When you just can't get any better  
 
    Home

Hardware:


Video gets routed to an LM1881 Sync Separator

The COMPOSITE VIDEO SIGNAL is capacitively coupled to pin 2 of an LM1881 Video Sync Separator. This part extracts timing information including VERTICAL SYNC (used to count frames), ODD/EVEN (used to tell if you're on an ODD FRAME or an EVEN FRAME), and BURST (used to count lines in a frame and for DC restoration). Each of these three signals are reduced in amplitude to 3.3V and routed to pins on the SPARTAN-3 board.

Next comes the Video Amplifier with DC Restoration

BURST and a copy of the original COMPOSITE VIDEO SIGNAL are routed to a ZETEX ZXFV4089 video amplifier w/ DC restoration chip. Usually video signals are capacitive coupled. This lets the video signal float all over the place. Not good. By using this part and grounding the REF pin-3, the video on the output pin-7 is again referenced to ground. Black is then always zero volts.

Now an OPAMP that sums the video signal and a HIGHLIGHT signal

Video output from the ZETEX part is fed to one leg of a summing amplifier made from an ANALOG DEVICES AD811. The other summing input is an output from the SPARTAN-3 board. When a logic high is fed to this point, the video signal is shifted up. On a monitor, you'll see a lighter section representing the area we're HIGHLIGHTING. The pot is used to set the highlight level.

The Video Signal now takes two paths: Path 1 - VIDEO OUT

Just a simple buffer made from another ANALOG DEVICES AD811. The output is the VIDEO OUT of this project and is connected to a monitor.

Video Signal: path 2 - Processing for the SLICING CIRCUIT

We're going to have to know when an area on the screen is white and when it's not. This way we'll be able to sense the presence of a puck. This is done with a Video Slicer circuit. Since 'white' on a monitor is represented by a high voltage and 'black' is represented by a low voltage, we need a comparator set for some cut-off value. A '0' output of the comparator will mean the video signal at a certain time is less than 'white', while a value of '1' at the output of the comparator will represent 'white' on the monitor. We really only use one of the two slicers on the board right now.

But before we can feed this signal to a comparator, we need to remove the chroma (color) information. The next part of the circuit is a low pass filter made from yet another ANALOG DEVICES AD811. It's used to remove the chrominance information from the video signal. Next in line is a buffer. Looking back at the schematic, I'm not sure why I put this in the circuit. Remember, this was an iterative process and I wasn't trying to save parts. If I needed a function, I just added it to the circuit. It's working fine and I'm not taking it out.

Problem is, it's referenced to ground so some of the video signal goes below ground and some of the video signal goes above ground. That wouldn't normally be a problem but the next circuit is referenced from ground to +5V. So, the next OPAMP is used to shift the entire video signal up by one diode drop - about .6 Volts.

Ok, we've got a slightly offset black and white representation of our original video signal. Feed that signal into a couple of National Semiconductor LMV7219 Comparators. These puppies are fast - their propagation delay is 7nS. By setting a reference voltage on the negative input, pin-4, the output will swing high whenever the video signal on the positive input, pin-3, moves above the reference. A couple of resistor divider networks and we've got a 3.3 Volt signal to sense white that we can feed to the Spartan-3 board.