VBPB165R20S
Product Specifications
VBPB165R20S
VBsemi — Discrete & Power Semiconductors
Product Overview
The VBPB165R20S from VBsemi is a discrete power semiconductor delivering efficient switching and power conversion. VBsemi specializes in discrete MOSFETs, power transistors, and protection devices at competitive cost for power supplies, motor control, and consumer electronics.
Superb Automation partners with leading brands including Qorvo, TI, ADI, XILINX, NVIDIA, Intel, ST, MACOM, Renesas, and Coilcraft to provide complete BOM and supply-chain solutions.
Key Features
- Low Rds(on) for minimal conduction losses in power switching
- Fast switching speed with low gate charge for high-frequency operation
- Avalanche-rated for robust operation under inductive load conditions
- High junction temperature rating for demanding environments
- Logic-level gate drive compatible for simplified design
- RoHS and AEC-Q101 qualified options available
Key Specifications
| Vds / Vce | V |
| Id / Ic | A |
| Rds(on) / Vce(sat) | mΩ |
| Qg | nC |
| Technology | — |
| Package | — |
| Configuration | — |
| Mounting | — |
Applications
Request Quote — In Stock, 100% New & Original, Full Warranty
VBPB165R20S: a 650 V super junction part in TO-3P with 0.161 ohm at 10 V
The VBPB165R20S is a 650 V N channel super junction MOSFET in the TO-3P body. The product summary prints two figures, 650 V from drain to source and 0.161 ohm typical at a 10 V gate, and the rest of the first page says what those two cost: a 20 A continuous rating at a 25 C case that falls to 12 A at 100 C, a 60 A pulsed limit, a 170 W dissipation row and a body diode with 6.4 microcoulomb of recovery charge.
The listed applications are the ones a 650 V switch of this size normally serves: server and telecom supplies, switch mode supplies, power factor correction, high intensity discharge and fluorescent ballast lighting, welding, induction heating, motor drives, battery chargers and solar inverters. That list is wide because this is a general purpose high voltage switch rather than a part tuned for one topology, and the figures on page two hold for all of them.
The 20 A rating is worth reading twice. It is a case temperature figure, and at a 100 C case the sheet already drops it to 12 A. Twenty amps through 0.161 ohm sheds 64.4 W, which the 0.65 C per watt junction to case path turns into 42 C of rise, so the die is comfortable on a heatsink and impossible on a bare board. The ambient rows are the ones that decide a fanless build, and the sheet keeps them on the same page rather than hiding them in a curve.
The gate is where the super junction character shows. Total gate charge is 78 nC typical and 122 nC maximum, a spread of 56 per cent on the same part number, and the gate drain charge of 20 nC against a gate source charge of 16 nC means the Miller plateau is a real interval rather than a spike. A driver sized on the typical figure will be slow on the maximum part, and at 65 kHz the difference is visible in the efficiency number.
The parameter table in full
| Drain to source voltage at 25 C | 650 V |
| Gate to source voltage, continuous | plus or minus 30 V |
| Continuous drain current, case at 25 C | 20 A |
| Continuous drain current, case at 100 C | 12 A |
| Pulsed drain current | 60 A |
| Single pulse avalanche energy, 30 mH and 8 A | 390 mJ |
| Maximum power dissipation, case at 25 C | 170 W |
| Linear derating factor | 1.7 W per C |
| Junction to ambient, maximum | 62 C per watt |
| Junction to case, maximum | 0.65 C per watt |
| Drain to source voltage slope at 125 C | 50 V per ns |
| Reverse diode voltage slope | 5.1 V per ns |
| Gate threshold at 250 microamp | 2.0 V minimum, 4.0 V maximum |
| Drain to source voltage temperature coefficient | 1.08 V per C |
| Gate leakage at 20 V and at 30 V | 100 nA and 1 microamp |
| Drain leakage at 480 V, 25 C and 125 C | 1 and 10 microamp |
| On resistance at 10 V gate and 10 A | 0.161 ohm typical |
| Forward transconductance at 6 A | 8.7 S typical |
| Input, output and reverse transfer capacitance | 2500, 81 and 9 pF |
| Effective output capacitance, energy related | 58 pF |
| Effective output capacitance, time related | 296 pF |
| Total gate charge at 10 V, 480 V and 5 A | 78 nC typical, 122 nC maximum |
| Gate source and gate drain charge | 16 and 20 nC |
| Turn on delay and rise time at 480 V, 5 A, 9.1 ohm | 24 and 25 ns typical |
| Turn off delay and fall time, same conditions | 28 and 25 ns typical |
| Gate input resistance | 0.3, 0.7 and 1.4 ohm |
| Continuous source drain diode current | 20 A |
| Pulsed diode forward current | 60 A |
| Body diode forward drop at 5 A | 1.2 V maximum |
| Body diode recovery time | 25 ns typical |
| Body diode recovery charge | 6.4 microcoulomb typical, 12.8 maximum |
| Peak reverse recovery current | 27 A typical |
| Operating junction and storage range | minus 55 to plus 150 C |
| Peak soldering temperature for 10 s at 1.6 mm from case | 260 C |
What the thermal rows allow in a 60 C cabinet
The 170 W row is the thermal path written twice. 170 W through 0.65 C per watt is 110.5 C of rise, which lands on a 135.5 C junction from a 25 C case, and the 192 W that a 150 C junction would allow is the true ceiling. The row is therefore a little conservative, which is the correct direction, and the 1.7 W per C derating factor is the same 170 W spread over the same 125 C.
In still air the ambient path takes over and the numbers collapse to something much smaller. 62 C per watt allows 2.02 W for a 25 C room, which through 0.161 ohm is 3.54 A; in a 60 C cabinet the allowance is 1.45 W and 3.0 A. Anyone who reads the 20 A headline and builds a fanless 15 A stage will find the junction at its limit within minutes, and the sheet has the row to show it before a board is cut.
The two output capacitance rows also matter more at 650 V than at 100 V. The energy related figure of 58 pF and the time related figure of 296 pF describe the same curve from two directions, and the factor of five between them is the reason a hard switched calculation and a resonant calculation of the same part give different loss numbers. Use 58 pF for turn on energy and 296 pF when the question is how long the drain takes to move.
Where the printed rows in this sheet disagree
The avalanche row is the safest kind of disagreement. Energy is printed as 390 mJ against a stated test of 30 mH and 8 A, and half of 30 mH times 64 is 960 mJ stored. The part is credited with less than half of what the test delivers, so the row is conservative, and a flyback clamp sized on 390 mJ will have margin rather than a surprise.
The body diode rows go the other way. Recovery time is 25 ns and recovery charge is 6.4 microcoulomb typical, and charge divided by time implies 256 A of peak reverse current on a 20 A device, while the peak reverse current row prints 27 A. Working the other direction, 27 A for 25 ns is 0.34 microcoulomb, which is nineteen times smaller than the charge row. One of the two rows does not describe this die, and our reading is that the charge row is the real one, because a 25 ns recovery belongs to a fast diode rather than to the body diode of a super junction switch with a 1.2 V forward drop.
The threshold window is the third item to note. Two volts minimum and four volts maximum is a two to one spread, and at the cold end the upper part needs a gate drive that is comfortably above 4 V before any current flows. A 10 V drive leaves 6 V of overdrive, which is fine. A 5 V logic gate on the same part is marginal at cold and is the sort of substitution that works on the bench and fails in the field.
Where a 650 V, 20 A part still fits
A 1 kW power factor correction stage at 65 kHz, a 500 W telecom front end, an HID ballast and a 3 kW induction heater all sit inside the envelope, because each of them either clamps the drain or keeps the current well below the rating. The part is a natural choice for the upper switch of a two switch forward, where the body diode never conducts and the slow reverse recovery costs nothing.
It is a poor choice for a hard switched bridge at 400 V with a high di/dt, not because of the switch itself but because of the 6.4 microcoulomb of recovery charge in the other leg. A snubber across each device or a phase shifted topology with a resonant transition is the way to use the part at that power level, and both cost board space that a smaller part would not need.
For the sourcing side, our power management cross reference work keeps the recovery charge, the output capacitance pair and the gate charge spread per approved alternate, because two 650 V parts with the same on resistance can differ by a factor of three in the charge that decides the switching loss. A cross reference that compares only voltage and resistance is not a cross reference at this rating.
Questions we answer before a reel is booked
The first is the gate drive. 122 nC at the maximum is 407 ns of charging time from a 0.3 A driver and 260 ns from the typical part, so the driver, not the MOSFET, sets the edge rate unless a 2 A driver is used. Add the 1.4 ohm maximum internal gate resistance and the external resistor has to be chosen against the worst case rather than the typical, or the turn off delay will exceed the 32 ns printed on the sheet.
The second is the avalanche margin. A flyback running in discontinuous mode with a 30 mH primary stores far less than 960 mJ, so the 390 mJ row is normally untouched, but a short circuit at the output can push the drain well above the clamp and the row is then the only thing standing between the part and a failure. We ask for the clamp voltage and the leakage inductance before we approve the part for that position.
The third is the alternate. A 650 V super junction part at 0.15 to 0.17 ohm comes from several suppliers with recovery charge between 0.3 and 7 microcoulomb and threshold windows from 2 V to 4.5 V, and the spread decides both the snubber and the driver. We hold those rows per approved alternate, the incoming inspection plan checks the arriving lot against them, and the pin to pin comparison starts from the TO-3P hole pattern, the tab thickness and the isolated mounting hole rather than from the current rating.