gm-C filter and nonlinearity

[wenadinho]

Hi,
as far as I know gm-C filter is an open-loop filter configuration, i.e. no DC feedback. Does this means that linearity become a problem for large input signal? The gain and cutoff freq itself is related to the accuracy of the gm of OTA, how accurate can the gm be? Does the open loop DC voltage gain of OTA needs to be very large (i.e. > 60 dB) or as long as we are able to implement the required transconductance, any DC gain would do?
Another non-related question is, can we implement second order filter, by cascading two first order filter? What are the disadvantages and advantages as compared to other second order filter topology (Sallen-Key, etc)
Thx very much

 

[pfd001]

Hi,
as far as I know gm-C filter is an open-loop filter configuration, i.e. no DC feedback. Does this means that linearity become a problem for large input signal? The gain and cutoff freq itself is related to the accuracy of the gm of OTA, how accurate can the gm be? Does the open loop DC voltage gain of OTA needs to be very large (i.e. > 60 dB) or as long as we are able to implement the required transconductance, any DC gain would do?
Another non-related question is, can we implement second order filter, by cascading two first order filter? What are the disadvantages and advantages as compared to other second order filter topology (Sallen-Key, etc)
Thx very much

Sure the linearity is a problem if a large input signal is used.
The DC gain of the gm cell will have a phase lead, and affect the Q of the filter.
Sure, You can implement the second order filter by cascading two first order filter