HXN40038
Product Specifications
HXN40038
Hisiwell — Discrete & Power Semiconductors
Product Overview
The HXN40038 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
HXN40038: 501 W on the same 2.7 to 2.9 GHz band
The HXN40038 is the largest member of the S band group in this family, and it doubles the output of the 54 dBm part next to it: at least 57 dBm of saturated power, at least 12 dB of gain and at least 48 per cent typical drain efficiency from a 50 V rail, over 2.7 to 2.9 GHz, in 100 microsecond pulses at 10 per cent duty, with plus or minus 0.8 dB of flatness and 50 ohm ports.
Fifty seven dBm is 501.2 W. At 48 per cent efficiency the DC input is 1044 W, the drain current at 50 V is 20.9 A, and the dissipation during the pulse is 543 W. That set of numbers is the whole story of this device: it is a 1 kW class DC load keyed in 100 microsecond bursts, and everything around it has to be built for 21 A of pulse current and half a kilowatt of instantaneous heat.
The efficiency row is the one that hurts compared with the 54 dBm member. Fifty per cent more output for two points less efficiency means the dissipation rises from 205.5 W to 543 W for the same 100 microsecond pulse and the same 10 per cent duty, a factor of 2.6 in heat for a factor of 2 in radio frequency power. That is the normal shape of a power ladder at the top, and it is why the largest device in a group is not automatically the cheapest way to reach a given output.
On a 10 per cent duty cycle the average dissipation is 54.3 W and the average drain current 2.09 A, which is still within a forced air budget on a reasonable carrier. The peak numbers remain the sizing driver, because the supply, the feed, the capacitor bank and the flange all see the peak.
The parameter rows Hisiwell publishes for HXN40038
| Part class | Gallium nitride power transistor, internally matched |
| Frequency range | 2.7 to 2.9 GHz |
| Saturation output power as published | at least 57 dBm |
| Saturation output power in watts, calculated | 501.2 W |
| Gain as published | at least 12 dB |
| Drain efficiency as published | at least 48 per cent, typical |
| Drain voltage | 50 V |
| Flatness across the band | plus or minus 0.8 dB |
| Operating mode | pulsed, 100 microsecond pulse width at 10 per cent duty |
| Input impedance | 50 ohm |
| Output impedance | 50 ohm |
| Output at 1 dB compression | not published, the index prints a slash |
| DC input at saturation, calculated | 1044.1 W |
| Drain current at saturation, calculated | 20.88 A at 50 V |
| Dissipation during the pulse, calculated | 543.0 W |
| Average dissipation at 10 per cent duty, calculated | 54.3 W |
| Average drain current at 10 per cent duty, calculated | 2.09 A |
| Drive power for the rated output, calculated | 45 dBm, 31.6 W |
| Charge drawn per pulse, calculated | 2.09 mC |
| Dissipation if efficiency falls to 45 per cent, calculated | 612.6 W |
| Thermal resistance, junction to case | not published |
| Transient thermal impedance | not published |
| Package outline and dimensions | not published in text |
| Scattering parameters | not published |
| Pinch off voltage and gate bias point | not published on this page |
| Operating and storage temperature | not published |
| Moisture level and floor life | not published |
| Orderable addendum and packing | not published |
| Same band, half the output, higher efficiency | HXN40037 at 54 dBm and 55 per cent |
| Same band, wider window, 48 V rail | HXN42002 over 2.7 to 3.1 GHz |
A kilowatt of DC in a hundred microsecond burst
The supply has to deliver 1044 W for 100 microseconds, which is 104 mJ of energy in each pulse at a 1 kHz repetition rate, or 104 W of average DC power. The charge drawn per pulse is 2.09 mC, so a rail allowed to sag 2 V needs 1.05 mF close to the device, and a rail allowed to sag 1 V needs 2.1 mF. Those capacitors sit inside the current loop with the feed and the flange, and at 21 A of edge current their ESR produces the droop that shows up as a leading edge power slump.
Twenty one amperes through a flange, a bond wire array and a printed feed is a packaging question as much as a circuit one. Nothing on the vendor page describes the flange, the pin arrangement or the current path, and at this level those details decide whether the part can be used at all: a flange with the wrong pin orientation will not fit the assembly, and a feed with too much inductance will limit the pulse edge before the device does.
The 543 W of instantaneous heat has the same problem in a shorter time. With no transient thermal impedance curve, the only defensible approach is a measured case temperature at the rated pulse condition on the first article, held against the 75 C case recommendation published in the family's one full data sheet, and a margin agreed with the customer before the order.
Why the largest device is not always the right one
Choosing between the 54 dBm and the 57 dBm member is not a question of output alone. The 501 W part needs 31.6 W of drive against 12.6 W, so it moves a packaged driver into a power stage. It dissipates 543 W against 205.5 W in the pulse, so the cooling and the feed change class. And it delivers 3 dB more, which for a radar in a fixed aperture is 3 dB of range, so the trade is real rather than academic.
The case where the larger part wins is a system that has already paid for the cooling and the driver, and needs the last 3 dB. The case where the smaller part wins is a compact assembly or a system with a limited DC budget: two 251 W parts do not sum to 501 W in a single chain, but a 251 W part with a good antenna can outperform a 501 W part with a compromised thermal path.
The case where neither wins is a linear application. Neither part publishes a compression point, so neither can be planned against a peak to average ratio from the data sheet, and a system that needs 501 W of average power needs a device several dB above it, which this family does not offer on this band.
Where a 501 W pulse stage at 2.7 to 2.9 GHz fits
The fit is a long range radar or a high power test amplifier where the pulse is short, the duty cycle is low and the cooling is engineered rather than inherited: S band surveillance radar, instrumentation radar for range and cross section work, satellite uplink ground equipment and laboratory load pull amplifiers. The 200 MHz band suits a system that owns the window, and the 50 ohm ports remove the external match.
It is a poor fit for a compact airborne unit without a liquid or heat pipe path, for any continuous wave use, and for any application where the driver cannot produce 31.6 W at 2.8 GHz with the stability that a 21 A pulse stage demands. It is also a poor fit where the supply cannot hold 50 V under a 1 kW transient, because the droop will show up as a power slump at the leading edge of every pulse.
The two S band members are listed together in our RF power transistor cross reference work with output, efficiency, rail and drive in one row each, so the step from one to the other can be priced against the cooling and the driver rather than against the datasheet headline.
Questions we answer about HXN40038 before a lot is booked
The first is the pulse condition. A hundred microseconds at 10 per cent duty is what the vendor publishes; a 1 ms key at the same duty cycle is a different thermal question at 543 W and needs a measurement, not an assumption.
The second is the flange. At 21 A of pulse current the flange, the mounting and the feed are part of the device, so ask for the outline drawing and the pin arrangement before the assembly is fixed. Our defence grade supply work covers the screening and temperature class questions that go with a radar build.
The third is the lot. With no moisture level, no marking code and no orderable addendum published, the purchase order names the band, the rail and the pulse condition, and the incoming inspection plan records the flange, the marking and the DC behaviour. The saturated power sweep at 2.8 GHz belongs on the customer bench or at a partner RF laboratory, and the RF procurement process is where the pulse width, the droop allowance and the cooling interface are agreed with the vendor before the first lot ships.