The input signal is simply composite sync signal, without any additional video information. It is CMOS - 0 to 5V levels. I need the same type of signal to appear in the cable along with the RGB lines.

The difficulties I face relate to the best way to tranform the signal - in terms of level and impedance.

First, the 5V CMOS signal can be easily transformed into 300 mV-pp or 600 mV-pp signal using simple resistor network, not introducing much currents and affecting logic output of the previous chip:

enter image description here

Vref will simply be a positive supply rail, and I can, and most probably will, DC bias the resulting signal somwhere within 1-2 V (relatively far away from the power rails).

As for the operational amplifier, I would be very happy using ADA4891 device for the purpose.

And then I have difficulties to decide how to proceed, as I need to match output impedance of the operational amplifier with the cables' impedance (75 Ohms).

  1. Use G=+2 configuration. I will make input signal 300 mV-pp, and it will be converted to 600 mV-pp, and then divided back to 300 mV-pp by the 75-Ohm resistor network.

However my environment does assume single supply power. I read somewhere (e.g. here) that putting gain resistor onto the ground in this configuration is bad idea as it creates unnecessary currents from the output to ground (=to negative power rail of opamp), and capacitor must be used. Then it is stated that gain and capacitor will form a filter which will change the frequency response of the circuit. That sounds somehow complicated. I reviewed a number of sources and some of them signify these problems, but most just connect Rg onto the ground without further exploring the issues.

  1. Use G=+1. It will be as simple as voltage follower, with no relation of the feedback to the power rails. I will make input signal as 600 mV-pp, and just translate it to 600 mV-pp at the output with low impedance - and then divided to 300 mV-pp by the 75 Ohm resistor network. That would be so simple if datasheet would not constantly recommend using this operational amplifier in G=+2 configuration for the video, even in its headline, e.g. Video specifications (G = +2, RL = 150 Ω).

Can I use the op-amp in G=+1 configuration as explained in point 2 and what risks will it present? Probaly datasheet has technical information on it, but I am unable to derive it out of there. Analysing plots does not give me any warning signs of possible problems...


1 Answer 1


Reading the datasheet, the amplifier is unity gain stable, so can be used with G=+1, where it should give you a -3dB BW of 250MHz. You will want your data rate to be somewhat slower than this to get good pulse shape.

The data sheet specifications expressed in video specific terms (diff gain, diff phase), have only been measured in the G=+2 configuration, as that's the way that most people will configure it for video.

As you are (presumably) not interested in meeting a particular video spec, you can choose which configuration you like.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.