HXN50245
Product Specifications
HXN50245
Hisiwell — Discrete & Power Semiconductors
Product Overview
The HXN50245 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
HXN50245: two hundred watts on 2.7 to 3.5 GHz with both ports matched
The HXN50245 is a gallium nitride HEMT from Hisiwell for the 2.7 to 3.5 GHz window, quoted at 200 W on a 28 V rail and stated to deliver more than 250 W at 32 V. The product prose says both the input and the output are matched, and that the device can be configured as a push pull pair or as a Doherty amplifier. The family index files it in the unmatched table with the mode column reading pulse.
Those two statements are in conflict in a way that matters to the board. A part with both ports matched is placed on the board and driven through a simple matching or combining structure. A part that is unmatched needs the network built from a load pull. The vendor text says matched twice, in the overview, and the table header says unmatched, so the page carries both and asks the factory to settle it.
Two hundred watts is 53.01 dBm and 250 W is 53.98 dBm, just under a decibel apart. No gain and no efficiency is published, so the drive and the heat cannot be derived from the vendor text. At 45 per cent drain efficiency 200 W draws 444.4 W of DC, which is 15.87 A at 28 V or 13.89 A at 32 V, and leaves 244.4 W of heat. At 55 per cent it draws 363.6 W, 12.99 A at 28 V or 11.36 A at 32 V, and leaves 163.6 W.
Those heat numbers are the reason this part is normally pulsed. At 200 W continuous with 160 to 240 W of dissipation, the sink is a liquid plate and the DC supply is a serious one. In a pulsed application the average dissipation falls with the duty cycle, and the same device becomes practical in an air cooled box.
The parameter rows behind the HXN50245
| Part class | Gallium nitride HEMT, described as matched at both ports and filed in the unmatched table |
| Frequency range | 2.7 to 3.5 GHz |
| Power at 28 V | 200 W |
| Power at 32 V | More than 250 W |
| Power in dBm, calculated | 53.01 dBm at 200 W and 53.98 dBm at 250 W |
| Operating mode, index row | Pulse |
| Operating mode, product prose | Linear and saturated, push pull or Doherty |
| Power gain | Not published |
| Drain efficiency | Not published |
| DC input at 45 per cent, calculated | 444.4 W, 15.87 A at 28 V, 13.89 A at 32 V |
| DC input at 55 per cent, calculated | 363.6 W, 12.99 A at 28 V, 11.36 A at 32 V |
| Dissipation at 45 per cent, calculated | 244.4 W |
| Dissipation at 55 per cent, calculated | 163.6 W |
| Supply rails | 28 V nominal, 32 V for the higher output |
| Load voltage at 50 ohm, calculated | 100.0 V RMS, 141.4 V peak at 200 W |
| Load voltage at 250 W, calculated | 111.8 V RMS, 158.1 V peak |
| Matching | Stated as matched at both ports in the prose, contrary to the table |
| Configurations named by the vendor | Push pull and Doherty |
| Band width, calculated | 800 MHz, which is 30 per cent of the centre frequency |
| Output per GHz of band, calculated | 250 W per GHz |
| Body and package | Not published |
| Thermal resistance, junction to case | Not published |
| Duty cycle and pulse width limits | Not published |
| Gate bias point and quiescent current | Not published |
| Orderable addendum, marking and packing | Not published |
Two hundred watts of heat if the duty cycle is one hundred per cent
The whole thermal question on this part is duty. At continuous wave with 45 per cent efficiency the device dissipates 244 W, which is a liquid cooled plate and a supply that delivers 444 W. At a 10 per cent duty the average dissipation falls to 24 W and the same device runs on a modest air cooled sink, which is why the index row marks the mode as pulse.
The prose does not give a duty cycle, a pulse width or a thermal impedance curve, so a pulsed application cannot be closed from the vendor text either. What can be said is that the peak junction temperature during the pulse is set by the transient thermal impedance, and without that curve the only honest route is a measurement at the intended pulse width and duty.
At 28 V and 200 W the DC current is 15.9 A at 45 per cent. That is a bus bar rather than a trace, a choke with a saturation rating above 20 A, and a bulk bank sized for the pulse current. At 32 V the current falls to 13.9 A and the output rises above 250 W, so the higher rail is the more efficient way to reach the top of the part.
The push pull and Doherty notes are useful because they explain the package and the application. A push pull pair needs a balun on both ports and doubles the impedance seen by the match. A Doherty configuration needs an input split with a phase offset and an output combiner with a load modulation network, and it is the standard route for a base station carrier at this frequency.
HXN50245 against the 2.7 to 4.2 GHz group
| HXN50245 | 2.7 to 3.5 GHz | 200 W | Not published | Not published | Reference point, pulse row |
| HXN50240 | 3 to 4.2 GHz | 200 W | Not published | Not published | Same power, band starts at 3 GHz, CW row |
| HX2040185F | 2 to 4 GHz | 185 W | 18 dB | 70 per cent | 0.34 dB below, publishes gain and efficiency |
| HX1020240P2 | 1 to 2 GHz | 280 W | 20 dB | 80 per cent | 1.46 dB above, lower band |
| HX446050F2 | 4.4 to 6 GHz | 50 W | 15 dB | 55 per cent | 6.02 dB below, upper band |
| HX2580258GX | 5.2 to 5.9 GHz | 25 W peak | 12 dB | 25 per cent | 9.03 dB below, upper band |
The HXN50240 is the same 200 W in the adjacent band, 3.0 to 4.2 GHz, and its index row reads continuous wave where this one reads pulse. The two parts overlap between 3.0 and 3.5 GHz, so a 3.2 GHz product has a choice, and the choice is the mode rather than the power.
Against the HX2040185F this part gives 0.34 dB more output in a narrower band and gives up a published 18 dB of gain and 70 per cent efficiency. Where the line up has to be sized from the datasheet rather than from a measurement, the part with the published rows is worth a third of a decibel.
Both routes are covered by the RF power transistor cross reference, which lists the band, the power and the mode side by side, and one stop RF supply explains how a push pull pair and its baluns are bought as one kit.
Board notes for a pulsed two hundred watt stage
The pulse is the first item. No duty cycle, no pulse width and no transient thermal impedance is published, so the pulse conditions have to be confirmed with the factory in writing. A part run at a longer pulse than it was characterised for fails at the die attach rather than at the junction, and the failure is not visible on a curve tracer.
The rail is the second item. At 32 V the output goes above 250 W but the gate clamp and the sequencer have to hold at that rail, and the bulk capacitors need the higher rating. A 28 V board that is later pushed to 32 V for more output needs the passives checked rather than just the supply turned up.
The match is the third item. If the part is matched at both ports, the board is a combining structure. If it is not, the board is a load pull and a network. The vendor text supports the first reading and the table header supports the second, so a written confirmation is the only way to order the right board.
The fourth item is the DC feed. Sixteen amps at 28 V through a choke that also has to present a high impedance at 3 GHz is a two section filter, and the resonance of that filter has to sit outside the band. At this current the choke is a wound part with a saturation rating and a self resonant frequency that both matter.
Questions buyers ask about HXN50245
Is it matched? The prose says both ports are matched and the table files it as unmatched. We quote it as matched on the vendor wording and require a written confirmation, because the answer changes the board.
Is it a pulse part or a continuous wave part? The index row says pulse and the prose says linear and saturated. The heat at continuous wave is 163 to 244 W, so a continuous wave application has to be justified by a thermal measurement rather than assumed.
What is the gain? Not published, so the drive cannot be derived. At this power level the drive is a watt or more, which is a power stage in its own right.
Can it be used in a Doherty configuration? The vendor names Doherty in the prose. A Doherty needs a split and a combiner with a load modulation network, and the peak and the carrier devices are biased differently, so the bias network is part of the board.
How do we start? Send the band, the rail, the pulse width and duty if it is pulsed, the configuration and the ambient. We will confirm the match condition and the mode with the factory, quote from the schedule and ship a sample measured at the band centre with the pulse conditions recorded.