Re: Design of High Frequency Integrated Analogue Filters (Y.
tok47 said:
Dear DoctorX,
firstly,thank for all your help.
secondly, i attach here a circuit. I have been trying to understand how is it functioning for some time. can some one kindly help me on this?
i found it in the Kansas State University
h**p://www.eece.ksu.edu/vlsi/bluetooth/
this is a good references site.
tok
N.P. First off, each block in the figure is a differential Gm (transconductor) cell with limited and controlled output resistance (the 50 kohm, etc). With the load capacitors, they are either lossy integrators (1st-order lowpass filter) or lossy gyrators (active inductors).
Block2 and Block3 are presumably a gyrator, however, it is postive feedback. Not sure if it is a design error or on purpose (to derail imitators maybe). Anyway, if the input is at Gyrator node 1, then the other node will be lowpass. For more explanation on this, please look at:
**broken link removed**
Similarly, Block 5 and Block 4 are another gyrator (with positive feedback unfortunately), acting as a lowpass filter. There is a wrong connection.
Thus, you get a 5th order lowpass filter by cascading the 1st-order LP, 2nd-order LP, and 2nd-Order LP. By properly choosing the Q's and F0's, it can be a Maximally-Flat filter, or Butterworth, when the cutoff frequency is defined at -3-dB.
I recommend you draw a block diagram of differential Gms, and trace it with the node notations. This is how I found all of the errors. I modified the circuit, and attached it here.