Hi everyone,
I'm working on a project that involves an analog video signal that needs to be switched on and off by a microcontroller. The signal will be coming from a camcorder, and going to a TV.

My question is: what the best way to switch the signal? I suspect that a relay would be the simplest, but I haven't got any 5v relays on hand (and don't feel like rewinding one, which I've done for previous projects). Where I live it takes at least a week for any ordered component to get here, so buying relays is not an attractive option.

After reading through a few wikipedia articles (http://en.wikipedia.org/wiki/Composite_video , http://en.wikipedia.org/wiki/Analog_television), it looks to me like a composite video signal is DC -- that is, the voltage on the signal wire never goes below the voltage on the shield. Is this the case? If so, could I switch the video signal using just an N-channel MOSFET? Would the MOSFET's on resistance interfere with the video signal, do you think?

Also, do you think it would work to run the video signal over a pair of wires in an ethernet-type cable, with RCA connectors at either end? Or would this hopelessly mess up the impedance situation?

If I don't hear otherwise, I'll probably try the MOSFET and see if it works.

Thanks for your help,
Adam

 

Many years ago I made a video switch by biasing a couple of diodes into conducting, once they were conducting the video signal would pass. But I forgot the details.
If you run the video over 8-10 foot you should use a shielded cable. Preferable copper center conductor and copper shield.

 

You probably could use a MOSFET in series with the signal if it is indeed DC coupled and doesn't go below 0V. Power MOSFETs have very low ON resistance which shouldn't significantly affect the signal. But many video signals are AC coupled and thus go equally around plus and minus with respect to common.

For RCA type connections you need a 75Ω coaxial cable. If you want to use a Ethernet twisted wire cable you will need a video balun on each end to convert the coaxial unbalanced signal to the balance twisted wire signal and back. Otherwise, as you say, the impedance will be hopelessly messed up.

 

The signal will be coming from a camcorder, and going to a TV.

The CVS is indeed unipolar, with the base set at 0 volts and important steps at +0.3 and +1 volt.

The zero volts needs to be rock solid, thus a gated video amp is a bit tricky.

The LM 1201 - 3 series of gated video amps migh help here.

http://www.ti.com/lit/ds/slfs048/slfs048.pdf

Or try the NTE7081.

I take it you realise the CVS does not contain audio signals, which are bipolar?

Would these also require gating?

 

Thanks to everyone for your ideas and suggestions. I tried an N-channel MOSFET in series with the shield wire, and it works quite well. The only defect is that when the video signal is turned off, there's a split second of a sort of distorted partial-frame that appears on the screen. I suspect that this has nothing to do with the method of switching, but is just a result of cutting off the signal without sending a frame of black (perhaps related in some way to interlacing of the signal frames). If anyone has a brilliant idea of how to get rid of this (without having to synthesize a blank video signal), let me know!

I have the video running over about 6' of ethernet cable, and it doesn't seem to be noticeably affected. It remains to be seen how it fares going over the 25' or so that will be required.

Thanks for your help,
Adam

 

To avoid the partial-frame you could try synchronizing the switch so it turns off with the vertical sync. That would obviously require a circuit to extract the vertical sync signal and some logic (perhaps a few gates and a FF) to control the switch from that.