Post Go back to editing

Why SAR+FIR ADCs are always Bipolar rather than True-Bipolar input? And how to convert.

Interested parts: LTC2512-24

Hi experts,

I'm looking for a high-precision ADC with the specification below:

SNR > 100 dB, the larger the better
INL about 1 ppm acceptable
Bandwidth > 3 kHz

I prefer SAR+FIR type ADC because it seems to have better sampling accuracy (I guessed because I heard that the SAR takes snapshots). But delta-sigma is also ok.

My input type is true-bipolar single-ended. The single-ended is acceptable in LTC2512's datasheet because of the 130dB CMRR, but the typical application chapter uses a single-to-differential converter. Can I infer that ADI recommends the differential input to achieve the performance written in the datasheet?

I found that there's no precise SAR+FIR ADC supporting the true-bipolar input. I wonder the reason (because of the SAR architecture?). How can I convert my true-bipolar signal to a bipolar signal? Will this conversion circuit significantly lower the ADC's SNR? (If so, I'll choose a delta-sigma one that supports the true-bipolar input)

Thank you.

Parents Reply
  • The true-bipolar to bipolar method used should not affect SNR. The LTC6363/LTC2500-32 circuit referenced achieves an SNR of 103.3dB. THD is -122.2dB. This is equivalent to a little better than a 1ppm INL.For slightly better performance you might replace the LTC6363 with the ADA4945-1. This amplifier has lower noise and distortion.