Design Con 2015

Wideband error correction elevates time-interleaved ADCs

-July 26, 2013

Digital post-processing employing digital filtering and interpolation can in principle equalize mismatches over frequency to an arbitrarily low well. However, due to the fundamental limitation of digital filters in representing arbitrary phase responses at integer multiples of fs/2, the digital mismatch error correction can only remove mismatch errors over a frequency region that is strictly narrower than a Nyqust frequency bandwidth. In practice, this does not constitute any real drawback since the frequency region where the error correction effect is reduced due to these digital filter limitations coincides with where transition regions are located for the analog anti-aliasing filters.

In Figure 5, measured SFDR performance levels are plotted for the first and second Nyquist frequency band. In the first Nyquist frequency band, the suppression of aliasing distortion due to time-interleaved ADC mismatch occurs in a “correction frequency band” having a low-pass character (see the left-hand part of Figure 5).  When the ADC analog input bandwidth is high enough, operation in the second Nyquist frequency band of the ADC array is possible. For Nyquist frequency bands above the first, the correction frequency band has a band-pass character (the right-hand part of Figure 5 shows the post-processing improvement in the second Nyquist frequency band). 

From Figure 5 one can conclude measured SFDR performance levels of around 90 and 80 dBFS for the first and second Nyquist frequency band respectively, which is what are typically expected for a single-core 14-bit ADC. The large benefit here is that the sampling frequency has in fact been doubled compared with the state-of-the-art single-core 14-bit ADC. Thus, one can draw the conclusion that digital time-interleaved ADC mismatch error correction makes the SFDR-performance of the time-interleaved ADC array to correspond to that of a single-core ADC. This is true both for solutions using discrete time-interleaved ADCs as well as singe-die ADC implementations.



Figure 5: Measured SFDR versus input frequency for the first (left) and second (right) Nyquist frequency band for the same time-interleaved system as above. Dashed lines show the unprocessed spurious performance level while the solid lines show the measured performance after digital mismatch error correction over frequency.

If you liked this feature, and would like to see a collection of related features delivered directly to your inbox, sign up for the Test & Measurement newsletter here.


About the Author

Per Lowenborg’s profile.


References
  1. http://www.analog.com/static/imported-files/data_sheets/AD9467.pdf
  2. http://www.ti.com/lit/ds/symlink/ads5474.pdf
  3. http://www.e2v.com/e2v/assets/File/documents/broadband-data-converters/EV12AS200.pdf
  4. Black W. C. and Hodges D.A., “Time interleaved converter arrays,” IEEE J. Solid-State Circuits, vol. SC-15, no. 6, pp. 1022-1029, Dec. 1980.
  5. Vogel C. and Johansson H, “Time-interleaved analog-to-digital converters. Status and future directions,” in Proc. IEEE Int. Symp. Circuits and Syst., Kos Greece, May 2006, pp.3386-3389.
  6. http://spdevices.com/index.php/products/silicon-and-fpga-ip

Loading comments...

Write a Comment

To comment please Log In