AD7621ASTZ
Product Specifications
AD7621ASTZ
Analog Devices — Analog & Mixed-Signal ICs
Product Overview
The AD7621ASTZ 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
AD7621ASTZ: 3 MSPS at 16 Bits From a 2.5 V Rail and 65 mW
The AD7621 is a 16-bit successive approximation converter with three speed modes: 3 MSPS in wideband warp and warp mode, 2 MSPS in normal mode and 1.25 MSPS in impulse mode. It runs from a single 2.5 V supply, has a 2.048 V internal reference, a differential input range of plus and minus the reference, an integral nonlinearity of 2 LSB maximum and 1 LSB typical, signal to noise and distortion of 89 dB at 100 kHz, total harmonic distortion of minus 101 dB, and dissipates 65 mW at 3 MSPS. It has no pipeline delay, offers a parallel 16-bit or 8-bit bus and a serial port at 5 V, 3.3 V or 2.5 V, and comes in a 48-lead LQFP or LFCSP.
62.5 uV per code and 14.5 effective bits at 100 kHz
With a 2.048 V reference the differential span is 4.096 V, so one code is 62.5 uV and 2 LSB of maximum nonlinearity is 125 uV. A signal to noise and distortion figure of 89 dB at 100 kHz corresponds to 14.5 effective bits, which means the top 1.5 bits of the 16-bit word are noise at that frequency. The 89 dB is a 100 kHz number and it falls as the input frequency rises, so a receiver digitising at 30 MHz of intermediate frequency will see fewer effective bits than the table suggests, and the degradation with frequency is the parameter that should be read before the headline rate is chosen.
Three speed modes and what changes between them
The mode pins trade conversion rate against noise and power. Warp mode runs the converter at 3 MSPS with the shortest acquisition window, which requires the driving amplifier to settle faster and produces a slightly higher noise floor. Normal mode at 2 MSPS gives more acquisition time and a quieter result. Impulse mode at 1.25 MSPS powers the converter down between conversions and is intended for a signal that is sampled in bursts, where the average power matters more than the instantaneous rate. The three modes are a single part doing three jobs, and the choice is made by the mode pins rather than by a different order code.
| Parameter | Value | What it decides |
|---|---|---|
| Resolution | 16 bits, no missing codes | 62.5 uV per code on a 4.096 V span |
| Throughput, warp | 3 MSPS | 333 ns per conversion |
| Throughput, normal | 2 MSPS | 500 ns per conversion, longer acquisition |
| Throughput, impulse | 1.25 MSPS | Burst operation with power down between samples |
| Integral nonlinearity | 2 LSB maximum, 1 LSB typical | 125 uV, twice one code |
| Reference | 2.048 V internal | 4.096 V differential span |
| Input range | Plus and minus the reference | Differential, 4.096 V peak to peak |
| Signal to noise and distortion | 89 dB at 100 kHz | 14.5 effective bits at that frequency |
| Total harmonic distortion | Minus 101 dB at 100 kHz | Spurs below the noise floor at that input |
| Power | 65 mW at 3 MSPS | 21.7 mW per MSPS |
| Supply | 2.5 V single | A low voltage analog rail, no 5 V needed |
| Interfaces | Parallel 16 or 8 bit, serial 5 V, 3.3 V, 2.5 V | Host choice without level shifting |
| Package | 48-lead LQFP or LFCSP | Two footprints for one function |
65 mW at 3 MSPS is a thermal and a supply question
Sixty five milliwatts in a 48-lead package is not a thermal problem, but it is a supply noise problem: the current is drawn in bursts at the conversion rate, so the supply sees a 3 MHz component that couples into the analog input network if the decoupling is not local and low inductance. The data sheet recommends the decoupling arrangement, and a board that shares a ferrite bead with a digital load will show spurs at the conversion rate. The same current burst is why the reference pin needs its own capacitor close to the part: the reference sees the switched capacitor load directly.
What a 2.5 V analog rail buys and costs
A 2.5 V analog supply lowers the power for a given conversion rate and removes the need for a 5 V analog rail on a board that is otherwise 3.3 V and below. The cost is headroom: the input range is set by the reference, which is limited by the supply, so the full scale is 4.096 V differential and no more. In a signal chain where the preceding amplifier runs from 5 V, that is fine, because the amplifier's output swing is larger than the converter's input. In a chain that runs from a 3.3 V rail, the amplifier's swing and the converter's range are close enough that the interface has to be checked carefully.
Buying a high rate 16-bit converter
At this rate the converter is bought together with the amplifier that drives it and the reference that defines it, and the three have to be evaluated as a set. Parts that can be compared on rate, input span and interface are listed in ADC DAC cross reference, the amplifier side of the same interface is covered in amplifier cross reference, and consolidating a fast converter bill of materials is described in one stop BOM sourcing.