I am examining a circuit from the Linn turntable Vahalla power supply where it use a crystal to generate a square wave at 3.2768MHZ, divide it by 65536 using counter and get a 50Hz square wave. From this it is feed to a five pole Sallen Key Low Pass filter to get a sine wave and then power amp it to feed to the AC motor. All looks good except when examining the LP filter I found the pole frequency is at something straight value and I am not sure why it is so.

Real pole: R=56K C=100n F0=28.42Hz
Section A: R1=82K, R2=8K2; C1=100N, C2=100N, F0=61.38Hz
Section B: R1=36K, R2=39K; C1=100N, C2=100N, F0=42.48Hz

I suppose we should be applying a LP filter with F0=50Hz to get a clean sine wave shaped from the 50Hz square wave??

I am try to use the same Sallen Key filter setup to shape both 67.5Hz square wave as well as 50Hz square wave into sine wave for driving the turntable for different rotational speed but with this figure I am puzzling should I use a 5 pole filter all with the same pole frequency or using different pole frequency at different stage, just like this Vahalla setup. But if using different pole frequency what should be adopted??

Other point I wondering is on these op amp the feed back circuit R3/R4 is isolated from the ground via cap. Any particular reason?

 

Instead of fiddling around with hundreds of parts to make a 5-pole Sallen and Key lowpass filter, why not use a 5-pole or 8-pole switched capacitor lowpass filter IC?
National Semi used to make some good ones. Now Maxim-IC make them.
Feed in a square-wave and out comes a sine-wave. The filter iC needs a clock that is 100 times higher and the output has a small amount of clock frequency that is easily removed with a resistor-capacitor filter. Some switched capacitor lowpass filter ICs have a built-in clock oscillator.