1、 Function of digital to analog converter

Real world analog signals, such as temperature, pressure, sound or image, are constantly converted into digital forms that are easier to store, process and transmit. But in many systems, digital information must be converted to analog signal to achieve some real-world functions. Digital to analog converter (DAC) can do this, and their output can also be used to drive a variety of devices, such as loudspeakers, engines, RF transmitters and temperature controllers.

DAC is usually placed in digital system. In digital system, some real-world signals are digitized and processed by analog-to-digital converter (ADC), and then need to be converted into analog signals again. The required DAC performance of these systems will be affected by the performance and requirements of other components of the system.

2、 Basic principles

The DAC generates a quantized (discrete phase) analog output in response to a binary digital input code. The digital input may be TTL, ECL, CMOS or LVDS, while the analog output is a voltage or current output.

To produce an output, a reference (voltage or current) is divided into binary or linear segments. The digital input then drives a switch that connects a certain number of segments to the output. The number of segments reflects the number of possible digital input codes, which is a function of the resolution of the converter, or can also reflect the number of bits (n) in the input code. There are n digits, and the number of possible codes is 2n.

The amplitude of DAC output can be expressed as follows:

Analog output = digital input code / (2n-1) × The reference input analog signal is a continuous time domain signal with infinite resolution. However, the output of DAC is a signal based on discrete value (quantization), which also has a unified but infinite time interval (sampling). In other words, the DAC output tries to show an analog signal with infinite resolution and bandwidth.

Quantization and sampling are another basic but predictable limitation of DAC performance. Quantization determines the maximum dynamic range of the converter and produces quantization error and noise. According to Nyquist standard, sampling determines the maximum bandwidth of DAC output signal.

In an ideal DAC, the analog signal is just a least significant bit (LSB) part, and an LSB is a full-scale analog output amplitude divided into 2n segments, where n is the DAC resolution in the form of bits.

However, the operation of DAC in the real world will also be affected by some non ideal effects, which can not achieve the ideal effects specified by quantification and sampling. These errors are characterized by some AC and DC performance specifications, which determine the static and dynamic performance of the converter.

1 offset error

Offset error is the deviation between the output of DAC and the output of ideal transfer function when gain error is zero. For all input codes, the offset error is constant.

2 gain error

The gain error is the deviation between the slope of the transfer function of the converter and the ideal transfer function, which is measured by compensating the offset error to zero.

3 differential nonlinearity (DNL)

Differential nonlinearity refers to the deviation between an actual step size and an ideal LSB step size. DNL error will produce additional noise and burr beyond the scope of quantization effect.

4 integral nonlinearity (inl)

The integral nonlinearity is the error of the actual output voltage and the ideal output voltage on the line between the ends of the transfer function. Inl is calculated after offset error and gain error are eliminated. The inl error results in additional harmonics and burrs.

5 monotonicity

If the output of a DAC increases or remains unchanged as the digital input code increases, then the DAC is monotonic. On the contrary, if the output of a DAC decreases with the increase of digital coding, then the DAC is nonmonotonic.

6 establishment time

The setup time is the time when the analog output responds to a step change on the digital input and determines a value within a limited error range.

7 false signal

False signal refers to the load added from input to output when the input and output of converter change state.

8 digital feedthrough

Digital feed through refers to the output noise caused by the high frequency logic signal penetrating into the output of the converter when the DAC is not started.

Noise has many characteristic parameters, such as parasitic free dynamic range (SFDR), total harmonic distortion (THD) and signal-to-noise ratio (SNR). The other parameter thd + n is the ratio of Total Harmonic Distortion plus noise to fundamental amplitude.

The following figures show several architectures of DAC.

Editor in charge: GT