HXN50120
Product Specifications
HXN50120
Hisiwell — Discrete & Power Semiconductors
Product Overview
The HXN50120 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
HXN50120: 80 W continuous from DC to 4 GHz on a 50 V rail
The HXN50120 is the largest of the three unmatched gallium nitride HEMTs that Hisiwell publishes on a DC to 4 GHz window and a 50 V rail. The saturation row reads 80 W and the mode column reads continuous wave. Like the 40 W and 50 W members beside it, the gain and efficiency columns are blank, so the catalogue gives a power figure and nothing else that a drive or heat budget could be closed on.
80 W is 49.03 dBm by calculation. Into 50 ohm that is 63.2 V RMS and 89.4 V peak, which is 1.79 times the 50 V rail. That is the highest peak to rail ratio of the three, and it is the number that decides the output network: the drain has to stand nearly 90 V of peak swing at the load, and the harmonic terminations have to be set so that the waveform does not push the drain voltage further still on the crest of the second harmonic.
One wording on the product page deserves attention. The 40 W and 50 W members say there is no guarantee of performance outside the stated window. This one says the performance is guaranteed for applications operating in the mentioned frequencies, and then repeats the no guarantee sentence. Read together, the two sentences say the same thing twice, but the first is a positive statement that the other two pages do not make, and a buyer should not read more into it than the factory intended.
The HXN50120 rows as the vendor publishes them
| Part class | Gallium nitride HEMT power transistor, unmatched |
| Frequency range, index row | DC to 4 GHz |
| Band ratio, calculated | 4 to 1 |
| Saturation output, index row | 80 W |
| Saturation in dBm, calculated | 49.03 dBm |
| Output per gigahertz of window, calculated | 20.0 W |
| Drain rail, index row | 50 V |
| Load voltage at 80 W into 50 ohm, calculated | 63.2 V RMS, 89.4 V peak |
| Peak swing against the rail, calculated | 1.79 times |
| Mode, index row | Continuous wave |
| Modes the product page claims | Continuous wave, pulsed and any modulation |
| Gain | Not published on the index row |
| Drain efficiency | Not published on the index row |
| Package and body | Not published on the product page |
| Thermal resistance, junction to case | Not published |
| Gate bias point and quiescent current | Not published |
| Load impedance at the output | Not published, no load pull table on the page |
| Applied application list | Wireless communication infrastructure, wideband amplifier, EMC testing, ISM |
| Data sheet link on the product page | A product data sheet button, target address resolves to the site root |
| Orderable addendum | None published |
| Packing quantity and moisture level | Not published |
What 1.79 times the rail means on the board
A drain peak that sits 79 per cent above the supply is not a fault. On a tuned class AB stage the drain voltage rises above the rail when the load network presents the right impedance at the fundamental and a short at the harmonics, and the crest factor of the waveform does the rest. What it does mean is that the drain to source rating has to be well clear of 90 V, that the decoupling on the drain has to hold through the whole pulse of current, and that the output capacitor has to survive the full RF voltage rather than the DC level.
The second consequence is on the network itself. An 80 W part needs a transformation from about 2 ohm at the die to 50 ohm at the connector, and that transformation is loaded by the parasitics of the package and of the board. At 3 GHz the same network that works at 1 GHz will be dominated by the lead inductance, so the practical ceiling of a single section match is lower than the window suggests, and many programmes end up with a two section network above 2 GHz.
The third is on the supply. At a nominal 55 per cent efficiency the part draws about 145 W from the rail, which is 2.9 A average at 50 V and 65 W of heat. A 2.9 A average on a modulated carrier becomes a peak current several times higher, so the feed has to be sized on the peak and not on the average, and the bulk capacitor has to hold the rail through the crest.
Where the HXN50120 sits against parts with published curves
The shelf around it splits into two groups. The unmatched group, which is this part and its two smaller brothers, publishes power only. The matched group publishes power, gain and efficiency, because a matched part is sold with its network already built and its numbers already measured at the band. The matched members include the HX04200F at DC to 4 GHz and 200 W with 20 dB and 70 per cent on a 48 V rail, and the narrower rows at 2.5 GHz and 3 GHz with 17 to 19 dB and 69 to 71 per cent.
That split is the real comparison. A buyer choosing between this part and a matched 200 W part is not choosing between 80 W and 200 W; the buyer is choosing between building a network and buying one. The unmatched route gives freedom of band and a lower part cost, at the price of a tuning pass and a load pull measurement. The matched route gives a published gain and efficiency immediately, at the price of a fixed band and a narrower choice of centre frequency.
Where the band is fixed and the schedule is short, the matched route usually wins. Where the band has to move, or where the same board has to cover two bands, the unmatched route wins, and the three DC to 4 GHz members are the natural shortlist. The trade is set out part by part on the RF power transistor cross reference page.
What an order for the HXN50120 has to state, and what it cannot yet
No orderable addendum is published for this part, and the product page carries no packing column, no marking and no moisture level. The purchase order has to state the part, the quantity, the band, the mode, the rail and the bias condition, and the factory has to confirm each in writing. On a part where the network is built around the die, the band field is not a formality: it decides which screening the part is taken from.
The datasheet question is live on this page. The product page presents a button labelled as a product data sheet, but the address behind it resolves to the site root rather than to a document. Until the factory supplies the sheet, the gain, the efficiency, the pin voltages, the thermal resistance and the load impedance are all unpublished, and a heat sink sized on an assumed 55 per cent is a guess with a tolerance of tens of degrees.
The schedule is likewise unpublished. Nothing on the page states a lead time, a minimum quantity or a packing unit, so the delivery date has to come from a factory reply. Quoting a date before that reply is the one thing that cannot be corrected later without a cost.
Questions buyers ask about the HXN50120
Can it replace the 50 W member directly? Not without a new network and a new thermal path. The 2.04 dB step is a 60 per cent increase in load voltage, from 70.7 V peak to 89.4 V peak, and the drain current rises with the output at the same efficiency. A board laid out around the 50 W part will run, but its margin at the top of the band and its junction temperature will both be worse than the numbers the 80 W part was chosen for.
Does the blank efficiency column mean the part is inefficient? No. It means the vendor has not printed the number on the index page. Across this shelf the unmatched members carry power only and the matched members carry the full set, which is a publishing habit rather than a performance statement. The efficiency still has to be obtained before the thermal path is signed off.
What package is it in? The page publishes no package, no flange and no pinout. For an 80 W die the body matters, because the thermal path runs through the flange and the source connection, and a ceramic versus a plastic body decides both the thermal resistance and the mounting torque. That information has to come with the data sheet.
How is it supplied? The route that holds for this shelf is a factory authorised channel with a batch trace, and the incoming check covers marking, body and a curve trace at the intended band. The RF component sourcing page sets out the trace, and the one stop RF supply page covers how the drivers and the passives around it are consolidated into one line.