HMC637A-Die
Product Specifications
HMC637A-Die
Analog Devices — Analog & Mixed-Signal ICs
Product Overview
The HMC637A-Die from Analog Devices is a high-performance analog / mixed-signal IC providing precision signal acquisition, conversion, and conditioning. Analog Devices (ADI) is a world leader in high-performance analog, mixed-signal, and DSP ICs for instrumentation, healthcare, automotive, and communications.
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-resolution data conversion with excellent SNR and THD
- Wide dynamic range and high linearity across operating conditions
- Integrated voltage reference with low temperature drift
- Flexible digital interface — SPI, I2C, or parallel options
- Industrial/automotive temperature range with guaranteed specs
- Low power consumption with power-down modes
Key Specifications
| Resolution | bits |
| Channels | — |
| Sample Rate | kSPS |
| Interface | — |
| Supply Voltage | V |
| Operating Temp | °C |
| Package | — |
| INL | LSB |
| SNR | dB |
| THD | dB |
Applications
Request Quote — In Stock, 100% New & Original, Full Warranty
HMC637A die: 29 dBm from 0.1 GHz to 6 GHz and four bypass pins that are not optional
The HMC637A is a gallium arsenide pHEMT distributed power amplifier die covering 0.1 GHz to 6 GHz with 13 dB of gain, 29 dBm of output power at one decibel compression, 44 dBm of output third order intercept and 400 mA from a 12 V supply. Gain flatness is 0.75 dB across the whole range, the input and output are matched to 50 ohm on chip, and four access points bring the drain and gate networks out to be bypassed at the die. Those four pins are where most first builds lose their gain flatness.
Published values
| Parameter | Value |
|---|---|
| Frequency range | 0.1 GHz to 6 GHz |
| Gain | 13 dB |
| Gain flatness | 0.75 dB from 100 MHz to 6 GHz |
| Output power at 1 dB compression | 29 dBm |
| Output third order intercept | 44 dBm |
| Supply | 12 V at 400 mA |
| Package | die, backside ground |
| Matching | 50 ohm input and output on chip |
| Bypass points | four access pins, ACG1 to ACG4 in the packaged equivalent |
4.8 W of DC against 0.79 W of output
12 V at 400 mA is 4.8 W. At the 29 dBm compression point the output is 0.79 W, so 4.0 W becomes heat on the die face. That is the figure that decides the carrier and the attach, and it is the reason this part is often bought as die for a module where the carrier is already there, rather than as a package on a laminate.
The 44 dBm of output third order intercept against a 29 dBm compression point is 15 dB of headroom, which is what a multicarrier or a chirped system needs. It also means the part is being run well backed off in linear use, where the dissipation does not fall much because the DC input stays at 4.8 W. A linear application at 20 dBm of average output still dissipates 4.7 W, so the thermal answer does not improve with the drive level.
Four bypass points and a gate that has two halves
The gate and drain networks are brought out at four points because the distributed structure is long in electrical terms and the two ends of the line are not at the same potential. Each point needs its own capacitor placed at the die, not shared with the next one and not moved back to the edge of the carrier. Sharing one capacitor across two access points puts the current of one half of the line through the bypass of the other, and the symptom is a ripple in the gain that moves with frequency.
The backside is both radio frequency and direct current ground, which simplifies the assembly: no backside via etching and no separate ground path to route. What it does require is that the carrier under the die is a solid ground plane with an attach that has no voids, because the ground return for 4 W of heat and several amps of current runs through the same interface.
Die or package, and the test that decides
The packaged equivalent brings the same amplifier in a 32-lead, 5 mm by 5 mm leadframe package with the access points already on pins. The die version is the one to take when the module already exists and the package height or the transition loss matters. The package is the one to take when the assembly flow cannot bond.
The acceptance test is the same either way: gain from 100 MHz to 6 GHz, gain flatness across the band, compression point and drain current at 12 V. A die measured in a fixture with 0.025 mm bonds and a package measured on a connectorised board will not agree to better than a few tenths of a decibel, and comparing the two sets of numbers as if they were the same measurement is where a substitution study goes wrong.
Buyers usually arrive at the RF power transistor cross reference to find the family alternatives, then use RF procurement to confirm a substitute before it is fitted; one stop RF supply is the routine for a part that has already stopped shipping.
Two questions buyers ask
What happens if one of the four bypass capacitors is left out? The gain ripple changes and the low frequency end of the band is affected first, because the access points are part of the drain and gate networks rather than test pads. It also raises the risk of a low frequency oscillation that is invisible until the part is put in a closed loop.
Is 4 W of dissipation a problem for die in a module? It is the carrier and the attach that answer that, not the die. A void free attach on a metal carrier is what the 4.0 W expects, and the same die on a thin laminate with a poor ground plane will run hotter than the case temperature suggests.