HXN50172
Product Specifications
HXN50172
Hisiwell — Discrete & Power Semiconductors
Product Overview
The HXN50172 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
HXN50172: 15 W across three octaves and the gain that a 1 to 8 GHz window will not give away
The HXN50172 is a gallium nitride HEMT filed on the 50 V shelf with a 1 to 8 GHz window and a 15 W row. Three octaves in one package is the whole story of the part: a window that wide cannot be matched for flat gain with a fixed network, so the transistor is sold as the constant in a chain whose network changes with the band, and the buyer should read every power figure as a figure at one point inside the window.
The page is more generous with words than the row is with numbers. It calls the part a 20 W internally matched device for applications up to 8000 MHz and says that in a typical multi octave application across L, S and C band it delivers above 15 W pulsed CW from 1.0 to 8.0 GHz. The row says 15 W. The two figures are 1.25 dB apart, which is exactly the kind of gap that disappears inside a measurement tolerance and reappears in a contract.
Nothing on either document says what the match inside the package covers. On a part this wide the usual arrangement is an input match that holds the gain up at the top of the band, with the output left to the board, but the page does not say so and the shelf it is filed on is the unmatched table. Both facts have to be reconciled before a board is committed.
The HXN50172 rows as the vendor publishes them
| Part class | Gallium nitride HEMT, broadband |
| Frequency range, index row | 1 to 8 GHz |
| Window width, calculated | 7 GHz |
| Band ratio, calculated | 8 to 1, three octaves |
| Saturation output, index row | 15 W, 41.76 dBm |
| Output in the product page prose | 20 W, 43.01 dBm |
| Gap between the two figures, calculated | 1.25 dB |
| Band figure on the product page | Above 15 W from 1.0 to 8.0 GHz, L, S and C band |
| Output per gigahertz of window, calculated | 2.1 W |
| Output per megahertz of window, calculated | 0.002 W |
| Drain rail, index row | 50 V |
| Load voltage at 15 W into 50 ohm, calculated | 27.4 V rms, 38.7 V peak |
| Peak against the rail, calculated | 0.77 times, below the rail |
| Mode, index row | Full band |
| Mode, product page | Pulsed CW, wording used for the band figure |
| Matching, product page | Internally matched |
| Matching, index shelf | Listed on the unmatched GaN table |
| Gain | Not published |
| Gain flatness across the window | Not published |
| Power at 1 dB compression | Not published |
| Drain efficiency | Not published |
| Noise figure | Not published |
| Package, flange and body | Thermally enhanced industry standard package, no drawing published |
| Thermal resistance | Not published |
| Load impedance and load pull | Not published |
| Data sheet | No document linked from the page |
| Orderable addendum | None published |
| Packing quantity and moisture level | Not published |
| Applications listed | Wireless infrastructure, wideband amplifier, EMC testing, ISM |
Why an 8 to 1 window costs gain, and where the loss lands
A network that presents a good load at 1 GHz presents a poor one at 8 GHz, and no single fixed network covers both ends well. The vendor answers that with a part that is usable across the window and a network that is built per band, which is why a 15 W row can carry a 7 GHz range without contradicting itself. What the buyer inherits is a tuning task at every band edge, and a gain slope that has to be flattened outside the package.
The voltage picture is the one easy number in the whole block. 15 W into 50 ohm is 27.4 V rms and 38.7 V peak, which is 0.77 times a 50 V rail, so the output never swings above the supply and the passive parts in the match are not voltage stressed. Compare that with the 400 W members of the same shelf, where the peak is four times the rail, and it is clear that this part belongs to a different class of board: a driver or a small signal stage rather than a final amplifier.
The thermal side is mild for the same reason. At 15 W and an assumed 40 per cent efficiency the DC input is 37.5 W, the current at 50 V is 0.75 A and the heat is 22.5 W, which a small flange and a modest heatsink can move. At 20 W the same assumption gives 50 W of DC input and 30 W of heat. Neither figure is published, so both are planning numbers only, and the difference between them is one third of the heat load, which is why the 15 W against 20 W question matters even at this power level.
The full band mode word, and what it does not say
The index puts Full band in the column where the other rows put CW or Pulse. That is not a mode, it is a statement about the window, and it leaves two questions open: whether the 15 W is a continuous figure or a pulse figure, and what the duty was when the vendor took it. The page uses the phrase pulsed CW for the same claim, which suggests both words are being used loosely.
For a buyer this decides the thermal path. A 15 W continuous stage with 22.5 W of heat is a small amplifier with a small heatsink. A 15 W pulsed stage at 10 per cent duty is 2.25 W of average heat, which a board can absorb without any heatsink at all. The difference between those two outcomes is not a preference, it is a chassis.
The honest way to buy this part is to specify the mode on the order and ask the factory to confirm the condition behind the 15 W row. Where the application is EMC work or a lab amplifier that runs continuously for hours, that confirmation is the difference between a part that survives the afternoon and one that does not.
Where the HXN50172 sits on the 50 V shelf
There is no direct neighbour. The other broadband rows on the shelf are the 5 to 6 GHz family at 30 W and above, and the 0.7 to 1.4 GHz and 0.96 to 1.4 GHz pulse parts, all of which cover narrower windows at much higher power. On power per gigahertz of window the HXN50172 is the lowest density row in the table at 2.1 W per GHz, and that is the price of the octave range rather than a sign of a weak part.
The nearest useful comparison is a two stage route: a narrower driver at 1 to 4 GHz followed by a band specific final. That route costs more parts and more board area but buys 20 dB of output, which is the difference between a 15 W stage and a 1500 W stage. Where the system needs a single part that covers L, S and C band for test or for a broadband exciter, this row is the one on the shelf, and the comparison against the narrower rows is laid out on the RF power transistor cross reference page.
What an order for the HXN50172 has to state, and what it cannot yet
The part publishes no gain, no compression point, no efficiency, no noise figure, no package drawing and no thermal resistance, and carries no orderable addendum, packing column or moisture level. The purchase order therefore has to carry the part number, the band the amplifier will cover, the rail, the mode with a pulse width and duty if pulsed, the quantity, and the test the buyer will run. It should also ask the factory for the gain at the two band edges, because that is the figure the rest of the chain is sized on.
For a broadband part the incoming gate is more useful than usual: a sweep at 1, 2, 4 and 8 GHz on the customer fixture says more about a delivered batch than a single point check ever will, and it catches a die that is fine at the low end and short at the top. The incoming inspection plan page sets out how to size that sweep to a lot, and the RF component sourcing page covers the trace and the batch record behind it.
Questions buyers ask about the HXN50172
Is it 15 W or 20 W? The row says 15 W, the page prose says 20 W and then quotes above 15 W across the band. The 1.25 dB gap is small enough that the safe reading is 15 W for the full window and 20 W as a class figure.
Does it really cover 1 to 8 GHz in one board? The part covers the window, the board does not. Expect a network per band of interest and a gain slope to flatten, and expect the 1 GHz end and the 8 GHz end to need different work.
Is the output matched inside the package? The page says internally matched and the shelf says unmatched, the same unresolved pair as the rest of this table. Assume the output match is on the board until the factory says otherwise.
What is it used for? The page lists wireless infrastructure, wideband amplifiers, EMC testing and ISM. In practice a 15 W octave part is a driver or an exciter, and the lab and EMC uses fit it better than a base station final does. Where the chain also needs its passives and its connectors against one date, the one stop RF supply page covers the consolidated buy.