raymond28j
Newbie level 1
Hi,
The signal and noise transfer function of a continuous time first-order delta sigma modulator using an simple active RC integrator both have a pole at 1/RC. The transfer functions are attached.
I was wondering, ideally, why can't we make RC very small, so the quantization noise is lower in the band of interest? while the signal transfer function is always one at low frequencies.
I simulated the circuit attached in cadence using ideal parts and found that changing RC does not improve the performance, but even introduce instability when I was using very small or large RC values.
Why is that? And what will the situation be in practical?
Thank you very much
Raymond
The signal and noise transfer function of a continuous time first-order delta sigma modulator using an simple active RC integrator both have a pole at 1/RC. The transfer functions are attached.
I was wondering, ideally, why can't we make RC very small, so the quantization noise is lower in the band of interest? while the signal transfer function is always one at low frequencies.
I simulated the circuit attached in cadence using ideal parts and found that changing RC does not improve the performance, but even introduce instability when I was using very small or large RC values.
Why is that? And what will the situation be in practical?
Thank you very much
Raymond
