HX0465F2
Product Specifications
HX0465F2
Hisiwell — Discrete & Power Semiconductors
Product Overview
The HX0465F2 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
HX0465F2: sixty five watts on either a 28 V or a 48 V rail
The HX0465F2 is an unmatched gallium nitride HEMT power transistor from Hisiwell for the DC to 4 GHz window, and it is the only member of the 28 V shelf whose index row prints two rails. The family table gives 65 W of saturation output, 13.6 dB of gain and 71 per cent drain efficiency, and puts 28/48 in the voltage column. The product page calls it a 65 W part for applications up to 4 GHz, states 28 V or 48 V operation, prints 69 W as the maximum saturated power and 71 per cent as the best drain efficiency, and names the body HX0465F2, flanged.
Sixty five watts is 48.13 dBm and sixty nine watts is 48.39 dBm, a quarter of a decibel apart, so the two readings agree for practical purposes. Thirteen point six decibels of gain puts the drive at 34.53 dBm, 2.84 W, on the 65 W reading.
The dual rail is the feature worth the page. At 28 V the drain current for 65 W at 71 per cent is 3.27 A; at 48 V the same output takes 1.91 A. The heat is the same 26.5 W either way, so the 48 V option buys lower current and thinner copper, not a cooler die.
What the HX0465F2 documents print
| Part class | Unmatched gallium nitride HEMT power transistor |
| Frequency range | DC to 4 GHz |
| Saturation output, family index | 65 W |
| Saturation output in dBm, calculated | 48.13 dBm |
| Maximum saturated power, product page | 69 W, 48.39 dBm |
| Power gain | 13.6 dB |
| Drive power at 65 W, calculated | 34.53 dBm, 2.84 W |
| Drive power at 69 W, calculated | 34.79 dBm, 3.01 W |
| Drain efficiency | 71 per cent on both documents |
| DC input at 65 W, calculated | 91.5 W |
| Drain current at 28 V, calculated | 3.27 A |
| Drain current at 48 V, calculated | 1.91 A |
| Dissipation at 65 W, calculated | 26.5 W on either rail |
| Drain rail | 28 V or 48 V, as printed in the index voltage column |
| Operating mode | Continuous wave |
| Body | HX0465F2, flanged |
| Matching | Unmatched, external network on the customer board |
| Scattering parameters | Not published in text |
| Load voltage at 50 ohm, calculated | 57.0 V RMS, 80.6 V peak at 65 W |
| Load voltage at 69 W, calculated | 58.7 V RMS, 83.1 V peak |
| Output per GHz of band, calculated | 16.3 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, same band | HX0445, 62 W, 17.4 dB, 76.2 per cent, one rail |
Why the 48 V column changes the board more than the die
At 48 V the current for the same output falls from 3.27 A to 1.91 A. That is a 42 per cent reduction, and it moves the drain feed from a two ounce wide trace into a one ounce trace, takes the series resistance loss down by the same ratio, and lets a smaller choke carry the current without saturating.
The heat does not move. Twenty six point five watts leaves the die on either rail, because the efficiency is the same. A 48 V rail therefore does not solve the thermal problem, it solves the current problem, and a board that assumes the two are the same will be disappointed.
The 48 V rail also changes the regulator. From a 48 V bus a buck to 28 V wastes nothing much at 3.3 A, but a 48 V direct feed needs a converter that can hold 1.9 A with the same load step, and the drain decoupling has to be rated at 48 V rather than 63 V parts carried over from a 28 V board.
The gain is unchanged between rails, 13.6 dB in the index, so the drive stays at 2.84 W. A part that needed 2.84 W on 28 V and 2.84 W on 48 V is convenient, because the driver does not have to be reworked when the rail changes.
Seventy one per cent is the lowest efficiency on this shelf apart from the band limited members, and on a continuous wave carrier that is 26.5 W of heat where the HX0445 leaves 19.4 W for the same output. On a pulsed application the average heat falls with the duty, which is where the missing efficiency matters less and the dual rail matters more. The fifteen decibel gain figure is also the lowest of the four gigahertz group, so this part is chosen for the rail and not for the numbers.
HX0465F2 against the other 4 GHz members
| HX0445 | 62 W | 17.4 dB | 76.2 per cent | 0.21 dB below, 3.8 dB more gain, one rail |
| HX0465F2 | 65 W | 13.6 dB | 71 per cent | Reference point |
| HX04100F2 | 126 W | 14.8 dB | 80.6 per cent | 2.87 dB above, 1.2 dB more gain |
| HX0625 | 34 W | 18.5 dB | 80.2 per cent | 2.81 dB below, band to 6 GHz |
Against the HX0445 the power difference is 0.21 dB, which is nothing, and the HX0445 wins on gain by 3.8 dB and on efficiency by 5 points. So why buy this part? Because of the 48 V column, and because a 48 V rail is what the rest of the rack uses.
Against the 126 W member it gives up 2.87 dB and takes a 4.4 point efficiency advantage, which on a continuous wave carrier is 26.5 W of heat against 30.3 W for a quarter more output. The 126 W part is the better transmitter; this one is the easier board.
The RF power transistor cross reference carries the rail column for each member, and the one stop RF supply page covers how a 48 V line up is bought together with its converter.
Board notes for a dual rail part
The bias network has to be built for the rail it will see, and the gate voltage window moves with the drain voltage. Since the vendor publishes no bias point for either rail, the negative supply has to be adjustable over the range the first article needs, with a clamp that stops it from going positive.
The drain decoupling is different on the two rails: a 63 V part on 28 V is comfortable, and 48 V plus a standing wave is where a 63 V part stops being comfortable. The bulk capacitance and the choke are chosen for the rail that will actually be used.
The match is the same on both rails in principle, because the die is the same, but the load line presentation moves with the drain voltage, so a match measured at 28 V is not automatically the match at 48 V. That is a measurement on the first article, not an assumption.
The flange carries 26.5 W on either rail. The bolted joint, the flatness and the interface are the same job as on the rest of the shelf.
Questions buyers ask about HX0465F2
Is 48 V the same part or a different die? The vendor prints one row with 28/48 in the voltage column and one product page that states both rails, so it is one part number with two operating points, not two parts. The bias and the match still have to be set per rail.
Does the 48 V rail improve efficiency? No. The efficiency is 71 per cent on both, so the heat is identical and only the current falls.
Is 65 W or 69 W the number to quote? The index says 65 W and the product page says 69 W maximum. Both are carried here, and the factory confirmation decides which one a schedule is set against.
Is there a pulse rating? No. The mode is continuous wave and no duty is published.
How do we start? Tell us the rail, the band, the output at the connector, the duty and the ambient. We will confirm the power reading, quote from the factory schedule, and ship a sample with data at 1, 2 and 4 GHz.