OTA followed by buffer

[samiurrehman]

Dear all,

We know OTA followed by a load driving buffer stage is what constitutes an operational amplifier. Can anyone show me how to design a buffer stage stage such that it doe not insert the dominant pole and the frequency response of the OTA remains undisturbed and the load is also driven!!

 

[FvM]

The question doesn't sound well considered. What do you expect for the "undisturbed frequency response" of an OTA followed by a buffer? Infinite bandwidth, infinite gain, or even both? An unloaded real OTA will always have a pole related to it's ouput impedance. If you don't manage to make it the dominant one and aplly pole-splitting to the others in return, the OP will be most likely useless as an amplifier, because it doesn't allow stable feedback operation.

If OTA output would be exposed as a terminal, it's reasonable to omit a compensation capacitor.

 

[samiurrehman]

Thanks for reply
Ok let me be a little more clear. It is a fact that OTA has practically very less output load driving capability and we need an output buffer stage to drive our load. I am confused with this output load stage which must drive a load of atleast 5pF. I want my unity gain frequency to be at 2GHz with my OTA driving the required load at 1GHz where I shall have a gain of about 25dB. You can construct the image in your mind.
My OTA is ready but it can not drive the capacitive load, as soon as i append some load at its output the unity gain frequency falls back to paltry 100-500 MHz from 2GHz. So I need a load driver between my OTA and load which must drive the load but not alter the unity gain frequency. How can I do that???

 

[leo_o2]

It should be hard to add a buffer without introducing a pole. However, a zero might be generated to compensation one of high frequency pole.