AMAXNEW Descrambler (Elvis The Great) version 1.0b+ ------------------- This schematics is a tested version of the Amaxnew Descrambler v. 1.0b+ The descrambler is for sync suppressed systems WITHOUT INVERSION. The descrambler takes into account the PDK pulses (missing horisontal syncs) which also the maxinew/maxipic does. When watching unscrambled channels the control signal to the RF stage is minimized, so the channel is not affected. When switching to scrambled channel, the analogue stage sets the control signal to the RF stage to such level, which descrambles it. I have built the additional stages on a veroboard for experimentation and connected it to the Maxinew descrambler. The asm code should also be found in this zip file. The Amaxnew descrambler is based on Automax 1.0, which was reversed engineered by ™rgen (Thanks for that !!! Truly great hack !) and the Maxinew Descrambler (Elvis The Great) Maxinew's design is based on Maxipic by Cableboss and Odin (That's were it all started !) Special thanks also to Hugobardu who has made great investigations into the cable TV world, and revealed the secrets of many strange designs. Thanks to Maxipic for the home page , and all the others contributing to the designs, improvements, testing etc. ! Bill of Material for AMAXNEW.PCB (Hope this is ok!) Comment Pattern Quantity Components ------------------------------------------------------------------------------ 100K AXIAL0.4 2 R4, R7 1K AXIAL0.4 2 R13, R17 22K AXIAL0.4 1 R2 27R AXIAL0.4 2 R14, R15 2M2 AXIAL0.4 1 R8 2M7 AXIAL0.4 1 R5 10K AXIAL0.4 2 R3, R10 4M7 AXIAL0.4 2 R9, R11 68K AXIAL0.4 1 R6 6M8 AXIAL0.4 1 R12 75R AXIAL0.4 1 R1 100NF C 9 C7, C11, C12, C15, C16, C100, C101, C102, C103 100PF C 1 C2 150pF C 1 C4 68PF C 1 C3 27pF C 1 C17 33 PF C 2 C5, C6, C19 330PF C 1 C10 10UF EK 2 C13, C14 47UF EK 1 C1 10K TRH1 1 RV2 25K TRH1 1 RV1 1N4148 D 6 D1, D2, D3, D4, D5, D6 4066 DIP14 1 IC4 4069 DIP14 1 IC2 8 MHZ XTAL1 1 XTAL MC78L06 TR 1 IC5 PIC16F84 DIP18 1 IC1 TL084 DIP14 1 IC3 Here's some info about the sampling stage The function of the Automax sampling stage (Örgen), implemented into Amaxnew (Elvis The Great). Please refer to the Amaxnew schematics. The capacitors C7, C11 are charged through the R5,R9 resistors. RA3 goes high for about 0.5 us, and samples the video signal at color burst in HBI (horisontal blanking interval) (color should actually be filtered away). This discharges the C7 to black level of the video signal. In a scrambled picture the black level in HBI is far above the original level. RC signal goes high a bit in on each video line. This means that the C11 cap. is discharged continously through D4 during the video line. The result is that the lowest level which occurs during the video line will be at the input to IC3b. IC3b and IC3a are voltage followers which outputs the same voltage from the caps C7 and C11, and these output signals are going to a differential integrator circuit, with a DC amplification added (R11,C12,IC3c,R12). The output signal of this stage is the DC correction signal to the RF stage. The DC correction is sampled out to the RF during video line. The R5-C7, and R9-C11, are slowly again charging, but the difference stays constant. This difference is the key thing here, as this determines the output signal to the RF stage. Another thing is the Resistor-caps (RC) or integrating circuits: The RC circuits (not the signal!) used in amaxnew makes changes go slow, so this means that even if the black level in a video line does not occur (due to the current picture is for instance white), this does not affect the correction signal instantly. Small variations, though, might occur, and in my case I note that a bright picture gives a bit brighter picture and vice versa. It is though possible to watch the picture without getting a high bloodpressure :-) Some thoughts on the autolevel thing: First some background info : Automax by ™rgen ---------------- This was the original design from which the Amaxnew evolved (together with Maxinew, which in turn was a mod from Maxipic by Cableboss and Odin). This device contains some 4538 monostables which only have the possibility to give certain control signals to the 4066 gates. Amaxnew ------- Amaxnew design uses the Automax autolevel adjust stage, but in the Amaxnew case we have a lot of more possibilities to make lots of other control signals to the 4066 stages, as well as sampling almost whatever we want in the video signal, because of the PIC processor. Duck's ideas: ------------- Duck (in a prev msg thread) shows that it is possible to make a simplier autolevel stage, using less components. I studied the ascii schematics, and would make some small mods, but the main idea is great ! (If I were from England I would probably yell "Marvellous"!) I hereby suggest a mod which reduces even further the components required. We still need 2 ICs to do the autolevel adjust, but as Duck says, we can use a trimpot for setting black level to the differential amp. which then should be more stable than the previous designs. It goes like this:
VCC | RA3 RA2 / | | \ 10K | | / | | video | |------| |\ | o--||------| 4066 |-------------|+\ | in | |------| | | \ |------| / | | >---|-| 4066 |---- RF correction out \ 10K Cs === | / | |------| / | |--|-/ | | ___ | |/ | | | | --- | | VCC | Ry | | |--\/\/--- | / | | \ 10K | | / | | | | | / black level set | || | 10K \ <------------------/\/\/---|--||------| pot. / Rx || | Cx / \ 10K / | _|_ Control signals in comparison to the video lines: |---HBI-----|------Video line ----------------| | RA3 samples black level of HBI |---------------------------------| RA2 opens gate to RF stage.It might be that there are some caps missing, for removing glitches etc. and I'm not sure about the values of the differential stage's Rx,Ry,Cx and the Cs but the main idea could work. The idea : Setting video dc level to approx half the VCC, sample back porch of horisontal blanking (Cs gets this voltage), compare the thing with the differential stage and produce the correction voltage. Sample out this to the RF when video line occurs. As you might notice I've removed the diodes in the sample and hold stages, as those were used in Automax for finding the lowest level of the video line, and here I only want to compare the HBI black level with the black level I get from the trimpot. (One of the diodes in Automax was for balancing that stage with the other as a diode makes the sample 0.7 V higher). Some things must be taken into account when designing the above stage, such as that 4066 has approx 1kohm impedance when active, and we must make sure that the cap, which gets the sample voltage, does get charged properly during the sampling (0.5 us), otherwise we'll have a low pass filter there. The input cap combined with the 2* 10K, might work as a high pass filter, reducing slow variations in the video, so this must also checked. etc.etc. Duck has tested some of the stages above and has come up with some great results. It might be possible to use a much simplier stage. Some thoughts on rf stages: I tried to implement the Maxinick RF stage, but I could not control it with the Automax stage so I changed back to the maxinew stage again. The Amaxnew 1.0 b+ has been modified from a version containing some design flaws, and have not been tested after that Please report any bugs to the Maxipic discussion board. The link to the board can be found at http://members.xoom.com/Maxipic TNX 2 all for their effort ! & Keep up the brainstorming ! regards, Elvis