need help on explaining the delta-sigma spectrum

[sapphire]

Hi,

I am designing a 2nd order delta sigma modulator using CIFB structure. The supply voltage is 1.5V, and feedback reference is also 1.5V. If my differential input voltage peak is 1.6V, the in-band quantization noise level looks OK, but there are also very high harmonic tones. Then if I reduce the differential input voltage peak to 0.8V, the harmonic tones are greatly suppressed, but in-band noise level raises by almost 50 dB. It's so weird. Please help me if you know anything about the issue.

Thanks

Sapphire

 

[YL.WANG]

Hi,

I don't know what is the real reason precisely, but if you are using 1-bit quantizer maybe it is because of the input-signal-dependent nonlinear behavior of the 1-bit quantizer. The equivalent quantizer gain of the binary quantizer is indeterminate as the multibit one.