We are designing a two-stage OPAMP using a 65nm process.
I set VDD = 1.2V, and the offset of the OPAMP output stage was set to 600mV, which is just the middle to get the maximum output swing.
When simulated, this gain is about 50dB, which is about 285.
Now let’s write down the questions.
1.In a situation where the maximum possible output swing width is only 1.2Vpp (since VDD=1.2V), if the gain is about 285, it means that if a signal over 1.2/285 = 4.21mV is input, it will start to saturate. Is it an OPAMP that can be used by inputting only a small signal of 4.21mV or less?
2. I know that the 741 OPAMP has a gain (= 100,000) of about 100dB. If so, is it possible to input only "VDD/10 million = very small waveform" to use the 741 OPAMP without saturation?
3. I know that the gains can be further increased by using the cascode format. If so, is it necessary to input and use only smaller signals when stacked in cascode format? Rather, I know that the range that can be output swing is further reduced. If so, I am wondering if it is necessary to input and use only very, very small signals.
4. Usually OPAMP
I know you use it this way.
How does the inside of this OPAMP triangle work?
-Pin and + pin must be applied with a certain amount of offset for the differential MOSFET pair to operate. If the + terminal is grounded, does the MOSFET not operate at all?
When solving the problem, I just solved -Rf/Rin in this way, but when I designed it myself, I was curious.
5. I think I am firmly mistaken, but if there is someone who can properly hold the concept, please tell me
Thank you.
In the 741 OPAMP you mentioned, the gain of 100dB means Open Loop Gain, that is, the maximum gain when no negative feedback is applied. When used without feedback, it is used as a comparator. In other words, if two inputs are compared and either side is a little higher, the output is used as an open loop when making the output high or low, and when used as an amplifier, it is used with an unconditional negative feedback.
If you configure it like the 4th circuit uploaded by you, the gain = Rf/Rin. Here, you select an appropriate resistance value and get the desired gain. For example, if you only want to amplify 10 times, you can do Rf = 10k, Rin = 1k. However, in order to do this
Open Loop Gain >= Closed Loop Gain is available only when the condition is satisfied.
And even if high gain is needed, only one OPAMP is used to increase the gain.
This is a graph for the gain and frequency band of 741 OPAMP. For example, if the gain is set to 100dB (100,000 times), the frequency that can be amplified is only possible up to 10Hz. If you want to amplify the audible frequency (20kHz) smoothly, the gain must be approximately 20~30 times or less to obtain a sufficient frequency band.
And most OPAMPs use positive power. Therefore, even if the + input is connected to GND, the bias is caused by -power. The internal circuit itself is completely different for OPAMPs that use single power.
When I look at the questions you have posted, I think I don't have any prior knowledge about the basic concept of OPAMP or the design method.