MOSFET Selection procedure

Hello all,

I have a switched capacitor converter circuit (attached part of the circuit, it has other 4 legs like that) I want to select the right MOSFET for it. I did the following steps and want to confirm of this is a right way to select the MOSFET or should be other way!

Step 1) Blocking voltage > 5V, Cont. Id> 0.35A (My Output power is 50mW @ 1.5V so Iout=35mA and Iin can be 5*Iout because this is 5x DC DC converter. I multiples by 2 for safety margin, so Id=(35mA)5*2=0.35A)

Step 2) Switching Capabilities, time delay of the mosfet should be less than ts by at least two times. I am using 1.77MHz oscillator, so ts=0.56us. So, there should be no problem here because most of MOSFET delay time are less than 200ns.

Step 3) Gate driving! I'm driving the MOSFETs using couple of logic gates. The logic gates are capable of providing 50mA. So, because some of these gates are driving 5 MOSFETs at one time (worst case), I assumed max current can be provided to each mosfet is 10mA. Then, the Gate charge total limit of the mosfet (Qg) can be found, Qg<(10mA/fsw) and I got 5nC. which means I need to find MOSFET of 5nC Qg or less.

Step 4) Conduction Power loss, I limit myself to 90% efficiency so the losses should be 5.5mW or less. I derived the losses formula and found Rds(on) of mosfet needs to be less than 6mOhms.

I checked DigiKey, Mouser and other online search but could not get Logic Gate Level MOSFET with Low Rds (<6mOhms) with Low Qg(<5nC) at the same time, there always a compromise! I think this is the best of MOSFETs that I found by NXP ( Rds=6.1mOhms, Qg=20nC!!):

https://www.mouser.com/ds/2/302/PSMN4R5-30YLC-843620.pdf


Would you please share with me your thoughts on my steps I've taken, do they make sense? Also, is the selection I made is considered OK?
and if you have other suggestions for these Low Rds, Qg, Logic gate driven MOSFET would you please share it with me?

I'm really sorry for the details I put, but I was trying to make it as clear as possible!

Hi

If you need low power then you should lower your switching frequency:
5nC * 5 Mosfets x 1.77MHz x 5V = 221mW
--> the switching loss will be far more than the conduction loss.

To reach 90% efficiency will be a hard way.

Klaus

you are right! I looked to other possible transistors and always there is a tradeoff between Qg and Ron. I found one with Qg=0.6nC and Ron=240mOhms. I could not get more than 70% efficiency (theoretically). I am not sure if this efficiency is expected in such low power level?

Step 3) Gate driving! I'm driving the MOSFETs using couple of logic gates.

Click to expand...

Wonder how you'll achieve it because pure NMOS switches require gate voltage above Vout, in other words an auxiliary power supply. Most charge pump converters are using CMOS switch topology to avoid it.