Contact Us

HX339160

Brand: Hisiwell
Hisiwell HX339160 is a amplifiers & comparators component designed for Signal Conditioning & Filtering and related applications. In stock at Superb Automation — request a quote for competitive pricing and fast delivery.
quote now

Product Specifications

HX339160 — Amplifiers & Comparators | Superb Automation

HX339160

Hisiwell — Amplifiers & Comparators

Product Overview

The HX339160 from Hisiwell is a precision amplifier / comparator delivering high gain, wide bandwidth, and low distortion for signal conditioning. 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

  • High gain-bandwidth product for wideband signal amplification
  • Low input offset voltage and drift for precision DC applications
  • EMI-hardened design for robust operation in noisy environments
  • Rail-to-rail input/output for maximum dynamic range
  • Low quiescent current for power-sensitive designs
  • Unity-gain stable with capacitive load drive capability

Key Specifications

Channels—
GBWPMHz
Slew RateV/µs
VosµV
Iq per ChannelµA
Input Type—
Supply VoltageV
Operating Temp°C
Package—

Applications

Signal Conditioning & Filtering Current Sensing & Monitoring Data Acquisition Systems Audio & Video Processing Sensor Interface Amplifiers Industrial Control Loops

HX339160: 1 W at 12 to 18 GHz, and the one dB that changes the heat

The HX339160 is a GaAs MMIC driver amplifier on a 2.63 x 1.26 mm die, covering 12 to 18 GHz with 22 dB of gain and 1 W of output at the 1 dB compression point from a 5 to 6 V supply at 1 A. The vendor overview calls it a wideband 1 W part, the device page prints 30 dBm, and the family table prints 31 dBm for the same part number.

What the vendor publishes

ParameterValue
Frequency range12 to 18 GHz
Small signal gain22 dB
Output P-1dB30 dBm on the device page, 31 dBm in the family table
Output IP340 dBm
Supply5 to 6 V, 1 A
Input VSWR1.5 to 1
Die size2.63 x 1.26 x 0.1 mm

Which number is the part

30 dBm is 1.000 W and 31 dBm is 1.259 W, so the gap is 26 percent in power for one dB on paper. The overview says 1 W and the device page says 30 dBm, which makes 30 the guaranteed or minimum figure and 31 the typical. That reading is consistent with how the family table is built: it prints single numbers for every row with no typical or minimum columns, so its figures read as typical.

The heat moves the other way. The DC input is 6 V at 1 A, 6.0 W. At 30 dBm of output the heat is 5.00 W and at 31 dBm it is 4.74 W. So the higher output figure is also the cooler one, because the DC current is fixed by the row rather than by the output. A buyer who quotes 31 dBm and sizes the heatsink for it will be 5 percent short if the part delivers 30.

Five watts on 3.3 square millimetres

The die is 3.31 mm2, so 5.00 W is 1.51 W per mm2 and 1.0 W of RF is 302 mW per mm2. That is high for a GaAs MMIC and it decides the mounting: copper molybdenum or diamond, a thin die attach with a low void rate, and a carrier that spreads the heat before it reaches the housing. The 22 dB of gain means the drive at compression is 8 dBm, 6.31 mW, so a single 10 dBm driver stage reaches it. At 31 dBm the drive is 9 dBm.

The band is 40 percent fractional around a 15 GHz centre, wider than the 8 to 13 GHz part and wide enough to cover the upper Ku satellite allocation and the lower K band radar allocation with one die. Input and output VSWR are both 1.5 to 1, which is 13.98 dB return loss and 0.177 dB of mismatch loss at each port, small enough to ignore in a first pass link budget.

Forty dBm of intercept without a reference level

OIP3 is 40 dBm, ten dB above the 30 dBm compression point. A ten dB spacing is normal for a part optimised for linearity, and this one says so in its own overview. What the row does not give is the two tone level and the spacing that produced 40 dBm. Intercept is a function of both, and a figure quoted at 20 dB of back off cannot be compared with one quoted at 6 dB. For a multi carrier application, ask for the tone level and the spacing before using 40 dBm in a spurious budget.

Two questions buyers ask

Do I quote 30 or 31 dBm? Quote 30 dBm, the device page and the 1 W description, and treat 31 as typical. On a purchase order that asks for a guaranteed minimum, the lower figure is the one that survives a test bench.

Can the die carry 5 W continuously? It can if the carrier is chosen for it and the junction temperature is held. Ask for the thermal resistance the vendor measured, and check the pulse duty cycle before running the part at compression in a radar chain. For a continuous application, the heat has to come down through the bias current rather than through the drive level, because the DC input is set by the bias and barely moves when the RF envelope falls.

SUPERB quotes the HX339160 against the device page figure and confirms 30 or 31 dBm with the vendor before an order is released. For a swap from a discontinued die, obsolete RF part replacement lists the parameters that have to match, including the carrier and the attach. Die attach and carrier supply sit under RF component sourcing, and stock is quoted from the amplifiers and comparators index.

HX339160: an amp of drain current at 12 to 18 GHz, and the supply that has to deliver it

Twenty two dB of gain and 1 W at compression from a 5 to 6 V rail drawing 1 A. The output figure and the heat have been covered. The remaining engineering is on the DC side: an amp of drain current through a bias feed at Ku band is a different problem from a hundred milliamps, and the third order intercept the row publishes at 40 dBm only becomes useful once it is converted into a tone spacing at a specific back off.

One amp changes the bias network, the regulator and the connector

At 6 V and 1 A the DC input is 6 W and the heat is 5 W at 30 dBm of output. The bias feed has to carry 1 A without saturating and still present a high impedance at 18 GHz, which in practice means a quarter wave high impedance line on a low loss substrate or a purpose made bias tee rather than a discrete inductor. The regulator has to hold 6 V while the current steps by hundreds of milliamps when the bias is sequenced, and the trace from the regulator to the die has to be short enough that the 1 A pulse does not droop by more than the gate can absorb. A 20 milliohm trace at 1 A is 20 mV, which is harmless; a 200 milliohm trace is 200 mV, which is not, because the drain voltage sets the compression point and the gain of the stage.

Per tone outputThird order productDistance below the tone
28 dBm4 dBm24 dBc
26 dBmminus 2 dBm28 dBc
24 dBmminus 8 dBm32 dBc
22 dBmminus 14 dBm36 dBc

What forty dBm of intercept means at a real back off

The row publishes a 40 dBm output intercept and no test level. The standard relation for two equal tones is that the third order product sits 2 times the difference between the intercept and the tone below the tone, so at 28 dBm per tone the products are 24 dBc down, at 26 dBm they are 28 dBc down, and at 24 dBm they are 32 dBc down. With 30 dBm at compression, a two tone test at 26 dBm per tone is already 4 dB into compression and the intercept relation stops holding, so the useful readings are the ones at 24 dBm and below. If the system specification is 30 dBc of third order suppression, this part works at about 25 dBm per tone, which is 28 dBm of total output and 2 dB below compression. That is a narrow window and it is the reason a modulated carrier at this band usually runs at 6 to 8 dB of back off.

The band is 40 percent wide and one die covers both allocations

Twelve to eighteen GHz covers the upper Ku satellite allocation and the lower K band radar allocation, and the family row gives one gain and one output for both. A 40 percent fractional band is wide enough that the gain at 12 GHz and at 18 GHz will not be the same, and the row does not say which end the 22 dB belongs to. On a radar the transmit band is usually narrow within that span, so the part can be optimised at the operating frequency by the external match, and on a satellite block the full band matters and a sweep at the two edges is the first measurement to make.

Two questions buyers ask

Can it be driven by a 10 dBm source? 22 dB of gain with 30 dBm out needs 8 dBm, so yes, with 2 dB of margin. At 31 dBm it needs 9 dBm, which is still inside a single driver stage.

Is 30 or 31 dBm the number to work to? Treat 30 dBm as the guaranteed figure and 31 as typical, and size the heat path for 5.0 W rather than 4.74 W. The parts in this group and their two published readings are compared on amplifier cross reference, the die attach and carrier question is on defense grade components, and the family supply position is on RF component sourcing.

Request Quote — In Stock, 100% New & Original, Full Warranty