A question about charge pump

[desperatejobseeker]

Conventionally, positive voltages are generated by a positive charge pump using N-channel MOSFETs, and negative voltages are generated by a negative charge pump using P-channel MOSFETs. Why and what is good for that?

Thanks in advance!

 

[BradtheRad]

The charge pump concept is that a capacitor charges when current goes in one direction, then it adds its charge to the applied voltage as current goes in the other direction.

Typically the current is steered from one stage to the next by diodes or transistors or mosfets, with the aim of stepping up a volt level (or sometimes to step it down).

I have a Youtube video of an animation of a simple charge-pump voltage doubler. Link:

www.youtube.com/watch?v=3czj7J_FE_k

You'll be able to find my other videos of voltage multipliers. Practically all use the charge pump concept.

 

[desperatejobseeker]

I know how charge pump works. What I wondered is why NMOS are conventionally adopted in positive charge pumps, and PMOS in negative ones? Make any differences?

Thank you though!

 

[BradtheRad]

The place where a P-device is liable to be needed is when the device is close to the positive supply rail.

The reason is that to use an N-device, the gate is referenced to the source terminal (where electrons enter). The gate needs to be pulled a few volts higher than the source terminal. The simple way to arrange this is if the source terminal is grounded, or has a clear path to ground at that time.

To use a P-device, the gate is referenced to the terminal where electrons exit. The gate needs to be pulled more a few volts lower than that terminal. It's easier to arrange this if that terminal is at the positive supply rail, or has a clear path to V+ at that time.

The above may be a consideration in constructing a charge pump. As the capacitor charges and discharges, it can create volt levels that would throw off operation, except if you're careful as to which polarity device you install next to which capacitor.