some questions aboout PTAT

[lhlbluesky]

in PTAT reference,
1\the opamp has two terminals, why does the terminal of pnp with resistor connect to the positive end of opamp? pls explain to me clearly.

2\the bjt can use pnp or npn, why does pnp normally be used? what's the advantage of pnp (in standard cmos technology)?

3\someone says the opamp should be single stage, but why? just because it is simple? or any other reason?

4\in order to improve psrr, a capacitor can be used between vdd (gnd) and the gate of pmos(that is, output of opamp)?how to choose the value of this capacitor?

pls help me. i need the answers urgently. thanks all.

 

[ytliang]

1. It depends on the structure of your bandgap. With your structure, it is the case.
For the opamp to be stable, we want to ensure overall negative feedback. The negative feedback factor has to be greater than the positive feedback factor. We want the side with the resistor to be connected to the positive end of the opamp so the negative feedback factor (pmos gives another inversion) is greater.

2. in nwell process, a pnp transistor is formed with p+, nwell, and psub.

3. a single stage op-amp is easy to compensate for stability.

4. the capacitor shorts at high frequency so your VGS of the pmos would stay the same when VDD varries, that is why it improves PSRR at high frequency. The size of the capacitor would depend on your spec and applications.

 

[saro_k_82]

3. A single stage opamp is easy to make in the case of self-bias amplifier configuration. More stages will ask for more startup circuits and invite trouble. But you can avoid this by having a fully differential first stage with a tail resistor.

4. The capacitor improves every spec except the wake-up time (and of course the area).

 

[lhlbluesky]

any other advice?