HXC50098
Product Specifications
HXC50098
Hisiwell — Discrete & Power Semiconductors
Product Overview
The HXC50098 from Hisiwell is a discrete power semiconductor delivering efficient switching and power conversion. Hisiwell is a recognized semiconductor supplier providing reliable components for global electronics manufacturing.
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
HXC50098: exactly 100 W, exactly 100 V peak, and why that rung is the calibration point
The HXC50098 is a 100 W single ended LDMOS RF power transistor on a 50 V rail, filed on the HF to 1.0 GHz line and marked CW. On the index it sits between an 80 W row and a 140 W row, which makes it the middle of three singles that differ by about a decibel and a half each. What makes it interesting is arithmetic rather than position: 100 W into 50 ohm is 50.00 dBm, 70.7 V rms and exactly 100.0 V peak, which is exactly twice the 50 V rail.
Those round numbers make this row the one to use when a bench has to be trusted. A 100 W load, a 100 V peak and a 2.00 ratio are easy to check with a scope, a directional coupler and a dummy load, and any error in the measurement chain shows up as a departure from a number that can be computed by hand. When a buyer is qualifying a fixture, this is the rung where a 0.5 dB error is visible instead of being lost in a table of measured values.
The row publishes no gain, no efficiency, no thermal resistance and no maximum ratings, and the vendor page carries no parameter block in the copy read this round. Everything below is the published row, the arithmetic that follows from it, and statements about what has to be measured before a build is released.
The HXC50098 rows as the vendor publishes them
| Part class | LDMOS RF power transistor, unmatched, single ended |
| Frequency range, index row | HF to 1.0 GHz |
| Output, index row | 100 W |
| Output in dBm, calculated | 50.00 dBm exactly |
| Step up from the 80 W row, calculated | 0.97 dB |
| Step down from the 140 W row, calculated | 1.46 dB |
| Drain rail, index row | 50 V |
| Mode, index row | CW |
| Load voltage at 100 W into 50 ohm, calculated | 70.7 V rms, 100.0 V peak |
| Peak against the 50 V rail, calculated | 2.00 times |
| Load voltage at a 2 to 1 mismatch, calculated | 200 V peak |
| Load voltage at a 3 to 1 mismatch, calculated | 300 V peak |
| DC input at 60 per cent efficiency, calculated | 167 W, 3.33 A at 50 V |
| Heat to remove at 60 per cent, calculated | 67 W |
| DC input at 55 per cent efficiency, calculated | 182 W, 3.64 A at 50 V |
| Heat to remove at 55 per cent, calculated | 82 W |
| DC input at 70 per cent efficiency, calculated | 143 W, 2.86 A at 50 V |
| Heat to remove at 70 per cent, calculated | 43 W |
| Gain | Not published |
| Efficiency | Not published |
| Thermal resistance, junction to case | Not published |
| Maximum ratings | Not published |
| Package | Not published on the row |
| Orderable addendum | None published |
| Moisture level and packing quantity | Not published |
What a 2.00 ratio to the rail means in practice
At full power the drain of a single ended device swings from near ground to twice the rail, and 100 W into 50 ohm happens to put the peak exactly at 100 V, which is exactly 2.00 times the 50 V supply. This is not a coincidence of this part, it is the geometry of a class B load line: for a device swinging to twice its rail the power available into a 50 ohm load works out at about 100 W for a 50 V supply. The ratio follows from the load line, and it is worth knowing because it tells a buyer what happens when the power changes.
Double the power and the peak rises by 1.41, so a 200 W part on the same rail peaks at 283 V, a 400 W part at 400 V. Halve the power and the peak falls by the same factor, so the 80 W row peaks at 89 V. That relationship, and not the power figure on the index, is what decides which parts on a 50 V shelf can share an output network topology and which cannot.
The mismatch case scales with it. At 2 to 1 this row sees 200 V peak and at 3 to 1 it sees 300 V. A 300 V capacitor is at its rating in the 3 to 1 case, which is a normal antenna fault condition, so the output network for a 100 W row is specified on 400 V parts rather than 250 V parts. This is the kind of decision that is cheap at the first article and expensive at the field return.
Thermal: 67 W of heat and a heat sink that fits in a hand
At 60 per cent efficiency the heat is 67 W. That is inside the range a 100 by 100 mm extruded sink with 25 mm fins can move by natural convection at a 40 degree rise, which is why this rung is popular for benchtop instruments and for equipment that has no fan. The calculation is not difficult, but it has to be done with the real ambient inside the box rather than the room temperature, because a sealed enclosure at 100 W of internal dissipation can run 25 degrees above the room.
The case to sink interface is the second term. A flanged ceramic package bolted to a sink with a thermal compound and a specified torque gives a fraction of a degree per watt; the same package with a dry interface and two screws gives several degrees per watt, which at 67 W is the difference between a junction at 160 and one at 200 degrees. Specify the interface material and the torque on the assembly drawing, and check both at the first article.
The third term is the one that is easy to miss at this power: the load resistor in the reject arm of a hybrid combiner. If two of these parts are combined, the reject load carries the difference between them, and a 0.5 dB amplitude error at 200 W combined puts about 10 W into a resistor that is often a 5 W part. Size the reject load on the worst case error, not on the typical one.
Gain and drive with nothing published
The planning figure for this class of unmatched 50 V LDMOS is 15 to 18 dB, which puts the drive for 100 W between 1.6 W and 3.2 W. On the same shelf, the 28 V rows of comparable power publish between 15 and 23 dB, and the 60 W pair on the 50 V rail publishes 23 dB. The spread is wide enough that a driver chosen on the midpoint can be 3 dB short at one end of the band, so the only safe route is to measure the assembled stage and set the driver specification from the measurement.
Gain flatness matters more than gain at this rung because the part is often used in a multiband instrument. A single network cut for the middle of a band gives more gain at the low end and less at the high end, typically 3 dB of tilt across a decade, and the driver has to have the inverse tilt or the amplifier needs a level correction loop. Measure the tilt on the first article, then decide whether the correction belongs in the driver or in the control loop of the instrument.
The input match is unmatched by definition, so the driver sees a load that varies from a few ohms to tens of ohms across the band. A 3 dB pad in front of the stage costs 3 dB of drive and buys stability that is otherwise difficult to guarantee. On a 100 W part with a 3 W driver, the pad is affordable; on a 400 W part with the same driver it is not, which is one more reason the ladder is not interchangeable between rungs.
How to use this rung as a calibration point
Because 100 W is exactly 50 dBm and exactly 100 V peak, this part is a good choice for the fixture that proves a measurement chain. Power meter, coupler, attenuator and scope can all be cross checked against a computed value rather than against each other, and a systematic error of 0.3 dB in an attenuator shows up as a 7 per cent error in the peak voltage reading, which is visible on a screen.
The same property makes it useful for incoming inspection of a batch. A 100 W part measured at 95 W is 0.22 dB low, which is inside normal part variation; the same part measured at 85 W is 0.71 dB low, which is outside it. Set the acceptance window on decibels rather than on watts, and the window will mean the same thing at every rung of the ladder.
The ladder itself, with the step between each rung and the current and heat at each power level, is on the RF power transistor cross reference page. The passives that go around this part, including the 400 V output capacitors the mismatch case demands, are on the RF component sourcing page, and the batch acceptance method described above is part of the incoming inspection plan page.
Questions buyers ask about the HXC50098
Why exactly 100 V peak at 100 W?
Because peak voltage into a resistive load follows the square root of twice the power times the resistance, which for 100 W and 50 ohm is the square root of 10000, or 100 V. That number is independent of the transistor and depends only on the power and the load, which is why the same calculation gives 89 V for the 80 W row and 118 V for the 140 W row.
Can it be used at 1 GHz?
The index says HF to 1.0 GHz, so 1 GHz is the top of the stated range and not a comfortable margin above it. Gain falls towards the top of the band, the input match changes, and the output network has to be cut for the sub band in use. Measure at 1 GHz before committing a build; do not assume the 100 W is available there with a network cut for 400 MHz.
How does it compare with the 140 W row?
The 140 W row is 1.46 dB above this one and needs output components rated for 355 V in the 3 to 1 mismatch case rather than 300 V. If the requirement is 90 W with margin, this row is the cheaper build; if it is 130 W, the 140 W row saves running a part at 93 per cent of its rating.