The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Mixed-Signal Design >> Why is the SNDR of signal-delta ADC is not dominated by mismatch?
https://designers-guide.org/forum/YaBB.pl?num=1516522644

Message started by Jacki on Jan 21^st, 2018, 12:17am

Title: Why is the SNDR of signal-delta ADC is not dominated by mismatch?
Post by Jacki on Jan 21^st, 2018, 12:17am

Hello,

Recently I read some documents about SAR-ADC and sigma-delta ADC. I find for SAR-ADC if mismatch calibration is not used, the SNDR is limited by mismatch, and it is hard for SAR-ADC to reach more than 12 bits. But for sigma-delta ADC, as I remembered, its SNDR is not dominated by mismatch, in other words, mismatch is not a big impact on the SNDR. Could anybody explain the reason for me, is it right for both CT-sigma-delta ADC and SC-sigma-delta ADC?
Thank you very much.
Jacki

Title: Re: Why is the SNDR of signal-delta ADC is not dominated by mismatch?
Post by sheldon on Jan 21^st, 2018, 8:25pm

Jacki,

It depends on the architecture of the Delta-Sigma ADC architecture.
Traditionally, the Delta-Sigma ADC uses a 1 bit DAC and a 1-bit DAC does
not have mismatch. However, it is very hard to achieve high dynamic
range using a 1-bit DAC. To achieve high dynamic range, multi-bit DACs
are used. When multi-bit DACs are used, special techniques are required
to reduce the impact of mismatch: element rotation, data-weighted
averaging, ... So to achieve high dynamic range, special methods are
required for Delta-Sigma converters.

Sheldon

Title: Re: Why is the SNDR of signal-delta ADC is not dominated by mismatch?
Post by Jacki on Jan 22^nd, 2018, 10:22pm

Hi Sheldon,

Thank you very much for your reply. I am new to sigma-delta ADC, want to learn this charming structure.

Title: Re: Why is the SNDR of signal-delta ADC is not dominated by mismatch?
Post by Jacki on Mar 20^th, 2018, 1:44am

Hi Sheldon,

About the multi-bit DAC, normally if we use binary DAC, the linearity is good. If we use the thermometer coded to binary DAC, I think we almost can guarantee the linearity of the multi-bit DAC, am I right?