Anybody got information on line driving a c-sync signal?

[FusionITR]

Not a Composite video signal, a composite sync signal.

I'm having some trouble with driving this signal two places because I dont really have information on what the termination impedence is suppose to be for rgb monitors/input devices.

The composite signal is a TTL function at about 4.6Vpp when not connected to anything. I need to send a composite sync signal to an encoder and to a rgb monitor at the same time. When I hook up the composite sync signal to the encoder, there is no voltage drop to the signal, but when I hook up the signal to just the monitor, it drops to about 760mVpp.

What I tired to do is hook up the signal directly to the encoder, then from the same node send the signal to an opamp and line drive it then hook up the output to the rgb monitor, but the voltage only drops to about 3.6Vpp instead of the 760mVpp it drops to when hooking up the signal directly to the monitor. I tried putting some resistors between the output of the opamp and the rgb monitor but its too random, some values dont drop the voltage enough and some values drop the voltage too much, resulting in noise and other artifacts that arent suppose to be there.

Any suggestions on what I need to do?

 

[Nigel Goodwin]

The input impedance of monitors is usually 75 ohms, and video signals are usually 1V p-p maximum, from sync tip to peak white.

You don't mention though what either the monitor, or the encoder, actually needs?.

As you say the output is TTL, you could use a TTL buffer to feed the monitor, assuming the original signal is fine to feed the encoder?.

You might also just remove the 75 ohm load resistor in the monitor! - quite often these were switchable in monochrome monitors.

 

[FusionITR]

The input impedance of monitors is usually 75 ohms, and video signals are usually 1V p-p maximum, from sync tip to peak white.

You don't mention though what either the monitor, or the encoder, actually needs?.

As you say the output is TTL, you could use a TTL buffer to feed the monitor, assuming the original signal is fine to feed the encoder?.

You might also just remove the 75 ohm load resistor in the monitor! - quite often these were switchable in monochrome monitors.

I know the input impedence for the rgb signals in monitors is 75 ohms, this is also the case for c-sync signals? As for the monitor, I want this to work in the general case, not just for my monitor.

As for the input impedence being 75, now that I think about it, that makes sense because the maximum current my opamp can provide is 45mA, so with a 75 ohm load, and given the fact that at 15.63Khz (the TTL signal freq) the output resistance of the opamp is pratically zero, it doesnt have enough current to provide the 61mA needed, so the voltage drop is only around 3.4V.

However, since the load is 75 ohms, and the drop is only around 760mV when hooking up the signal directly up to the monitor, does it mean the c-sync signal has a ouput resistance of about 400 ohms or than it can only supply 10mA of current?

 

[FusionITR]

BTW, the encoder than i'm using is the AD725.
The signal is from an arcade board which outputs RGB + C-Sync at 15.75Khz horizontal scan frequency.
The opamp i'm using is an AD8054.

 

[Nigel Goodwin]

However, since the load is 75 ohms, and the drop is only around 760mV when hooking up the signal directly up to the monitor, does it mean the c-sync signal has a ouput resistance of about 400 ohms or than it can only supply 10mA of current?

Probably so, as you've now mentioned it's for an arcade board it explains why - it's not intended to feed an external monitor, so no need for 75 ohm impedances and the extra expense it incurs.

 

[FusionITR]

However, since the load is 75 ohms, and the drop is only around 760mV when hooking up the signal directly up to the monitor, does it mean the c-sync signal has a ouput resistance of about 400 ohms or than it can only supply 10mA of current?

Probably so, as you've now mentioned it's for an arcade board it explains why - it's not intended to feed an external monitor, so no need for 75 ohm impedances and the extra expense it incurs.

Actually, these signals are intended to hook up directly to arcade monitor, which I assume have the 75 ohm load impedences just like every other monitor. So since the C-Sync signal is at 4.6Vpp but only has the capability to supply 10mA of current, does this mean monitors expect around 760mVpp for the composite signal? I would assume if this is true, that its just easier internally to make a TTL function of 4.6Vpp instead of the intended 760mVpp and limit the current to 10mA to get the desired voltage at the load. If that is the case, does that mean I just have to put a resistor of about 385 ohms at the ouput of the opamp (when line driving the c-sync signal) to get the desired voltage a at the load (monitor)? If so, this seems alittle random as resistance values can vary and a small deviation would result in a noticibly noisey/distorted syncing and picture. Is there anyway I can make the signal a controled current source using a BJT or something?

Sorry about the long post.

 

[Nigel Goodwin]

Actually, these signals are intended to hook up directly to arcade monitor, which I assume have the 75 ohm load impedences just like every other monitor.

Quite possibly not!, because the need isn't there, the 75 ohm loading is intended to prevent 'ringing' down long cables, this doesn't apply with the monitor internal.

So since the C-Sync signal is at 4.6Vpp but only has the capability to supply 10mA of current, does this mean monitors expect around 760mVpp for the composite signal? I would assume if this is true, that its just easier internally to make a TTL function of 4.6Vpp instead of the intended 760mVpp and limit the current to 10mA to get the desired voltage at the load. If that is the case, does that mean I just have to put a resistor of about 385 ohms at the ouput of the opamp (when line driving the c-sync signal) to get the desired voltage a at the load (monitor)? If so, this seems alittle random as resistance values can vary and a small deviation would result in a noticibly noisey/distorted syncing and picture. Is there anyway I can make the signal a controled current source using a BJT or something?

I can only suggest the best thing is to scope the signals and see what happens - the 75 ohm output is usually simply fed through a 75 ohm resistor from a low impedance output amp, to give a 75 ohm output impedance. If you think this will result in a 50% signal drop, you are completely correct! - the output stages usually have a gain of 2 in order to compensate for the drop.

So adding your 385 ohm resistor would be perfectly acceptable, assuming that's what's needed.

 

[FusionITR]

Nigel, im simply puzzled by the operation of this c-sync signal, maybe you can help me out.

This is what the signal looks like without anything connected to it:

**broken link removed**


Here is what the signal looks like directly connected to the monitor:

**broken link removed**

It's already interesting to see that the top completely flattens out at the top when its connected to the monitor, but whatever, based on what I see here it seems like the c-sync out from the source can only provide about 10mA of current since thats why the voltage is so low.

However, when I did some testing to confirm this, I wasnt getting the results I should get. Based on this I should be able to hook up a 1k resistor as the load and see the full drop on it since the current would be less than 10mA, however the drop across it is 3.18V, meaning that it can now only supply 3mA of current? That doesnt make any sense, but moving on.

When I try to line dirve the signal through the op-amp, for some reason when I connect the signal to the opamp the original signal gets slighly distorted, not visible until I turn I change the sec/div to only see one square wave and make the volts/div fill the screen, but the distortion is there. Anyhow, ignoring the signal distortion, when I hook up around 325 ohms of resistance to the ouput to get 760mV to the monitor, the signal doesnt look exactly the same as it does as when the c-sync signal is directly connected to the monitor (the second picture), and results in the monitor not syncing properly. Heres what im talking about:

**broken link removed**

As you can see, the square wave type signal isnt getting to the monitor while trying to line drive it, which is why (i think) the monitor wont sync properly when trying to drive it this way.

Any suggestions?

 

[Nigel Goodwin]

A sync input should be very non-critical, the monitor would normally slice it to ensure it gets good sync.

Instead of an opamp, how about trying a TTL buffer, as it appears to be logic level?.