HX0445
Product Specifications
HX0445
Hisiwell — Discrete & Power Semiconductors
Product Overview
The HX0445 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
HX0445: sixty two watts over DC to 4 GHz with seventeen decibels of gain
The HX0445 is an unmatched gallium nitride HEMT power transistor from Hisiwell for the DC to 4 GHz window on a 28 V rail. The family index prints 62 W of saturation output, 17.4 dB of gain and 76 per cent drain efficiency. The product page names the two bodies HX0445F for the flanged outline and HX0445P for the unflanged one, prints 62 W as the maximum saturated power and 76.2 per cent as the optimal drain efficiency, and states that the part runs in pulsed or continuous wave mode. The documents agree on this member, which is not true of the two larger parts on the same shelf.
62 W is 47.92 dBm. Seventeen point four decibels of gain puts the drive at 30.52 dBm, which is 1.13 W. That is the friendliest drive level in the whole shelf: a single packaged amplifier of about a watt can push this part to saturation, so the line up is two stages rather than three.
The band is DC to 4 GHz, which is the widest absolute window among the parts above 60 W in this family. As a fraction of centre it is 200 per cent at 2 GHz, so the octave language does not apply: this is a decade style part with a match that has to hold from the low megahertz to 4 GHz.
The parameter rows published for HX0445
| Part class | Unmatched gallium nitride HEMT power transistor |
| Frequency range | DC to 4 GHz |
| Saturation output | 62 W |
| Saturation output in dBm, calculated | 47.92 dBm |
| Power gain | 17.4 dB |
| Drive power at saturation, calculated | 30.52 dBm, 1.13 W |
| Optimal drain efficiency, product page | 76.2 per cent |
| Drain efficiency, family index | 76 per cent |
| DC input at saturation, calculated | 81.4 W |
| Drain current at 28 V, calculated | 2.91 A |
| Dissipation at saturation, calculated | 19.4 W |
| Drain rail | 28 V |
| Operating mode | Pulsed or continuous wave, as stated on the product page |
| Bodies | HX0445F flanged, HX0445P unflanged |
| Matching | Unmatched, external network on the customer board |
| Input and output impedance | Not published |
| Scattering parameters | Not published in text |
| Load voltage at 50 ohm, calculated | 55.7 V RMS, 78.7 V peak |
| Output per GHz of band, calculated | 15.5 W per GHz |
| Thermal resistance, junction to case | Not published |
| Gate bias point and quiescent current | Not published |
| Operating and storage temperature | Not published |
| Orderable addendum, marking and packing | Not published |
| Shelf neighbour above | HX04100F2, 126 W, 14.8 dB, DC to 4 GHz |
| Shelf neighbour below | HX0625, 34 W, 18.5 dB, DC to 6 GHz |
Two point nine amps and nineteen watts for sixty two out
At 76.2 per cent the DC input is 81.4 W and the drain current at 28 V is 2.91 A. The heat is 19.4 W, which is a spreader and a modest fin stack rather than the liquid cooled block that a 138 W part needs. This is the point in the shelf where a natural convection heatsink stops being optimistic but is still possible.
The supply is a 3.5 A converter with headroom, not a 6 A one. That changes the board: a smaller inductor, fewer output capacitors, and less switching residue to filter out of the gate line. It also changes the enclosure, because 19 W can be spread over a larger area with a thinner base.
The output match sees 78.7 V peak into 50 ohm. That is below the 100 V class of capacitor, so the output network can be built from parts that are easy to buy, which is one of the quiet advantages of this member over the larger ones.
The drive is 1.13 W. A packaged gallium arsenide or a small gallium nitride stage at 31 dBm covers it, and the driver can sit on the same board with its own 28 V feed, which is what makes a two stage line up realistic here.
Where HX0445 sits between the small and the large members
| HX04100F2 | 126 W | 14.8 dB | 80.6 per cent | 3.08 dB above, band to 4 GHz |
| HX03120 | 138 W | 15.3 dB | 81 per cent | 3.48 dB above, band to 3 GHz |
| HX0445 | 62 W | 17.4 dB | 76.2 per cent | Reference point |
| HX0625 | 34 W | 18.5 dB | 80.2 per cent | 2.61 dB below, band to 6 GHz |
| HX0610 | 16.6 W | 21 dB | 80.2 per cent | 5.72 dB below, band to 6 GHz |
Going up to the 126 W member buys 3.08 dB, which is a doubling of power, and costs 2.6 dB of gain, 4.4 points of efficiency and 2.7 times the supply current. Going down to the 34 W member costs 2.61 dB and buys 2.7 points of efficiency and a two gigahertz wider band.
Two point six decibels is a factor of 1.8 in power and a factor of 1.34 in range on a free space path, so a link that is on the edge at 34 W closes at 62 W and does not close at 16.6 W. That is the practical reason this member exists in the middle of a ladder that already covers both ends.
The RF power transistor cross reference lists the ladder with the band and the gain, and the RF component sourcing page covers the driver, the isolator and the passives that go with it.
Board notes for a two stage line up on one rail
The driver and the final share a 28 V bus in the usual arrangement. The final draws 2.91 A and the driver a fraction of an amp, so the feed is sized by the final but the sequencing is shared: the negative gate on both stages has to be present and clamped before either drain rises.
The gate bias is unpublished, so the quiescent current is set on the first article. On a continuous wave part of this size a few hundred milliamps of quiescent current is normal practice, but the number comes from the measurement and the factory confirmation rather than from this page.
The input match has to present the source impedance the die wants across four gigahertz, which is the hardest item on this board: a match that is fine at 1 GHz can be a short at 4 GHz. Without published scattering parameters the network is measured, not calculated.
The output match works at 78.7 V peak and carries 62 W, so the capacitors and the combiners are chosen for voltage and current rather than for size. The isolator in front of the antenna protects the final from the mismatch that a broken connector produces.
Fifteen and a half watts per gigahertz is the density of this part across its four gigahertz, which is a third of what the 138 W member carries over three. That is the arithmetic behind the easier match: the same output spread over more band is a lower impedance transformation ratio at each frequency, and a lower ratio is a network that can be built with fewer sections and less loss.
Questions buyers ask about HX0445
Is the drive really one watt? At 17.4 dB of gain and 62 W of saturation, the calculated drive is 1.13 W. That is a comfortable level for a packaged driver, and it is the main reason this part is chosen over the larger members when the line up has to be short.
Does it run pulsed? The product page states pulsed or continuous wave. No duty limit is printed, so a pulsed application needs the factory to confirm the duty and the pulse width rather than assuming the continuous wave figure holds.
Which body do we order? HX0445F flanged for a bolted spreader, HX0445P unflanged for a solder down carrier. The electrical row is the same.
Why is the gain higher than the bigger parts? Because the die is smaller and the match is easier to hold across the band. Gain falls as the periphery rises, which is why the 138 W member prints 15.3 dB and this one 17.4 dB.
How do we start? Send the band, the output at the connector, the duty, the ambient and the rail. We will confirm the member, quote from the factory schedule, and ship a sample with test data at 1, 2 and 4 GHz.