How to Select Analog-to-Digital Converters

A / D devices and chips are frequently-used peripheral devices,used to achieve single-chip data acquisition .The kinds and performance of  A / D converter are various.When we design data acquisition system, the first things we should consider  is how to select the appropriate A / D converter to meet the system design requirements. Select the A / D converter device needs to consider the quality of the device  and the application of the occasion requirements.We can basically find the appropriate A/D device according to the following indicators.

(1) A / D converter bits

To confirm the A / D converter bit,we should pay attention to two aspects,one is the  static precision ,and then is the dynamic smoothness. For the static precision, we should consider the original error of the input signal to the output of the error, which is the main cause of generating the analog signal digitization error. The quantization error is related to the number of A / D converters. Generally,we called A / D converter as low-resolution when the bit is equal or less than 8-bit , 9 to 12-bit called the resolution of the converter, more than 13-bit called as high-resolution converter . The 10-bit A / D chip usually follow a larger error, and 11-bit or more always are not useful for reducing the error,with a high-requirements for the A / D converter. Therefore, it is appropriate to take 10 or 11 bits.

The analog signal would firstly through the measuring device, and then converted by the A / D converter.As a result, the total error is  a combination of measurement errors and quantization errors. The accuracy of the A / D converter should  match the measuring device.In other words , on the one hand, it is required that the quantization error should be a smaller one in the total error  . On the other hand, we should Make the appropriate request to A/D bit according to the precision level of the current measuring device .

At present,most of the measurement devices' accuracy is not less than 0.1% to 0.5%.we can set the  A / D converter accuracy between  0.05% and 0.1% ,with 10 to 11-bits corresponding binary code and the sign bit,then we get 11 to 12 bits. if you have special applications ,required a larger number of bits, you can change to double-precision conversion program.

(2) The conversion rate of A / D converter

it would take certain conversion time to output a stable digital from the beginning to the end of conversion. The reciprocal of the conversion time is the number of conversions that can be done per second, called as the conversion rate.

When determining the conversion rate of the A / D converter, the sampling rate of the system should be taken into account. For example, if the conversion time is 100us, the conversion rate is 10kHz. According to the sampling theorem and the actual needs , a cycle of the waveform would be taken 10 samples. It means that the maximum processing frequency of these A / D converters is only 1KHz analog signal. Reducing the conversion time, the signal frequency can be increased. For the general single-chip,it is difficult to complete other works such as reading data, restart, save data after achieving the A/D convert.

(3) Sampling / holding device

you don't need to use track-and-hold when the DC' analog signals are very slow . But most time ,sample and hold are necessary. If the signal frequency is not high, A / D converter conversion time is short,we change to use high-speed A / D sampling rather than sample / hold.

(4) A / D converter range

A / D conversion is bipolar ,but sometimes is unipolar. The minimum value of the input signal can be zero-based or non-zero. Some converters provide different range of pins.The factors attracting the A / D converter range include: range conversion and bipolar bias; the dual reference voltage;  the correct use of A / D converter internal comparator input.

(5) Full-scale error

The difference between the corresponding input signal and the ideal input signal value at a full-scale output.

(6) linearity

The maximum shift of the transfer function of the actual converter to the ideal straight line.

   Author ：Darcy Beaufort — WIN SOURCE ELECTRONICS