Fullscreen HTML5Med Res (??MB)HTML4

By JIM ROWE Dr Video An Easy-To-Build Video Stabiliser Do the pictures on your TV or video projector jitter and jump around when you’re trying to watch a movie on VHS or DVD? This is usually caused by the hidden pulses that are added to a lot of pre-recorded video software, to prevent illicit copying. Here’s a low cost circuit that removes most of these nasties, cleaning up the video for more stable viewing. 30  Silicon iliconCChip hip Y OU’RE PROBABLY AWARE that nowadays a lot of pre-recorded video software is “copy protected”, to stop people from making their own pirate copies. In principle that’s fair enough, too – having spent millions of bucks making a movie, the producers are entitled to get a fair return on their investment. What complicates the situation is that the system that’s used to prevent copying involves adding extra “dancing pulses” to the normal video signal. Unfortunately, this can stop quite a few TV sets and projectors from displaying a steady picture during legitimate viewing. In particular, the extra pulses can cause problems with large-screen TVs that display the picture at 100 fields per second (100Hz) to reduce flicker and also with projectors that perform line and pixel doubling to improve picture clarity. They can cause problems with older conventional TV sets, too. If you have one of these sets or projectors, often the only way to get a steady picture is to somehow remove those extra pulses. The idea is to “clean up” the video and let the set’s sync circuitry do its normal job without interference. And that’s exactly what this little project is designed to do. Note that once the offending pulses are removed, it may also become possible to record the video. However, we want to stress that this project is NOT designed to allow recording – it’s intended purely to allow you to achieve stable and steady pictures for viewing. It is illegal to record copyright material and there are heavy penalties for doing this. We must therefore warn you specifically against using the project to do so. As well as removing most of the copy protection pulses, Dr Video also allows you to apply a small amount of high-frequency boost to the video, to “sharpen” the picture a little when you’re watching movies on older VHS tapes (which are often a little “soft”). However, you can switch off this sharpening when it’s not needed – when you’re watching DVDs, for example (these are usually quite sharp enough already). Dr Video is housed in a compact low-profile instrument box, and runs from a nominal 12V DC source – such as a battery or plugpack. You should also be able to build it for considerably less than other stabilisers. Fig.1: this scope shot shows the extra sync & “dancing” pulses (righthand end of top trace) that are added following the vertical sync block. These pulses constantly change amplitude. Fig.2: a close-up of the “fake” sync and dancing pulses on one of the lines in the vertical blanking interval. How it works Before we look at the circuit diagram, it may help if I explain a little about the copy protection pulses we’re trying to remove. By the way, we’re talking here about the pulses added to video signals in the Macrovision copy protection system, as this is the one most commonly used. To thwart illicit recording, the Macrovision system adds three main sets of pulses to the video signal – two of them essentially combined. First, there’s the “dancing” pulses, which are added to as many as 14 of the normally “black” lines which follow the vertical sync pulse “block”, in the vertical blanking interval (VBI). This is a group of lines that correspond to the vertical retrace time, when the scanning electron beam in the picture tube is being returned from the bottom of the screen back to the top, to begin the next video field. To each of these 14 or so VBI lines, the Macrovision system adds as many as seven extra “fake” horizontal Fig.3: the end of field (EOF) pulses consist of a series of narrow positive pulses that are added to the lines at the very bottom of the picture. April 2001  31 32  Silicon Chip Fig.4 (left): the circuit diagram for the Dr Video. Sync separator IC4 and its associated circuits generate gating signals which operate CMOS switches IC2c and IC2d, to strip off any extra sync and dancing pulses present on the vertical blanking interval lines. sync pulses, each of which is immediately followed by a short “fake video bar” pulse – which can have an amplitude anywhere between black and peak white. And it’s these “fake video bar” pulses which slowly vary up and down in amplitude or “dance”, usually in two or three groups. Figs.1 & 2 show the details of this. The top trace in Fig.1 shows the groups of pulses on eight lines after the vertical sync block, while Fig.2 is a close-up of the pulses on one line. In theory, these VBI pulses shouldn’t upset the operation of the sync separator circuit in a TV or projector – but they are intended to play havoc with the sync locking servo and recording level AGC circuitry of a video recorder. In particular, the extra sync pulses muck up the sync locking, while the “dancing video bars” fool the recorder’s AGC circuitry into varying the record­ing gain up and down to compensate. All of which they indeed do but unfortunately the havoc isn’t restricted just to VCRs! The remaining set of pulses that are added into the video signal are the socalled “EOF” or end-of-field pulses. These consist of a series of narrow positive pulses which are added to the lines at the very bottom of the picture and are timed to coincide with the colour synchronising “bursts” (ie, they are inserted just after the horizontal sync pulses). In effect, these pulses push the colour bursts for these lines right up into the peak white region, so that the black level and colour locking circuitry of a VCR are again tricked. Fig.3 shows what the EOF pulses look like on an oscillo­scope. The EOF pulses are considerably harder to remove than the fake sync and dancing-video-bar pulses in the VBI group. Luckily, though, they don’t seem to cause nearly as much havoc with TV sets and projectors as the VBI pulses. So in the Dr Video pro­ject, we take the same practical approach adopted in many other video stabilisers: we concentrate on removing the VBI pulses and allow the relatively innocuous EOF pulses to remain. By now, you should have a good idea as to what we’re trying to achieve in the Dr Video circuit. Now let’s see how it’s done. Circuit details Fig.4 shows the full circuit of the Dr Video project. It might look a bit complex at first glance but it’s really not as bad as it looks. We’ll describe it section by section. The incoming video arrives at CON1, where we terminate it with the correct 75Ω load. We then couple it to the non-inverting input (pin 3) of IC1, a 5534 op amp used here as a wideband video input buffer. The 0.22µF coupling capacitor removes any DC com­ponent in the incoming video and, together with the 1MΩ resistor and BAW62 diode, forms a simple “DC restorer”. This clamps the sync pulse level to about 0.6V above ground. (D4 and D5 produce a DC level of 1.2V but this is offset by a drop of 0.6V in D6). IC1 is connected as a voltage follower with a gain of one, so a replica of the incoming video therefore appears at its output (pin 6). This signal is then taken in three directions. We’ll look at two of these shortly but first we’ll concentrate on the path that leads April 2001  33 Fig.5: install the parts on the PC board as shown here. Note that some of the links are quite close to each other so use insulated wire for these. Note that the ICs don’t all face in the same direction. down via the 680Ω resistor. As you can see, this feeds the video signal to the input of IC4 (via a 0.1µF capacitor. IC4 is a very handy LM1881 video sync separator chip. The 680Ω series resistor and paralleled 470pF and 39pF capacitors to ground form a low-pass filter, to “lose” the signal’s colour information (which can disturb the LM1881’s operation). The 0.1µF coupling capacitor simply blocks the DC component, while the 680kΩ and 0.1µF capacitor from pin 6 of the LM1881 to ground set the chip’s internal timing circuitry for accurate and stable sync separation. The LM1881 provides a number of outputs but here we only need two of them. First, from pin 3, we get a negative-going vertical sync pulse about 230µs wide. Second, pin 5 gives a series of narrow pulses (again negative-going) which correspond to the video signal’s colour sync bursts – ie, “burst gating” pulses. Next, we invert both these pulses 34  Silicon Chip using IC5e and IC5f, to convert them into positive-going form. We then pass them through differentiator circuits, to obtain narrow negative-going pulses derived from their trailing edges. For the vertical sync pulses this is done by the 390pF capacitor, 10kΩ resistor and D9, while for the burst gating pulses it’s done by the 270pF capaci­tor, 2.2kΩ resistor and D10. Each of these narrow pulses is then used to trigger a simple non-retrigger­ able monostable or “one shot” circuit, to produce longer pulses of carefully set length. Each one-shot consists of a flip-flop formed by two cross-coupled NAND gate elements, plus an RC timing circuit and a Schmitt inverter. The one-shot formed by IC6d, IC6a and IC5a is used to pro­duce a pulse about 1.1ms long, starting at the end of the verti­cal sync pulse from IC4. The end of this output pulse corresponds closely with the end of the VBI, so therefore it “covers” all of the lines which should ideally be “black” but can have added Macrovision nasties. The second one-shot formed by IC6c, IC6b and IC5b is used to produce a much shorter pulse, about 50µs long, starting at the end of each colour burst gating pulse from IC4. This one-shot’s output pulse therefore lasts for most of the “active” part of each horizontal line, and certainly “covers” that part of the VBI lines where the extra sync and “dancing” pulses occur. As you can see, the output of the upper one-shot is then gated with an inverted version of the vertical sync pulse from IC5e using NAND gate IC3a – which together with IC5c forms a positive-logic AND gate. This gating is done because the LM1881 can itself be disturbed by the Macrovision pulses, which occa­sionally cause its vertical sync pulse output from pin 3 to begin early. This can in turn cause our one-shot to trigger early. However, the gating ensures that if this occurs, the oneshot’s output pulse is “blocked off” until the end of the vertical sync block. (We don’t want to change this part of the video signal, of course). So the output from IC5c is a pulse which is “high” for all of the lines between the end of the vertical sync pulse and the end of the VBI. And this is gated with the 50µs pulses from the lower one-shot using NAND gate IC3b. This means that the output of IC3b will go low for the active part of each line between the end of the vertical sync pulse and the end of the VBI – but ONLY for those lines. We’ll get back to these pulses shortly. For the moment, though, let’s turn our attention to NAND gate IC3d. As you can see, one input of this gate is fed with the positive-going burst gating pulses from IC5f, while the other input receives a nega­tive-going 50µs pulse from the output of IC6b, in the lower one-shot. What’s the idea of this gating? Again, it’s needed because of the way that the LM1881 (IC4) can itself be upset by the Macrovision pulses. In this case, “extra” burst gating output pulses can be produced during the active part of the VBI lines, at some points in the “dancing pulses” cycle. By using IC3d to gate the burst pulses with the complementary output of the 50µs one-shot, we make sure that these unwanted extra pulses are “gated out”. As a result, the output of IC3d only goes low for the 12µs duration of the Everything fits on the PC board, so there is no external wiring to the front panel components or to the sockets on the rear panel. “real” colour bursts. IC3d’s output is used to drive the gate of analog bipolar switch IC2b. This switch is used as a simple pulse inverter, with its “output” pin connected to the +5V supply via a 2.2kΩ resis­tor. So the output (pin 3) provides a train of positive-going burst gating pulses. These in turn are used to turn on switch IC2a, which therefore conducts during the colour burst period of every video line. And when IC2a turns on, it allows the following 0.22µF capacitor to charge via the 2.2kΩ resistor, to the current average value of the video signal from IC1. Why on earth is this done? Well, by convention, the average value of a video signal during the colour bursts is used to establish the signal’s black/ blanking level. So by turning IC2a on only during the burst periods, we ensure that the 0.22µF capacitor charges to a voltage which corresponds closely to the video signal’s black level. The last step Right, so now we have the 0.22µF capacitor providing a black level voltage, plus some pulses available from IC3b which go low only during the active part of the VBI lines. The last step in cleaning up the video signal is to put these pulses to work. As shown on Fig.4, the pulses from IC3b are fed directly to the gate of analog switch IC2c, which is in series with the “top” video path from IC1. As a result, IC2c will be turned off during the active part of the VBI lines but left on at all other times. At the same time, IC3c is used to invert the pulses from IC3b and supply these to the gate of CMOS switch IC2d – which is connected between the output of IC2c and the 0.22µF capacitor. This means that when IC2c is turned off to block the video, during the active part of the VBI lines, IC2d is turned on to clamp the video output to black level. Still with me? Essentially, all of the circuitry around IC3, IC4, IC5 and IC6 acts to produce some fast gating signals. These signals operate CMOS switches IC2c and IC2d, to strip off any extra sync and dancing video pulses present on the VBI lines and turn the lines back into nice plain black. As a result, we get a “cleaned up” video signal across the 100kΩ resistor at the output of IC2c and IC2d. Output amplifier The circuitry to the right of the 100kΩ resistor is a wide­band video output buffer amplifier, with transistors Q1 and Q2 forming the input stage and Q3 the output stage. Q4 forms a constant current load for Q3, to allow it to drive a relatively low impedance external load via a 75Ω series or “back April 2001  35 The rear panel carries two RCA sockets (Video Out & Video In) plus a 2.5mm DC panel socket for the external plugpack supply. terminat­ ing” resistor. Note that the reference voltage for the base of Q4 is established by the “power” LED (LED1), which therefore also acts as a pseudo-zener reference diode. The voltage gain of the output buffer amplifier needs to be 2.0, to compensate for the loss in the back terminating resistor. This gain is set by the two 470Ω resistors, which provide nega­tive feedback to the base of Q2. However, as you can see, there’s also a 330Ω resistor, connected via a 47µH RF inductor to one of the poles of switch S1. When S1 is switched to the “Sharp­ from +5V DC, while the input and output video ampli­fiers run from ±5V. As a result, the power supply is fairly straightforward, since we can easily derive these rails from any suitable source of 10-12V DC. Diode D1 is connected in series with the supply input, to protect the circuit against reverse polarity damage. This then feeds a 1000µF filter capacitor and a 3-terminal regulator (REG1), which provides the +5V supply rail. This rail is filtered using a 100µF capacitor plus several 0.1µF bypass capacitors scattered around the circuit. The -5V supply rail is produced using 555 timer IC7. This is used as a self-oscillating “commutator switch” en” position, an 82pF capacitor is connected to the inductor, forming a series resonant circuit at about 2MHz. The “Q” is quite low though, because of the series 330Ω resistor. As a result we only get a “boost” of about 4dB – enough to give a useful amount of sharpening to a “soft” video picture. The second pole of S1 is used to turn on a second indicator LED (LED2), to show when this sharpening is taking place. Power supply The sync separator chip (IC4) and most of the logic circui­try operates Table 1: Resistor Colour Codes  No.   1   1   3   2   4   1   1   2   3   1   2   1 36  Silicon Chip Value 1MΩ 680kΩ 100kΩ 10kΩ 2.2kΩ 1.5kΩ 1kΩ 680Ω 470Ω 330Ω 75Ω 47Ω 4-Band Code (1%) brown black green brown blue grey yellow brown brown black yellow brown brown black orange brown red red red brown brown green red brown brown black red brown blue grey brown brown yellow violet brown brown orange orange brown brown violet green black brown yellow violet black brown 5-Band Code (1%) brown black black yellow brown blue grey black orange brown brown black black orange brown brown black black red brown red red black brown brown brown green black brown brown brown black black brown brown blue grey black black brown yellow violet black black brown orange orange black black brown violet green black gold brown yellow violet black gold brown Table 2: Capacitor Codes             Value IEC Code EIA Code 0.22µF   224  220n 0.1µF   104  100n .012µF   123   12n .01µF   103   10n .0082µF   822   8n2 470pF   471  470p 390pF   391  390p 270pF   271  270p 220pF   221  220p 82pF    82   82p 39pF    39   39p and drives a “charge pump” rectifier circuit consisting of D2, D3 and the two 220µF capacitors. This produces a source of unregulated -10V DC, which is then passed through 3-terminal regulator REG2 to produce the -5V rail. In this case, we can use such a simple charge-pump circuit to generate the negative rail because the current needed from it is quite low. Construction Building the Dr Video project is very straightforward since all the parts are mounted on a PC board coded 02104011 (117 x 112mm). This board assembly fits snugly into a standard low-profile plastic instrument box measuring 140 x 110 x 35mm. The front panel is anything but intimidating. There’s only one control (ie, the Normal/Sharpen switch S1) plus the Power and Sharpening indicator LEDs. On the rear panel, there’s just the video input and output sockets, plus the DC input connector. All these connectors mount on the PC board, along with S1 and the two LEDs, so there is no off-board wiring at all. Fig.5 shows the assembly details. There are eight wire links on the board and it’s probably a good idea to fit these first. You can use tinned copper wire for many of these, although I suggest you use insulated wire for at least one link where there are two running close together (just to the left of IC6, for example). This will help prevent unwanted shorts. Next, I suggest you mount the DC input connector and the two video sockets. Note that the holes for these may need enlarg­ ing slightly with a jeweller’s rat-tail file before the connector lugs will fit through. Make sure the connectors are bedded down squarely against the top of the board before you solder the lugs to the board pads. Switch S1 can also be mounted at this stage. Push is down squarely against the board before soldering its leads and don’t forget the two “hold down” lugs near the front of the switch (these lugs stop the switch from moving when it is operated). With this done, you can add the various electronic parts, in the usual order. Start with the resistors and small capacitors, then fit the diodes and electrolytic capacitors – taking care with their polarity. The next stage involves fitting the transistors and ICs, again taking care with their polarity. As usual, take steps to minimise the risk of ESD (electrostatic discharge) damage when handling and fitting the CMOS devices in particular. Use an earthed soldering iron and wear a wrist-grounding strap if you like. You should also solder the supply and ground pins of each IC to the board pads first, before soldering the remaining pins. Use a 10mm long M3 machine screw and nut to secure the positive regulator (REG1) – this device does get warm and the screw and nut provide a small amount of heatsinking in conjunc­tion with the board copper. It isn’t strictly necessary to do this for REG2, as this device runs virtually cold. However, it’s still a good idea to secure it, just to stop it “flapping around” and placing strain on the solder joints. Finally, fit the two LEDs. Note that these mount in mirror image fashion, with the longer anode lead of each towards S1 in the centre of the board. They should initially be soldered in vertically, with the bottom of each LED about 15mm above the board. After soldering, each pair of leads is bent forwards by 90° about 7.5mm up from the board, so that the LEDs can be pushed into matching front panel holes. Final assembly The front and rear panels for this project will be supplied prepunched, with screened lettering. These panels can now be fitted to the PC board hardware and the entire Parts List 1 PC board, code 02104011, 117 x 112mm 1 low-profile plastic instrument case, 140 x 110 x 35mm 1 miniature DPDT toggle switch, 90° PC-mounting (S1) 2 RCA sockets, 90° PC mounting 1 2.5mm DC connector, 90° PC-mounting (CON3) 2 M3 x 8mm machine screws, with M3 nuts 6 small self-tapping screws, 6mm long 1 47µH RF inductor (L1) Semiconductors 1 NE5534 op amp (IC1) 1 74HC4066 quad switch (IC2) 2 74HC00 quad NAND gates (IC3,6) 1 LM1881 video sync separator (IC4) 1 74HC14 hex Schmitt inverter (IC5) 1 LM555 timer (IC7) 1 7805 5V regulator (REG1) 1 7905 -5V regulator (REG2) 3 BC548 NPN transistors (Q1,Q2,Q4) 1 BC640 PNP transistor (Q3) 2 3mm red LEDs (LED1-LED2) 3 1N4001 or 1N4004 power diodes (D1-D3) 6 1N4148 diodes (D4-5, D7-10) 1 BAW62 fast switching diode (D6) Capacitors 1 1000µF 25VW PC electrolytic 2 220µF 25VW PC electrolytic 2 100µF 16VW PC electrolytic 3 2.2µF TAG tantalum 2 0.22µF MKT polyester 10 0.1µF monolithic ceramic 1 .012µF MKT polyester 1 .01µF MKT polyester 1 .0082µF MKT polyester 2 470pF ceramic 1 390pF ceramic 1 270pF ceramic 1 220pF ceramic 1 82pF NP0 ceramic 1 39pF NP0 ceramic Resistors (0.25W, 1%) 1 1MΩ 1 1kΩ 1 680kΩ 2 680Ω 3 100kΩ 3 470Ω 2 10kΩ 1 330Ω 4 2.2kΩ 2 75Ω 1 1.5kΩ 1 47Ω April 2001  37 assembly installed in the bottom half of the case. The panels slide into the mould­ed case slots, while the board is secured using 6mm long self-tapping screws which mate with matching plastic spigots in the base. A total of eight 3mm mounting holes are provided in the board pattern and you can fit screws to all eight if you wish. It’s a certainly a good idea to fit the four along the back, to anchor the board firmly so that it doesn’t move when plugs are fitted to or removed from the sockets. On the other hand, two mounting screws will be quite sufficient at the front. Your Dr Video should now be ready for checkout. Checkout time There’s no actual setting-up required for this project. However, it’s a good idea to check that the power supply circuits are working correctly before you fit the top cover and put it to work. First of all, try applying 12V DC to the power input from a battery or plugpack supply. The Power LED should glow fairly brightly and the Sharpening LED should also light when S1 is switched to the “Sharpen” position. If one or both LEDs don’t glow, remove the power immediately and investigate because you have a problem. If neither LED glows, your 12V DC source may be connected with reverse polarity so that diode D1 is preventing any current flow. Reversing the supply connections will fix this problem. If only one LED refuses to glow, the odds are that it’s fitted to the board the wrong way around. So check this possibil­ity first and correct the problem if necessary. If you need to, the +5V and -5V supply rails can be checked with a multimeter. Both rails should be within a few tens of millivolts of their nominal values. If the positive rail is fine but the negative rail isn’t, look for a fault in the circuitry around IC7 and REG2. You may have fitted one of the electrolytic capacitors or diodes D2 & D3 the wrong way around. Another possibility is a solder bridge that’s preventing IC7 from oscillating. If the LEDs glow as they should and the two 5V supply rails measure correctly, your Dr Video is probably working fine and is ready for business. As mentioned earlier, there are no setting-up adjustments, because in this project we’re relying on close tolerance resis­tors and parallel capacitor combinations to ensure that the only parts of the circuit that are “critical” function as they should. We’re confident that this should be the case with almost any combination of components. Problems & cures There are only two possible problems that we can envisage, neither of them very likely. One is that if the timing components attached to the input (pin 1) of IC5a (in the VBI one-shot) are all excessively high in value, you may see a few black lines at the extreme top of the picture – and then only with movies in “full screen” format, as opposed to widescreen/letterbox. If this happens, it can easily be fixed by replacing the .0082µF capaci­tor with one of lower value (say .0068µF). The other equally faint possibility is that if the same component tolerance problem should occur in the timing circuit for the “burst gate” one-shot (ie, at the input of IC5b), the output pulses from this one-shot might be extended enough so that switches IC2c and IC2d begin to damage the horizontal sync pulses – causing horizontal jitter or tearing. This is most unlikely to happen but if it should, the remedy would be to replace the 220pF capacitor with a smaller one (ie, 180pF). One final comment – if you want to change the amount of high frequency video boosting given by the “Sharpen” switch, or the actual peaking frequency, this is easy to do. The amount of boosting is set by the series resistor, so varying it up or down in value from 330Ω will reduce or increase the boosting respec­tively. Similarly, the peaking frequency is set by the series capacitor, which can be changed from the current 82pF if you wish. A smaller value will increase SC the frequency and vice-versa. Where To Buy The Kit The copyright on this project is owned by Jaycar who will have complete kits available shortly after publication. these kits will include pre-punched front and rear panels with screened lettering. DON’T MISS THE ’BUS! Do you feel left behind by the latest advances in com­puter hardware and software? Looking for an easy-to-read book that explains the technology. Don’t miss the bus: get the ’bus! Includes articles on troubleshooting your PC, installing and setting up computer networks, hard disk drive upgrades, clean installing Windows 98, CPU upgrades, a basic introduction to Linux plus much more. AVAILABLE FROM SILICON CHIP PUBLICATIONS, PO BOX 139, COLLAROY, NSW 2097. PRICE $12.50 Inc P&P (Aust. only – see order form for overseas rates). To order your copy, call (02) 9979 5644 9-5 Mon-Fri with your credit card details! 38  Silicon Chip www.siliconchip.com.au SILICON CHIP’S 132 Pages $ 95 * 9 ISBN 0 95852291 X 9780958522910 09 09 9 780958 522910 COMPUTER OMNIBUS INC LUD ES FEA TUR E LIN UX A collection of computer features from the pages of SILICON CHIP magazine NO AVA W Hints o Tips o Upgrades o Fixes ILAB LE IRENT Covers DOS, Windows 3.1, 95,D98, CT o FR SILIC ON OM just $ CHIP 125O INC RT P&P