FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 

simple video clamp
MUFF WIGGLER Forum Index -> Video Synthesis  
Author simple video clamp
Hi, I got a few of the genlockable cameras Lars suggested in a previous thread and would like to mess around with them, despite not having a 'Color Video Interface' yet. If I can't directly drive the sync input in these cameras with the composite out of another camera, I've got a piece of gear that will give me black burst.

My question is, once I've got the cameras sync'd, will I need a clamp/DC restore circuit and how can I make one simply? Will a diode clamp be sufficient? (these are black & white cameras, by the way, and I don't know if they've got AC- or DC- coupled outputs) I'm guessing I'd need to have the diode connected to a post-AC coupled signal, and a -900mV reference, to get the sync at -300mV?

Or I could use an op-amp with the diode with a -300mV reference voltage as here:

I've got a gig in a few days, wanted to get this to work by then so I was hoping someone with some experience could chime in and maybe save me some trial & error.... I'm not looking for perfect right now, just quick & dirty.
Here's a simple DC-restoration amp that I used a lot in the 70s and 80s.

The best quality dc-restoration involves extracting the sync from the video and generating a timing pulse for the back porch portion of the video. Then using either a clamp or DC error amp to make the back porch sit at a reference level.

But you said simple, and doing all that isn't simple.

The above circuit is a sync tip clamp. It's not as accurate or stable as a back porch clamp, but it's a lot simpler to make. But the output video can move up and down a bit if the sync level fluctuates.

This circuit puts out video with the black at around -0.3 to -0.4 volts, and the whites at around +0.3 to +0.4 volts. To offset the signal to get black at zero volts would either involve a DC mixer after this to add some voltage, or to use the output of an op amp to drive the collector of Q3 instead of connecting it to ground. Set the op amp up with a trimpot so it's output is adjustable from about zero to 1 volt. With it at around +0.3 or +0.4 volts the output of the circuit should have black at around zero volts.

You may also want a video amp after this circuit to increase the signal a bit, if you want black to white to be 1 volt.
The circuit Dave posted and advice on biasing the signal/adjusting amplitude sounds good to me! That would work for a quick build. The DC-restoration circuit in our Video Sync Generator and Color Video Interface is a transistor buffer followed by a clamp circuit involving a CD4053 switch, and an LM1881 to extract the back porch signal (which then uses the CD4053 to clamp the buffered input signal to 0 Volts.) This is then buffered again with a video op-amp.

In the VSG's circuit, this is then followed with a static notch-pass filter centered around the subcarrier frequency for NTSC or PAL, a sync tip-clipper (half-wave rectifier that clips off all the sync information) and then a gain-setting amp to set the scale from 0V to 1V. There's also a trimmer to adjust for ~53.7mV offset going into the sync clipping for NTSC (since black level is slightly higher than clamp level for NTSC signals.)

All of that said, if you're wanting just a temporary solution, you're not going to harm anything by just getting an RCA to 1/8" adapter and plugging your camera video directly into your Video Blending Matrix or other LZX module. As long as the video is genlocked, it will be stable. You will get some fluctuation in brightness level without it being DC-restored, and possibly other weird performance issues, but it will still work.

The "Color Video Interface", (which may get renamed "Triple Video Input" -- any thoughts on that?) will be out immediately following this current production run we're pushing through, so not too much longer.
Could someone point me to the thread that "smrl" mentions the Genlockable cameras that Lars Suggested? I have been trying to do research on this same thing. much appreciated!
Here are some relevant threads:
Thanks for the input, guys. I got one of these working... Now, I was wondering about op-amp bandwidth. I just put a TL072 at the collector of Q3 thinking that bandwidth wouldn't be too much of an issue for a current source. But then I'm not 100% certain because HF variations in the current there would end up modulating the output. I think at least. (I don't have any formal electronics training so bear with me)

For certain a video amp after the circuit would demand something better. The only through-hole IC I have on hand that works at that supply voltage is a HA5024IP - which is current-feedback. I haven't really put any video through a current-feedback op-amp before and reading a little bit I'm concerned about driving cable capacitances and stray capacitance in my shitty stripboard design. I also see one can limit the bandwidth of CFA's with the feedback resistance network, which might make my life easier?

One question I have for Lars - looking at your Sandin schematics I notice no 75 ohm terminating resistor on the input. Is this typical for your modules? I suppose I should put a 75 ohm series resistor at the output of my circuit either way, for compatibility, but it looks like typically you would drive an output with a gain of 2, with the 75 ohm resistors splitting the voltage. Does this make sense? i wouldn't be surprised if there's a difference between your video input schematics and your module inputs, but I wonder about this from a design perspective. Is the 75 ohm impedance matching pointless if you're not driving 75 ohm coax? Does it do any good?

Here I am asking for quick and simple and of course now I want to figure out how to make it all work -WELL- smile
Hey smrl! Glad to hear you got something working. I will PM you some more schematics.

One question I have for Lars - looking at your Sandin schematics I notice no 75 ohm terminating resistor on the input. Is this typical for your modules?

Yes. Within the LZX module patch, the outputs are 500R impedance and the inputs are 100K impedance -- this is pretty much the same as normal EuroRack/Modular synth standards for audio and CV. All inputs are also buffered, usually by a video op-amp right at the input, before any attenuation.

There were a lot of design concerns that went into this decision. We found that the bandwidth of short cable runs buffered very well throughout the system, when dealing with RGB channels, does not make a noticeable difference in image quality (composite video with subcarrier would be more sensitive to signal degradation). Another factor is power consumption, 75R terminated I/O requires much more power -- and we wanted a system very heavy on the number of I/O points on each module without a ridiculous power load. Another factor is the creative ease of use that's afforded by being able to passively mult one output to many inputs simultaneously rather than requiring a video distribution amp to achieve this.

But of course, going into the system, and out of it (designated in the system by using RCA jacks rather than 1/8" jacks) will be normal 75R terminated inputs, and Gain X2 75R driven outputs. So for a DC restoration circuit for video from a camera entering the system, you will want to have the 75R terminated input, but could drive the output through 500R at a 1V scale (rather than Gain X 2.)

Hope that makes sense!
A TL072 should work fine at the collector of the dc restore amp. The currents are very low there so I don't expect there to be any issues.

A current feedback video amp should be fine for further amplification. They are quite sensitive to high frequency oscillations, so best to build on breadboard with direct connections rather than on a prototyping strip board.

It would be good to put a termination resistor on the output of the video amp chip. 150 ohm to ground from the output should be fine. I'd also add a 150 series resistor to limit current when the output is shorted to ground. You don't need to do the common 75 series, 75 to ground of the data sheets since, if going to an LZX module you won't be going to a terminated input.

If you use this same circuit at some point in the future with my upcoming video modules, you will need to use the 75 series resistor on the output since my modules do all have 75 ohm terminations.But with mine you will also need more gain since my system uses larger video signals to help decrease signal noise.
Hello,This is an old topic, but it applies to something I'm working on.

I've built a 2-input video keyer (well 3, for the control voltage), and it is working so far.... but of course, I now have to figure out a way to properly make valid sync, since I'm not working with video sources that have a genlock. Thanks to Linear Technology's AN57 whitepaper, there is a schematic (and I have the parts) to make a circuit that will take the composite sync (from an LM1881 in my case) and sum that into the video output of the keyer. That circuit includes a DC restore.....

My question is about whether or not to in addition DC restore the incoming two video inputs first too, before feeding them into the keyer circuit. It seems to me that would be a good idea to prevent the incoming sync from messing with the sync that I'm inserting later in the The keyer requires 1 volt input for each input, but it never specifies that this can't be the -0.3 to +0.7 of composite video, and I've seen that it is working. The reason for doing the DC restore would be simply to ensure that I'm getting a solid sync signal without interference later in the circuit.

Is that correct? Or am I overthinking it?
MUFF WIGGLER Forum Index -> Video Synthesis  
Page 1 of 1
Powered by phpBB © phpBB Group