beargebinjie
Junior Member level 3
Hi, guys, i am designing a sigma-delta ADC for digital audio application. The signal bandwidth is 25KHz, and the design SNR is to achieve 90dB.
I chose the MASH2-1 architecture, and the OSR is 128, Fs=6.4MHz, so in the theory the dynamic range will be 127dB, and after parameter scaling, the mash modualtor in the MATLAB had the SNR equal to 116dB.
I built the modulator with the ideal opamp, switch and comparator in the HSPICE, when the switch on resistace was 1ohm, the snr was 104dB, the first stage(second order) snr was 86dB, close to the result in the MATLAB which was 89dB. But the second stage(first order) snr was only 12dB, less than the result in the MATLAB which was 19dB, the low frequency noise floor was much high than that in the MATLAB's result.
Here is the big problem when i change the switch on resistance to 50ohm(or 100ohm,200ohm....), the modualtor's snr will drop dramatically to 65dB!! the inband noise was horizontal, was not shapped at all.But the tran wave showed that the opamp 's outputs were settle completely, and all the time seems right, 50ohm is very small resistance in the SC circuit in my opinion, i donot know why this happen, do you guys know the reason, thanks a lot!
I chose the MASH2-1 architecture, and the OSR is 128, Fs=6.4MHz, so in the theory the dynamic range will be 127dB, and after parameter scaling, the mash modualtor in the MATLAB had the SNR equal to 116dB.
I built the modulator with the ideal opamp, switch and comparator in the HSPICE, when the switch on resistace was 1ohm, the snr was 104dB, the first stage(second order) snr was 86dB, close to the result in the MATLAB which was 89dB. But the second stage(first order) snr was only 12dB, less than the result in the MATLAB which was 19dB, the low frequency noise floor was much high than that in the MATLAB's result.
Here is the big problem when i change the switch on resistance to 50ohm(or 100ohm,200ohm....), the modualtor's snr will drop dramatically to 65dB!! the inband noise was horizontal, was not shapped at all.But the tran wave showed that the opamp 's outputs were settle completely, and all the time seems right, 50ohm is very small resistance in the SC circuit in my opinion, i donot know why this happen, do you guys know the reason, thanks a lot!