VBA3303
Product Specifications
VBA3303
VBsemi — Discrete & Power Semiconductors
Product Overview
The VBA3303 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
VBA3303: two 30 V N channels at 2.6 milliohm for the low side
The VBA3303 is a dual N channel MOSFET in the SO-8 body, printed at 30 V drain to source and plus or minus 20 V on the gate, with 100 per cent gate resistance and unclamped inductive switching testing. Each leg is 2.6 milliohm at a 10 V gate with 15 A flowing and 3.2 milliohm at 4.5 V with 10 A flowing, and the typical gate charge is 26.5 nC at the 4.5 V point or 62 nC at 10 V. Continuous drain current is 25 A with the case at 25 C, 21 A at a 70 C case, 18.5 A at a 25 C ambient and 15.1 A at 70 C, with a 70 A pulsed rating. The vendor lists DC to DC conversion, low side switch, notebook PC and gaming, which is a fair description of a part that has the lowest resistance in this band.
This is the part to use when the resistance has to be small and the voltage does not. At 2.6 milliohm and 20 A the die loses 1.04 W per leg, and two legs of a synchronous stage lose 2.1 W, which is the reason a notebook converter running at 500 kHz can stay inside a thin chassis. The die is large for the body: 3545 pF of input capacitance, 650 pF of output capacitance and 240 pF of reverse transfer capacitance, with 62 nC of gate charge at a 10 V drive. That gate charge needs a real driver, and a resistor driven from a microcontroller pin is not enough above a few tens of kilohertz.
The ratings a buyer checks first
| Drain to source voltage | 30 V |
| Gate to source voltage | plus or minus 20 V |
| On resistance at 10 V gate, 15 A | 2.6 milliohm |
| On resistance at 4.5 V gate, 10 A | 3.2 milliohm |
| Continuous drain current | 25 A at 25 C case, 21 A at 70 C case |
| Continuous drain current in still air | 18.5 A at 25 C ambient, 15.1 A at 70 C ambient |
| Pulsed drain current | 70 A |
| Continuous source to drain diode current | 5.4 A at 25 C case, 2.7 A in still air |
| Single pulse avalanche energy | 80 mJ, avalanche current 40 A into 0.1 mH |
| Power dissipation | 6.0 W at 25 C case, 3.0 W at 25 C ambient |
| Junction to ambient, junction to foot | 33 and 16 C per watt typical, 42 and 21 maximum |
| Gate threshold voltage | 1.2 V minimum, 2.5 V maximum |
| Total gate charge at 10 V and 4.5 V | 62 nC and 26.5 nC typical |
| Input, output and reverse transfer capacitance | 3545, 650 and 240 pF |
| Body diode forward voltage at 3 A | 0.72 V typical, 1.1 V maximum |
| Junction and storage range | minus 55 to plus 150 C |
The gate drive is the whole story on a 2.6 milliohm die
Two switching sets are published. At a 4.5 V drive the turn on delay is 35 ns, rise 16 ns, turn off delay 48 ns and fall 16 ns; at a 10 V drive the same rows read 18 ns, 8 ns, 41 ns and 8 ns. The 16 ns rise at 4.5 V against 8 ns at 10 V is a doubling of the transition time, and on a die with 650 pF of output capacitance that is where the switching loss hides. At 20 A and 30 V, an 8 ns edge pair is about 2.4 microjoule of switching energy per cycle, and at 16 ns it is roughly double. At 500 kHz that difference is 1.2 W per leg, which is more than the conduction loss of the die at 20 A. The conclusion is simple: this part wants a 10 V gate drive, and a 4.5 V drive wastes the advantage the resistance buys.
The threshold window is 1.2 V to 2.5 V at 250 microamp, so a 4.5 V drive gives 2 V of overdrive on a worst case die and a 10 V drive gives 7.5 V. Gate resistance is 0.2 ohm minimum, 1.1 ohm typical and 2.2 ohm maximum, a low internal value that means the external resistor sets the edge and a 2.2 ohm to 4.7 ohm part is the usual choice for a converter. With 240 pF of reverse transfer capacitance the Miller plateau is visible at a 20 V drain, so the driver needs 2 A of peak current or better to keep the plateau short. A small driver in a SOT-23 body with a 1 A peak will work at 300 kHz and will show a rounded edge at 500 kHz, which is the point where the loss budget starts to suffer.
Thermal data supports the current rating better than on the smaller dies in this band. Junction to foot is 16 C per watt typical, so at 20 A the die loses 1.04 W and the junction sits 17 C above the tab; junction to ambient is 33 C per watt on a 10 second pulse and 42 C per watt maximum, so the same loss puts the junction 35 C above ambient in still air. That is why the vendor can print 18.5 A at a 25 C ambient: the dissipation at that current is 0.89 W and the rise is 30 C. In a 70 C box the ambient figure drops to 15.1 A, and a pair of legs sharing one lead frame has to count both losses before the first one is derated.
A dual 30 V part as the low side of a converter
A two phase buck for a notebook processor uses both legs as low side synchronous rectifiers and a separate part for the high side. The 2.6 milliohm figure is what makes a 20 A phase efficient, and the 80 mJ avalanche rating with a 40 A pulse covers the phase node ringing when the high side turns off into a cold inductor. The 650 pF output capacitance adds to the resonant energy at the switch node, so a snubber or a slower high side edge is often needed, and that is a decision made on the bench rather than from the table. The same two legs also work as a single 1.3 milliohm switch when paralleled, which is the pattern used for a battery disconnect on a gaming notebook where the load can reach 25 A in bursts.
Outside computing, the pair is used as a low side switch for a small motor or a heater on a 24 V rail, where the 30 V rating covers the inductive kick and the 3.2 milliohm at a 4.5 V drive is good enough because the current is under 5 A. A second use is a load switch on a 24 V rail feeding a DC to DC module: the 80 mJ avalanche rating absorbs the module input capacitance discharge when the switch opens, and the 70 A pulse covers the inrush. Where the product has a 28 V battery, the 30 V rating is too close to the top of the charge curve and a 40 V part is the answer, which is the reason the same family carries a 40 V member with a similar die.
Sourcing notes and the second source question
A 2.6 milliohm dual in SO-8 is close to the bottom of what the body can do, so the number of real candidates is small and the differences between them are in the gate charge and the output capacitance rather than in the resistance. That makes the bench more important than the table: a candidate with 2.8 milliohm and 40 nC of gate charge can be the better part on a 300 kHz converter, and one with 2.4 milliohm and 90 nC can be the worse. Our power management cross reference keeps the gate charge and the capacitance beside the resistance, and the audit route in supplier quality audit covers what we check at the source before a lot is booked, including the on resistance sample at both gate voltages and the unclamped inductive switching test the document claims.
For a build that mixes this part with a driver, an inductor and a handful of passives, the kitting service cuts the whole set from one date code, which is the practical way to keep a two phase converter consistent between first article and production. We hold the incoming record per lot and can quote the on resistance distribution rather than a single value, which matters when two legs have to share current.
Questions we answer before a reel is booked
Buyers ask whether 25 A is available on a normal board. It is a 25 C case figure, and the ambient derating gives 18.5 A at 25 C air with the junction at 150 C, so the real number for a two layer board in a 60 C box is nearer 12 A per leg. The second question is whether a 4.5 V drive is enough. It switches, with 3.2 milliohm, but the edge doubles and the switching loss at 500 kHz rises above the conduction loss, so the drive voltage is the first thing to fix before the routing of the gate loop.
The body diode question matters on this part because the 5.4 A continuous rating with a 0.72 V drop is a real path during dead time. At 10 A of dead time current the diode dissipates 7.2 W, well beyond the package, so the dead time has to be short or the diode has to be helped by the channel being on. A parallel Schottky is the usual answer in a 500 kHz converter, and the 33 ns reverse recovery at 10 A tells the reader that the body diode is not a fast one. Avalanche is generous at 80 mJ, but it is still one pulse into 0.1 mH, and a repeated event needs a clamp.
On supply, we quote the reel and the date code and hold the unclamped inductive switching record per lot. For a customer that needs matched legs, we can supply from one date code so the threshold windows are close, which is what a two phase converter with a shared gate drive needs.