Question regarding two SigmaDelta circuit difference

[hyleeinhit]

HI , Guys,

I have a question regarding current mode SigmaDelta circuits. Two circuit structure are attached. Compared to "Circuit1", OTA is removed in "Circuit2". What is advantages and disadvantages of these two designs, respectively? To my knowledge, the first circuit is more popular that the second one. WHY?
Circuit1

Circuit2

 

[helpmejerry]

At a quick glance I would say that in the second figure power consumption is lower and the input independences are different.

BR Jerry

 

[hyleeinhit]

What do you mean by "input dependence"? I know The second circuit input node voltage swings, which the first circuit input node voltage fixes at "virtual ground" of OTA.

The first circuit is more popular that the second one. WHY?

At a quick glance I would say that in the second figure power consumption is lower and the input independences are different.

BR Jerry

 

[helpmejerry]

What do you mean by "input dependence"? I know The second circuit input node voltage swings, which the first circuit input node voltage fixes at "virtual ground" of OTA.

The first circuit is more popular that the second one. WHY?

Something to get you started https://en.wikipedia.org/wiki/Input_impedance

BR Uroš

 

[yuvan]

Hi,

I think the first of all the parasitics of the current dac or any other @ the input will have no effect due to the virtual ground effect of the integrator (ACtive integrator), which results in close to ideal integration, which in CTDSM is very important.
Secondly if you use the cap directly instead of the integrator then the voltage change at the input is very large which practically makes the DAC implementation very tuf, which is not the case due to the virtual ground effect of the active integrator. (And if the integrating voltage in the cap is very small the small signal error due to the parasitic cap of the integrator will cause a error of the form (1-Z-1), so may not be so severe if it exactly cancels out, but even small mismatch in the error will increase the noise floor in the band of interest)

 

[hyleeinhit]

Thank you a lot.

Something to get you started https://en.wikipedia.org/wiki/Input_impedance

BR Uroš

- - - Updated - - -

You gave really good points. Thank you!

Hi,

I think the first of all the parasitics of the current dac or any other @ the input will have no effect due to the virtual ground effect of the integrator (ACtive integrator), which results in close to ideal integration, which in CTDSM is very important.
Secondly if you use the cap directly instead of the integrator then the voltage change at the input is very large which practically makes the DAC implementation very tuf, which is not the case due to the virtual ground effect of the active integrator. (And if the integrating voltage in the cap is very small the small signal error due to the parasitic cap of the integrator will cause a error of the form (1-Z-1), so may not be so severe if it exactly cancels out, but even small mismatch in the error will increase the noise floor in the band of interest)