Post History
It's got nothing to do with the MOSFET's miller capacitance. Miller capacitance causes problems in common-source circuits but, your circuit is common-drain (or source follower) hence, with a steady...
Answer
#3: Post edited
by
Andy aka
·
2024-03-10T12:16:45Z (8 months ago)
- It's got nothing to do with the MOSFET's miller capacitance. Miller capacitance causes problems in common-source circuits but, your circuit is common-drain (or source follower) hence, with a steady DC voltage on the drain (your power supply), there can be no problematic feedback via the miller capacitor to the gate.
- Any issues with the stability (from to a step change in demand) are due to the 50 kΩ gate resistor and the gate-source capacitance. You might say "hey, it's a source follower so gate-source capacitance doesn't come into play" and, that would be a fairly valid point should the MOSFET source follower have near unity voltage gain (like a BJT). But, it doesn't so, about 50% of the gate-source capacitance can be modelled as sitting between gate and 0 volts.
This produce a decent phase lag that approaches 90° in the feedback loop and takes you close to instability. In fact, many of these types of circuit are so unstable that local feedback around the op-amp are needed to stabilize them.Then you should ask yourself, do you really need a very fast response in load current from a demand change and, if not, then put an RC filter between demand input and non-inverting input of the op-amp.
- It's got nothing to do with the MOSFET's miller capacitance. Miller capacitance causes problems in common-source circuits but, your circuit is common-drain (or source follower) hence, with a steady DC voltage on the drain (your power supply), there can be no problematic feedback via the miller capacitor to the gate.
- Any issues with the stability (from to a step change in demand) are due to the 50 kΩ gate resistor and the gate-source capacitance. You might say "hey, it's a source follower so gate-source capacitance doesn't come into play" and, that would be a fairly valid point should the MOSFET source follower have near unity voltage gain (like a BJT). But, it doesn't so, about 50% of the gate-source capacitance can be modelled as sitting between gate and 0 volts.
- This produces a sizable extra chunk of phase lag that approaches 90° in the feedback loop and takes you close to instability. In fact, many of these types of circuit are so unstable that local feedback around the op-amp are needed to stabilize them.
- Then you should ask yourself, do you really need a very fast response in load current from a demand change and, if not, then put an RC filter between demand input and non-inverting input of the op-amp.
- Regarding the variation in overshoot with supply voltage, this might be because the drain-source capacitance increases as drain-source voltage decreases. That capacitance can be regarded as being in parallel with the load hence, stability changes as supply voltage changes.
#2: Post edited
by
Andy aka
·
2024-03-10T10:01:29Z (8 months ago)
It's got nothing to do with the MOSFET's miller capacitance. Miller capacitance causes problems in common-source circuits but, your circuit is common-drain (or source follower) hence, with a steady DC voltage on the drain (your power supply), there can be no feedback via the miller capacitor.Any issues with the stability (from to a step change in demand) are due to the 50 kΩ gate resistor and the gate-source capacitance. You might say hey, it's a source follower so gate-source capacitance doesn't come into play and that would be a fairly valid point should the MOSFET have near unity voltage gain (like a BJT). But, it doesn't so, about 50% of the gate-source capacitance can be modelled as sitting between gate and 0 volts.- This produce a decent phase lag that approaches 90° in the feedback loop and takes you close to instability. In fact, many of these types of circuit are so unstable that local feedback around the op-amp are needed to stabilize them.
Then you should ask yourself, do you need a very fast response in load current form a demand change and, if not, then put an RC filter between demand input and non-inverting input of the op-amp.
- It's got nothing to do with the MOSFET's miller capacitance. Miller capacitance causes problems in common-source circuits but, your circuit is common-drain (or source follower) hence, with a steady DC voltage on the drain (your power supply), there can be no problematic feedback via the miller capacitor to the gate.
- Any issues with the stability (from to a step change in demand) are due to the 50 kΩ gate resistor and the gate-source capacitance. You might say "hey, it's a source follower so gate-source capacitance doesn't come into play" and, that would be a fairly valid point should the MOSFET source follower have near unity voltage gain (like a BJT). But, it doesn't so, about 50% of the gate-source capacitance can be modelled as sitting between gate and 0 volts.
- This produce a decent phase lag that approaches 90° in the feedback loop and takes you close to instability. In fact, many of these types of circuit are so unstable that local feedback around the op-amp are needed to stabilize them.
- Then you should ask yourself, do you really need a very fast response in load current from a demand change and, if not, then put an RC filter between demand input and non-inverting input of the op-amp.
#1: Initial revision
by
Andy aka
·
2024-03-10T09:58:58Z (8 months ago)
It's got nothing to do with the MOSFET's miller capacitance. Miller capacitance causes problems in common-source circuits but, your circuit is common-drain (or source follower) hence, with a steady DC voltage on the drain (your power supply), there can be no feedback via the miller capacitor. Any issues with the stability (from to a step change in demand) are due to the 50 kΩ gate resistor and the gate-source capacitance. You might say hey, it's a source follower so gate-source capacitance doesn't come into play and that would be a fairly valid point should the MOSFET have near unity voltage gain (like a BJT). But, it doesn't so, about 50% of the gate-source capacitance can be modelled as sitting between gate and 0 volts. This produce a decent phase lag that approaches 90° in the feedback loop and takes you close to instability. In fact, many of these types of circuit are so unstable that local feedback around the op-amp are needed to stabilize them. Then you should ask yourself, do you need a very fast response in load current form a demand change and, if not, then put an RC filter between demand input and non-inverting input of the op-amp.