I want to build a device that does the following:

- Input 50MHz PAL video or s-video signal
- Combine odd and even frames (color average)
- Output 50MHz PAL video or s-video signal

Sort of a simple deinterlacer. But I would use it with Commodore 64 to view flickering/interlaced images as static eye-friendly images on a 1084 monitor.

Is there a simple IC that would do this? I'm quite a beginner in electronics so I would be grateful for any help. Thanks!

 

It's quite a complex circuit. You'll have to figure out how to:

• Sync with the video signal - easy, use an LM1881. The LM1881 provides an odd/even field output which is very handy! 5V-12V supply range at a few mA and comes in a DIP8.
• Extract the Y, U, and V fields from the fields. Either store the Y, U and V, or convert it to R, G, B. The YUV -> RGB is relatively easy, using some resistors and opamps.
• Store each field. This will be tough but not impossible. It would seem digital is the only way to go. So for PAL which is nominally 720x576 you need at least 1.19MB of memory, at 8-bit colour. You will need to store two fields so you need 2.37MB in total.
• Digitise the signal. Digital to analog. RGB to YUV.
• Average the two fields. There are two ways to do this: you can use a DSP, or average the two signals with analog circuitry.
• Convert each field to video again, including timing.

Now you can see why commercial deinterlacers are so expensive!

It is important to note that there is probably a better way of doing this. I could imagine a circuit which simply shifted the picture up/down by one line every odd field, which would be a poor man's deinterlacer. It could be fairly easy to do with some logic chips and a LM1881.

 

Thanks!

What about this one: MC44011.

What good could I do with this one? It seems to have alot of features already.

 

That's just a video processor. So what, it provides you with R, G and B or Y, U and V. You still have to store the fields, average the fields, and output them.

Give me a few hours, I have an idea. I reckon it should be possible to deinterlace a video signal by shifting up the odd (or even) fields by one line. This would be done by blanking the sync for one line. But this might produce a shaky picture.

 

Here is my idea.

I am not even sure it will work properly. It might cause the picture to appear shaky, or it might not work properly at all. We shall have to see. I suspect some modifications will need to be made before it works.

Let me know what you think. The components in this design are readily available, unlike that chip which could be difficult to get hold of especially in Finland.

 

Thanks, I'll have to try it out. It doesn't seem too complicated. It takes some time for me to get hold of the components. It's worth a shot.

 

After thinking a few things over here is an improved schematic with some better choices for components and certain specifications added. Also, I decided the 75 ohm resistor could be tricky, so I substituted it with an 82 ohm resistor, the next highest value. It's not critical, but keep it above 50 ohms and below 100 ohms. The input filter capacitor has also been changed because I figured the low power consumption (25mA max., 10mA typ.) of the device negates a big capacitor. The filter capacitor doesn't significantly change video quality (it is only turning lines on and off), though if it is too small it will not work properly.

When the battery goes low, it is quite likely to stop fixing the video signal instead of interfering with it, but to prevent failure anyway I recommend an external wall wart of about 9 volts DC. Because of the diodes, if you have a supply less than (battery voltage+0.7V), you cannot have the battery installed, because the battery will override the lesser supply. It will also accept AC though it is not full wave rectified so it might struggle, the filter capacitor might not be enough.