HXN50253
Product Specifications
HXN50253
Hisiwell — Discrete & Power Semiconductors
Product Overview
The HXN50253 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
HXN50253: forty watts peak on 0.7 to 2.7 GHz for LTE infrastructure
The HXN50253 is an unmatched gallium nitride HEMT from Hisiwell aimed at the 700 to 2700 MHz window, which is the LTE, LTE-A and LTE-U band group. The product page opens with a sentence that calls it a 45 W part, then prints no parameter table at all. The family index prints 40 W in the peak power column, leaves the saturation column, the gain column and the efficiency column empty, and marks the operating mode as continuous wave.
Three things in that row are worth naming. The only power figure the index publishes is a peak figure, 40 W, on a row labelled continuous wave. The prose says 45 W. The gap between the two is 0.51 dB, which is small enough that a buyer would not notice it on a level plan and large enough that a purchase order written against 45 W would not be met by a 40 W lot. Both numbers are carried on this page rather than one being smoothed into the other.
40 W is 46.02 dBm and 45 W is 46.53 dBm. No gain is published, so the drive power cannot be derived from a vendor figure. For a part of this class and this band, 14 to 18 dB is the usual window, which would put the drive between 0.6 W and 2.5 W. That is a range, not a specification, and the page treats it as one.
No efficiency is published either. At 45 per cent drain efficiency the DC input is 88.9 W, the drain current is 3.17 A at 28 V and the heat is 48.9 W. At 50 per cent the same output takes 80 W, 2.86 A and leaves 40 W of heat. At 55 per cent it takes 72.7 W, 2.60 A and leaves 32.7 W. All three are calculated from an assumption, and the spread between the best and the worst case is 16 W of heat, which is the difference between a small extruded heatsink and a substantial one.
The parameter rows behind the HXN50253
| Part class | Unmatched gallium nitride HEMT power transistor |
| Frequency range, index row | 0.7 to 2.7 GHz |
| Frequency range, product prose | 700 to 2700 MHz, the same window |
| Power, index row | 40 W, published in the peak column |
| Power, product prose | 45 W |
| Power in dBm, calculated | 46.02 dBm at 40 W and 46.53 dBm at 45 W |
| Saturation output | Not published |
| Power gain | Not published |
| Drive power | Not published, cannot be derived |
| Drain efficiency | Not published |
| DC input at 45 per cent, calculated | 88.9 W, 3.17 A, 48.9 W of heat |
| DC input at 50 per cent, calculated | 80.0 W, 2.86 A, 40.0 W of heat |
| DC input at 55 per cent, calculated | 72.7 W, 2.60 A, 32.7 W of heat |
| Drain rail | 28 V |
| Operating mode, index row | Continuous wave |
| Load voltage at 50 ohm, calculated | 44.7 V RMS, 63.2 V peak at 40 W |
| Load voltage at 45 W, calculated | 47.4 V RMS, 67.1 V peak |
| Matching | Unmatched, external network on the customer board |
| Body and package | Not published |
| Thermal resistance, junction to case | Not published |
| Input and output impedance | Not published in text |
| Scattering parameters | Not published in text |
| Gate bias point and quiescent current | Not published |
| Orderable addendum, marking and packing | Not published |
| Data sheet link on the product page | Points to the vendor download directory, no document served |
Three amps on the drain and the assumption that decides the heatsink
Because the efficiency is not published, the thermal budget cannot be closed from the vendor text. The three calculations above bracket the answer between 32.7 W and 48.9 W of heat, and the only way to narrow that band is a measurement on the first article at the intended output and the intended duty.
Three amps at 28 V is a copper question rather than a connector question. At one ounce copper a 100 mil trace carries about three amps with a modest rise, so the drain feed is a 100 mil to 150 mil run, and the choke and the bulk capacitors sit on that geometry. Nothing about the DC side of this part is difficult. The difficult part is the match.
Seventy hundred to twenty seven hundred megahertz is two gigahertz of band, and 40 W across it needs a broadband match. The usual route on a 28 V GaN device is a tapered output transformer combined with a shunt feedback input network, built on a board with a defined reference plane and measured at both ends. Because no scattering parameters are published in text, that network is either supplied by the factory as a reference or built from a load pull, and the difference between those two routes is weeks of schedule.
At the output, 40 W into 50 ohm is 44.7 V RMS and 63.2 V peak. A standing wave of two to one raises the peak to about 89 V, so the output capacitors have to be rated above 100 V rather than the 50 V parts that fill a drawer. This is where a 40 W part costs more in passives than a 10 W part.
HXN50253 against the rest of the 0.7 to 2.7 GHz group
| HXN50253 | 0.7 to 2.7 GHz | 40 W peak | Not published | Not published | Reference point |
| HXN50254 | 0.7 to 2.7 GHz | 60 W peak | Not published | Not published | 1.76 dB above |
| HXN50256 | 0.7 to 2.7 GHz | 120 W peak | Not published | Not published | 4.77 dB above |
| HX1020120P2 | 1 to 2 GHz | 140 W | 20 dB | 80 per cent | 5.44 dB above, one gigahertz of band |
| HX2040185F | 2 to 4 GHz | 185 W | 18 dB | 70 per cent | 6.65 dB above, upper band only |
| HX2580258GX | 5.2 to 5.9 GHz | 25 W peak | 12 dB | 25 per cent | 2.04 dB below, different band |
The three members of the 0.7 to 2.7 GHz group form a clean ladder: 40 W, 60 W, 120 W, which is 1.76 dB and then 3.01 dB apart. All three carry the same blank parameter table, so the choice between them is made on power and price rather than on gain or efficiency, and none of the three can be compared with the wider shelf on those two rows.
Against the narrow band HX1020120P2 the trade is stark. That part publishes 20 dB of gain and 80 per cent efficiency on one gigahertz of band, and this part publishes neither figure on two gigahertz of band. A customer who needs 1 to 2 GHz only should not be buying this part; a customer who needs the whole LTE group has no alternative in the shelf.
For a line up that has to cover several bands, the RF power transistor cross reference puts the band and the power side by side, and the component lead time solutions page covers what happens when one member of the ladder is out of stock and a higher power member has to be backed off instead.
Board notes for an unmatched part with no published impedance
The first item is the reference board. Vendor pages in this family sometimes carry a data sheet link that resolves to a download directory rather than a document. On this part the link on the product page does exactly that, so the match has to be requested from the factory as a measured set or a reference board, and the request should name the band edges and the board material rather than the part number alone.
The second item is the bias. No gate voltage, no quiescent current and no bias sequence is published for this part. A 28 V GaN HEMT wants the negative gate present and clamped before the drain rises, and the drain removed before the gate falls. If the sequencer is not on the board, the first hot plug destroys the part and the failure looks like a bad batch rather than a missing circuit.
The third item is the output network. With 63 V peak at 40 W and 89 V peak under a two to one mismatch, the capacitors in the output match have to be rated for the mismatch case and not the matched case, and the series element has to carry the full load current. A 100 V part with a temperature derating of 50 per cent is the honest choice.
The fourth item is the heatsink interface. No junction to case resistance is published, so the case temperature at rated output in the final enclosure is the number that sizes the heatsink. That is a measurement, and an open bench reading taken at 25 degrees ambient is not a substitute for it.
Questions buyers ask about HXN50253
Is it 40 W or 45 W? The index prints 40 W in the peak column and the prose says 45 W. We quote 40 W as the figure with a published row and we ask the factory to confirm the 45 W sentence before a schedule is set against it.
What is the gain? Not published. The page gives no gain row, so the drive power has to come from a measurement or from the factory, and a line up sized on an assumed 15 dB may be one stage short at the top of the band.
Is the 40 W a peak or an average? The index puts it in the peak column on a row whose mode reads continuous wave, which is contradictory. For a continuous wave carrier the two are the same; for a modulated signal they are not, and the crest factor of the signal has to be added before the average is compared with 40 W.
Which band exactly? 700 to 2700 MHz, which is the LTE group from band 12 to band 41 in the common numbering, and the vendor states that performance outside that window is not guaranteed.
How do we start? Send the band edges, the output wanted at the connector, the signal type and its crest factor, the ambient and the rail. We will confirm the power figure with the factory, quote from the current schedule and ship a sample with the match and the gain measured at the two band edges.