0

IRFP250N vs IRF3415

I'm making a boost converter which converts 12 V to 100 V maximum. The maximum input current is 15 A. Power of the converter is 120 W. I will drive the MOSFET with a TC4429.

At first I chose the IRFP250N because it has low parasitic capacitances and shorter turn on/off times, but it has a higher RDS(on) (0.075 Ω) compared to the IRF3415 (0.042 Ω.) This difference creates a big power dissipation difference at higher currents. Also, the IRF3415 will cause more ringing when switching.

I'm unsure of which MOSFET would give me better efficiency. I am planning to add a R = 6.8 Ω, C = 470 pF snubber across the MOSFET which may decrease efficiency even more. Do I really need a snubber?

What should I look for in a MOSFET? Which MOSFET is better for my application, the IRF3415 or the IRFP250N?

My calculations (f = 100 kHz, P = 120 W):

tdon = 12*10^-9;
tr = 55*10^-9;
tdoff = 71*10^-9;
tf = 69*10^-9;
Rds_on = 0.042;


Loss_conduction = Iimax^2 * Rds_on
Loss_turnon = P/6 * tr * f
Loss_turnoff = P/6 * tf * f
TotalLoss = Loss_conduction + Loss_turnon + Loss_turnoff

IRFP250N datasheet

IRF3415 datasheet

The circuit: enter image description here

ocrdu
  • 8,705
  • 21
  • 30
  • 42
CaveScientist
  • 375
  • 1
  • 10
  • "Do I really need snubber?" This will take into account the actual layout as well. If there is a bunch of ringing from d-s, then yes, it can quite that down. – Aaron Nov 14 '22 at 19:42
  • Well I don't know parasitic caps or inductances because the circuit will be hand printed and I can't measure those small values, I will play safe and put one, I can change later. It is not critical for now, this is a student project. But really want to learn how important the parasitic capacitance difference of MOSFETs really is. – CaveScientist Nov 14 '22 at 19:45
  • For a hand printed board, I would put the place holders in. If you don't need them, then the open traces shouldn't present a problem. – Aaron Nov 14 '22 at 19:48
  • To answer this we'd also have to see the circuit it was being used in as well as the driving frequencies. – Voltage Spike Nov 14 '22 at 20:14
  • Okay, I've added the circuit. What I realised is that if a MOSFETs voltage rating is high, it has bigger Rds(on) and if it has bigger current rating it has bigger capacitances. I've looked other MOSFETs and I noticed this general tendency. – CaveScientist Nov 14 '22 at 20:28

1 Answers1

0

Inefficiency can come from many places. Your Rds(on) is one source; multiply by the current to see how much power is being lost (mean square current, not average).

Another is switching currents. Switching a capacitive load C between V1 and V2 at frequency f results in a current of |V1-V2|fC. Multiply by the voltage difference again to express it in terms of power.

Now, you'll probably want to set up a spreadsheet so you can enter parameters of various devices and optimize the overall power loss.

Cristobol Polychronopolis
  • 9,622
  • 9
  • 17