Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Need circuit or IC for composite video from separate sync / video

Status
Not open for further replies.
Hi,

I am trying to repair a 1980's era home automation system for a friend. The CRT type monitor went bad and is unrepairable (major heat/flame damage to the circuit board and components). This system feeds separate sync and video signals to the original monitor. I need a circuit or an IC that can combine the H & V sync with video for an NTSC composite video output. That way we can use just about any security monitor or even a TV with composite video in to operate that system.

Just in case it's important, I have verified that a standard monitor with composite input can put the image on the screen, it is just very jagged because there is no sysc component in the signal.

Thanks for any help you can provide.

Mike
 
Last edited by a moderator:
**broken link removed** a Maxim circuit that adds a separate sync signal to the Green signal in an RGB video circuit. But it should work for combining the sync into your video signal also. Just make the green channel the video channel and leave the Red and Blue channels unused. You can use a MAX9586 for this, if you like, which has only one channel.

Place the chip on a board with a ground plane and decouple the circuit to the ground plane directly at the pins with short capacitor leads. Connect all components with short leads. A wall-wort (with a possible 317 regulator to generate 3.3V) can provide the power.

As a simpler (but less robust) alternative, you could try the summing resistor network at the input to the IC but run the output of the network directly to the monitor, with no IC. This will reduce the signal amplitude, however. Leave off the 75Ω resistor to ground if the signal level is too low. If that gives a satisfactory signal, then you don't need to add the IC.
 
Thanks... That looks like a promissing solution. I do have a question though. The chip appears to assume a composit sync signal, but this system has three inputs, H Sync, V sync, and video. Will this circuit work if I use a pair of diodes or resistors to combine the two sync pulses into a composit sync pulse?

Thanks again.
 
OK, I didn't realize there were separate sync signals also. It's a little messy to combine three signals at once. You may need to combine the two sync signals into one input of the Maxim circuit. Then take the combined sync output signal from the Maxim and combine that with the the video into a second Maxim input channel.

Do you have an oscilloscope to measure the video and sync output levels (into a 50 ohm load)?
 
Yes, I do have a scope. I am currently looking up the NTSC signal spec so I can make sure I get the pulses going in the right direction and correct amplitude.

Thanks.
 
If the sync signals are intended to be wired separately, combining them with simple logic blanks the H pulses during the vertical interval. Some NTSC monitors recover horizontal synchronization in the first few lines at the top, but others do not.

I've seen genlock created with Atmel AVR chips, but can't seem to find one at the moment.
 
Sorry, but that was over my head. This may end up being tougher than I thought. I guess I thought there would be a simple solution, but maybe it's a bit more difficult.
 
mneary was talking about combining the signals with digital logic gates. The Maxim method is an analog sum approach and should not suffer from the problem he mentions.

I asked about the signal amplitudes (and polarity) so the value of the summing resistors can be properly determined to generate the correct NTSC levels.
 
Carl, the signals are all approximately 1.1 v into 50 ohms. The vertical sync pulses are negative and the horiz pulses are positive. The vertical pulses are about 18 ms apart and the horiz pulses are about 62 usec apart.

Also, we are dealing with only B & W here so perhaps a 4 channel MAXIM chip could use 2 channels to combine the sync and then feed that + video into the other two channels to acheive the desired result?

Let me know if you need anything else.

Thanks,

Mike
 
Last edited by a moderator:
The horizontal positive sync pulses add some complexity, since both NTSC sync pulses must be negative. Thus you will need to invert the horizontal sync before combining it with the video. This can likely be done with a simple transistor inverter.

What are the width of the sync pulses?

Are they basically square-edge pulses that go from zero to 1.1V and back?
 
Carl, yes, they are basically square pulses of abut 7 usec wide. They have some overshoot on both the rising and falling edges, but 1.1 volts is correct for the square top.
 
Carl, I put pictures of the signals on a web page.

Goto **broken link removed**
You can delect which picture you wantto see and enlarge them. Hope that helps.
 
Well, this keeps getting more complicated. NTSC vertical sync is defined as:
  • pre-equalizing pulses (6 to start scanning odd lines, 5 to start scanning even lines)
  • long-sync pulses (5 pulses)
  • post-equalizing pulses (5 to start scanning odd lines, 4 to start scanning even lines)
Each pre- or post- equalizing pulse consists in half a scan line of black signal: 2µs at 0V, followed by 30µs at 0.3V.
Each long sync pulse consists in an equalizing pulse with timings inverted: 30µs at 0V, followed by 2µs at 0.3V.

This is obviously more than can be done with a simple sync summing circuit. I'm not sure how to do this without adding some rather complex circuitry. And I haven't found any commercial device that will do it.

But It may be that the monitor will sync without all those equalizing pulses. From your posted pictures I see that the vertical sync pulse is about 1ms which is roughly equal to the vertical blanking interval that normally contains all the vertical sync and equalizing pulses.

So the scheme would be to invert the horizontal pulse with a transistor inverter and sum that into one input of the 4 channel Maxim chip with the vertical sync. Then sum that with the video into the second channel.

So that may work, depending upon whether the monitor will sync properly on that long vertical sync pulse without the equalizing pulses. No guarantee.

It's your call as to whether it's worth the effort to try.
 
Thanks Carl... it is definately worth the effort to try. I'll get the chip on order as I already have the transistors, resistors etc that are needed. I'll let you know how this all turns out.

Thans for your effort to help with this.
 
Good luck. If you need any help with the circuit post a question.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top