert subtitles etc. into a broadcast TV picture. In such a case the 12.0 MHz master crystal would have to be reduced in frequency slightly, or perhaps more practically, the interlaced option of the timing generator should be chosen for this application instead of non-interlaced )
Ostensibly the interlaced format is the superior one, and indeed it is for moving conventional broadcast TV pictures. Viewed from a few metres away the scan lines can hardly be seen. However things are a little different for close-up viewing of computer style material. Many people report a disturbing flicker which makes the interlaced display less than ideal for viewing for a long time. The probable reason is that each of the two types of frame ("odd” and "even") in an interlaced display is only repeated every other frame, i.e. twenty five times per second. This frequency is low enough to cause visible flicker, particularly when, as is the case in a typical computer display, there are only sharp changes from white to black, rather than hazy changes of intensity from white via grey to black.
When this point is considered it can be sean that there is a chance that a non-interlaced display may prove to be without the disadvantages mentioned above. As the same frame is repeated fifty times a second rather than the twenty five times a second, which is the case when the "even" frame is interlaced with the "odd" one, the resulting picture is much steadier, and in most cases is preferred by users who spend a lot of time close up to their screen.
If special computer display monitors with long-persistence phosphors are used then the effect of flicker is much less pronounced and the interlaced display might give subjectively superior results. (Note that although long persistence phosphors are often green, the reverse is not true. Green screens on TV monitors are most often not long-persistence types. This could have a lot to do with the description "TV monitor"; if a monitor is actually to be used for TV pictures, it cannot really have a long persistence screen, because if it did it would give a most unacceptable "smearing" of quick movement in pictures on the screen )
Interlaced Display Odd field commences one half-line earlier than the even field. ______Odd field - - - - Even Field |
Non-Interlaced Display |
| The 312.5
lines in the odd field are 'interlaced' with the 312.5
lines in the even field to form the conventional 625-line
TV display (312.5 + 312.5 = 625). Not all the lines are
visible, and the half-line in each frame is absorbed
during the frame flyback time, which is not a whole
number of lines. Note that the original picture elements are refreshed only in alternate frames, hence a long-persistence phosphor is needed to avoid flicker when the screen is viewed close up. However a long persistence phosphor will cause 'smearing' if the picture includes any animated elements, so some trade-offs are involved in interlaced displays.
|
A 'hybrid'
field (starts 'odd', finishes 'even'), is repeated in
every consecutive frame. Individual picture elements are
therefore refreshed in every frame, i.e. twice as often
as in the interlaced display. In many cases a steadier,
more pleasing picture is the result, particularly on a
screen with a short-persistence phosphor (e.g. a video monitor,
with or without green screen, or a domestic TV receiver).
|
| Drawn: DMP Date: 1982-11-09 |
|
At the end of this Section is a <diagram which illustrates the effect. It is a symbolic diagram rather than an actual picture of a video waveform, and it shows the effect of a.c. coupling in video circuits. In general, the effect of a.c. coupling is detrimental, and for this reason the video output from the VDU-K card is not connected via a coupling capacitor.
From the diagram it can be seen that the average (sometimes called the "d.c.") level of a video signal which is a.c. coupled will depend quite markedly on the picture content, (This effect is due to a property of a.c. coupled circuits, and any text-book on basic a.c. theory will provide further explanation should the reader require it.) If there is a lot of white in the picture, the average level will sink down, and it will rise again if the picture contains a lot of black instead of white. The TV receiver or monitor will have contrast and brilliance controls which will permit a good picture to obtained for one or other of these extremes but not both. The diagram shows that if a setting is found which is right for one condition it will not be right for the other.
This subject is raised here simply so that the user who finds this trouble will mislead himself into thinking there is something wrong with his VDU-K card; the "fault" is in the TV or monitor.
The effect will be demonstrated most clearly if the TV or monitor controls are adjusted for a picture which is predominantly black, with just one or two words displayed normally. If a program is then run which turns most of the screen white (e.g. a program which writes a lot of inverse video characters on the screen), then it is often the case that the orignal (normal video) characters will become almost illegible until the brilliance and contrast controls are adjusted on the TV monitor. This is due to the shift in average level in the video signal, which is demonstated on the diagram.
It is generally not possible to d.c. couple the video throughout in the TV receiver or monitor because of the difficult circuit design problems this would cause, however a top-quality TV receiver or video monitor will have a few extra components added in its circuit to provide a feature known as "black-level clamp". With this feature the undesirable effect of a.c. coupling the video signal will be removed, as the black level clamp forces the black level (and thus the white level) to be held at a fixed potential, so that it does not change with picture content.
Almost all TV receivers are built on a very low budget and their designers cannot justify the luxury of "black level clamping”. Partly the reason is to do with the demands this would place on the regulation of the set's EHT (Extra High Tension, i.e. high voltage) supply. As a.c. coupling in the video circuits results in an all black screen being lightened to grey, and/or an all white screen being darkened, there is far less change than there would otherwise be in the demands made on the EHT supply (which would otherwise be very low for a black screen and very high for a white screen). The very poor EHT regulation on some sets is very visible on those which change the size of the picture according to the intensity of the picture being transmitted.
It might be thought that a video monitor would not suffer these defects being built to far more stringent standards (and thus costing a lot more into the bargain!). However nowadays most so-called monitors are nothing more than TV receiver designs simply with the aerial and tuning circuitry stripped out and a green phosphor screen fitted, and so they inherit the deficiencies of the TV receivers from which they are derived.
There is little that can be done to improve matters if the described effect causes any inconvenience; even if the minor circuit changes required to provide "black level clamp" can be devised, they should not be undertaken lightly since they may result in an intolerable burden being placed on the insulation and general construction of the EHT supply if its voltage is caused to rise significantly.
As stated before in this manual, the authors do not feel highly qualified as teachers when it comes to discussing matters of this kind, and so the reader is directed to seek guidance from a any suitable reference book on the subject of television receiver and TV monitor design and kindred matters.