Analog Dialogue: A forum for the exchange of circuits, systems, and software for real-world signal processing.
Home     Analog Devices     Feedback     Subscribe     Archives     简体中文     日本語
Advanced Search

Rarely Asked Questions Ė Issue 103, March 2014

Programmable ADC Input Range Provides System Benefits

adjustable_input

Q: Whatís the advantage of an adjustable or programmable analog input range?

A: While older high-speed analog-to-digital converters required the external reference voltage to be adjusted to change the analog input range, most of the latest designs include programmable or adjustable input ranges. A simple register write can program the range from 1 Vp-p to 2 Vp-p in 100-mV steps, for example. For most devices, the default input range is set to the maximum, as this will maximize the signal-to-noise ratio (SNR), which is a simple fraction: signal/noise. The noise, typically dominated by the converterís own thermal noise, will not change when the input range is adjusted, so maximizing the input range also maximizes the SNR.

Occasionally, an engineer will review the data sheet and ask why anyone would want to adjust the range, as the value of reducing it isnít clear. Thatís a fair question, as the reduced signal range does not provide a significant trade off with the converterís other specifications. You might expect that the distortion would improve with the lower signal swing, just as it would with op amps or other linear components. The improvement is typically not very dramatic in an ADC, however, especially near full scale where the converterís nonlinearity will dominate its distortion. As an example, the AD9467 16-bit, 250-MSPS ADC provides typical performance specifications for both 2 Vp-p and 2.5 Vp-p input ranges. Reducing the input range lowers the SNR by 1.7 dB, while increasing the spurious free dynamic range (SFDR) in the first Nyquist zone by 1 or 2 dB. The 1.7 dB SNR trade for 1 to 2 dB of distortion may sound similar, but the spur is at a discrete frequency in the Nyquist zone, while the lower SNR raises the average the noise floor across the entire Nyquist zone. A 1 or 2 dB improvement at a discrete frequency may be negligible in most applications, while the 1.7 dB degradation across the entire band of interest can be significant.

So why lose the SNR if you arenít gaining anything? Well, the converter is only one piece of the puzzle at the system level. When budgeting for cost, performance, and power dissipation of the entire system, the lower signal range means that 1.7 dB less gain is needed upstream of the ADC, perhaps allowing the use of a lower power op amp with a lower gain-bandwidth product. So if the reduced noise performance from the converter is acceptable, you may find a positive tradeoff elsewhere in the system.

References

Apples and Oranges - How is the input referred noise of an ADC related to its signal-to-noise ratio (SNR)?

MT-003: Understand SINAD, ENOB, SNR, THD, THD + N, and SFDR so You Don't Get Lost in the Noise Floor

MT-228: High Speed ADC Analog Input Interface Considerations

 

Author
david buchanan  David Buchanan [david.buchanan@analog.com] received a BSEE from the University of Virginia in 1987. Employed in marketing and applications engineering roles by Analog Devices, Adaptec, and STMicroelectronics, he has experience with a variety of high-performance analog semiconductor products. He is currently a senior applications engineer with ADIís High Speed Converters product line in Greensboro, North Carolina.

 Have a question involving a perplexing or unusual analog problem?

Submit your question to: www.analog.com/askdavid

 

Copyright 1995- Analog Devices, Inc. All rights reserved.