Modulation in analog domain / Demodulation in digital domain

[madlock84]

Hi Guys,

i hope i put this under the right topic.
I am currently working on the topic of digital calibration of A/D-Converters and i am ivestigating different approaches of calibration.
In one of these approaches the analog signal is modulated with a Pseudo-Random-Number-Sequence (PRN-Sequence). In other words, the PRN-Sequence is added to the analog input-signal of the A/D_Converter. Afterwards the modulated signal is processed by a 1-Bit-Delta-Sigma-Converter. The output signal of this converter is a bit-stream, of course. Then, in the digital domain the added PRN-Sequence is removed and Conversion-Errors can be determined by correlating the outpur-signal with the PRN-Sequence.
Although this procedure sounds pretty easy i have no idea how to remove the PRN-Sequence from the digital output signal of the A/D-Converter. Adding a wighted PRN-Sequence to the analog signal is no problem, but how can i remove this sequence from the digtal output of the converter? I am obviouly having two one-bit-signals. How do i substract them and if i do so, do i get the correct converter output?
Can anybody help me on this problem?

Thanks in advance.

PS: Calibration approach is given in the following paper (Access to IEEE-Xplore required): https://ieeexplore.ieee.org/xpl/art...tal+calibration+of+nonlinear+memory+errors+in

 

[FvM]

In one of these approaches the analog signal is modulated with a Pseudo-Random-Number-Sequence (PRN-Sequence). In other words, the PRN-Sequence is added to the analog input-signal of the A/D_Converter.

Modulation will be hardly an additive operation. You should clarify about the intended modulation method.