# fully differential class AB OTA Design

#### mirror_pole

Joined Nov 14, 2020
8
Hello guys,

First of all sorry for my english,i hope u can understand most of it.

Im new here and have some questions concerning the design on an class AB OTA.
Im relativly new to this topic and so far only studied basic stuff like basic circuit theory, basic operational amplifier topologies (Miller OTA, folded cascode..),
large signal, small signal models, feedback theory and so on from my university classes.

At the moment im trying to design a fully class ab differential ota in cadence which provides very high slew rate using an input transconductor in class AB, which is not limited by a fixed tail current.
My problem is that i have trouble to find some good literature sources concerning this topic to understand circuit behavior on a deep level. Most standart literature describes class A opamps mostly. Class Ab descriptions are mostly limited to output stages (push pull).

Also im kinda lost when it comes to the question how to chose dimensions for my devices.
So far i only calculated stuff like small signal gain, input/output common mode range, stability and so on and recognized that choosing the dimensions requires a systematic approach and is some totally different stuff compared to calculation.

Therefore i would like to know if there is some good literature going deep into approaches how to start with the design.

Thanks guys

#### MisterBill2

Joined Jan 23, 2018
9,243
I am trying to understand exactly what a "class AB OTA" actually is. Class AB means that the output devices conduct for more than half a cycle, more than 180 degrees of phase angle, but much less than 360 degrees.
My interpretation of "OTA" is Operational Transconductance Amplifier, and that is a type of device that I have avoided using at all. And certainly class AB operation would be limited to output stages, primarily audio power or other power, push-pull operation, OR radio frequency single-ended stages with a resonant circuit output to complete the sine wave.

Thus the actual design you are seeking is not at all clear, at least to me. So to avoid pointing in the wrong direction we will need more details of what is required. Otherwise all you will be getting are guesses.

#### mirror_pole

Joined Nov 14, 2020
8

As you can see the input transconductor is not fixed by a tail current source, therefore providing high slew rate when applying high differential signals. So far i had the same understanding of this class AB description, that it is only limited to output stages. From several ieee papers this is what is called fully class ab transconductor (input stage is also in class ab).

Most of this papers are written on a pretty high level and i wonder if there is literature concerning this type of amplifiers.
I try my best to understand these papers but its pretty challenging.

#### MisterBill2

Joined Jan 23, 2018
9,243
What frequency range and what power levels would the amplifier be for? Modern semiconductors are fast enough for most applications at lower power levels. I am trying to imagine anything other than a radio frequency amplifier in battery powered equipment where class AB would be used for anything aside from an output stage.

#### mirror_pole

Joined Nov 14, 2020
8
Low voltage/power supply: Vdd around 1.2V and quiescent power of around 250 uW.
Gain-Bandwidth at around 240 Mhz
Small signal differential gain of roughly 50 db

Using 40nm technology.

I have to say that my task is to design this amplifier and i really dont know much about where it can be used sorry. All im looking for is some advice concerning good literature so i can try to study it step by step. I didnt do anything beside hand calculations of different circuits and basic analysis by deriving transfer functions so far..then i started to think about how to design an opamp and then i realized that i actually know nothing.

Last edited:

#### MisterBill2

Joined Jan 23, 2018
9,243
There is another forum in this same site that is much more likely to be able to provide more useful information than this one..