1 Features

• Dual channel

• 14-bit 125 MSPS ADC

• Noise floor: –153 dBFS/Hz

• Ultra-low power with optimized power scaling:

82 mW/ch (125 MSPS)

• Latency: 1 clock cycle

• 14-Bit, no missing codes

• INL: ±1.5 LSB; DNL: ±0.5 LSB

• Reference: external or internal

• Industrial temperature range: –40°C to +105°C

• On-chip digital filter (optional)

– Decimation by 2, 4, 8, 16, 32

– 32-bit NCO

• DDR and Serial CMOS interface

• Small footprint: 40-WQFN (5 mm × 5 mm) package

• Single 1.8-V supply

• Spectral performance (fIN = 5 MHz):

– SNR: 74.0 dBFS

– SFDR: 90-dBc HD2, HD3

– SFDR: 100-dBFS worst spur

• Spectral performance (fIN = 70 MHz):

– SNR: 72.5 dBFS

– SFDR: 70-dBc HD2, HD3

– SFDR: 85-dBFS worst spur

2 Applications

• High-speed data acquisition

• Industrial monitoring

• Thermal imaging

• Imaging and sonar

• Software defined radio

• Power quality

• Communications infrastructure

• High-speed control loops

• Instrumentation

• Smart grids

• Spectroscopy

• Radar

3 Description

The ADC3644 device is a low-noise, ultra-low power,

14-bit, 125-MSPS dual-channel, high-speed analogto-

digital converter (ADC). Designed for low power

consumption, the device delivers a noise spectral

density of –153 dBFS/Hz, combined with very good

linearity and dynamic range. The ADC3644 offers IF

sampling support, which make the device an excellent

choice for a wide range of applications. High-speed

control loops benefit from the short latency of only one

clock cycle. The ADC consumes only 82 mW/ch at

125 MSPS, and power consumption scales well with

lower sampling rates.

The ADC3644 use a DDR or a serial CMOS interface

to output the data offering lowest power digital

interface, together with flexibility to minimize the

number of digital interconnects. These devices are

a pin-to-pin compatible family with different speed

grades. These devices support the extended industrial

temperature range of –40 to 105⁰C.

ADC3642属于集成电路（IC） > 模数转换器（ADC）。美国德州仪器公司制造生产的ADC3642模数转换器（ADC）模数转换器（ADC、A/D 或 A 转 D）对模拟信号进行采样，例如由麦克风采集的声音或传感器的输出，然后将其转换为数字信号。一般来说，数字输出是二进制补码二进制数，与输入成比例。输入类型可能是差分、伪差分或单端输入。模数转换器可按照位数、采样率、输入数、接口、转换器数和架构（例如自适应增量、双斜率、折叠、流水线、SAR、三角积分、两步式）进行选择。