Gain values reported are wrong for high-pass MFB filters >2 poles. For exmaple, a G=50 4-pole design reports each 2-pole stage gain as 7.071 where as it is far different than that. There does not appear to be a pattern to the actual gains vs the reported gains. Very confusing and disappinting.
Harvey,
You may need to give me a more specific example if gains are being reported wrong. But if you have two stages, each with a gain of 7.071, then the overall gain of the filter is 7.071*7.071 = 49.999041, which I'd say is pretty close to 50. FilterPro generally tries to evenly distribute gain amongst stages (an observation, I don't recall the actual algorithm, it may also factor Q in there where doing gain distribution).
If you can share a design with me that shows the wrong behavior, I'd like to investigate.
Rick,
Attached is a design for a filter that is High-pass, 800 Hz, Gain 50, MFB, 4 poles (2 stages). Look at the synthesized resistor values and calculate the stage gains. They are not .7071, and the product does not equal 50. I did the same design at 6 poles, 3 stages, and on that one at least the product of stage gains is correct but the stage gains are reported wrong.
Obviously the stage gains and total gain could be reported from the synthesized values. Don't know where the values shown came from ?? Something is drastically wrong in that part of the synthesizer.
Thanks, Harvey
Short of doing the math to find the transfer functions of each stage (which I'll leave to you), I built the circuit in TINA-TI, which you can download from http://www.ti.com/tina-ti. I've attached the TINA circuit:
2781.hpfilterg50.tsc
Using real op amps, you don't quite get the gain of 50, but it's close:
Using real op amps instead of ideal (and I just chose a FDA sort of at random, so may not be the ideal choice), there is a gain error, but you'll note that each stage has a gain of about 6.6. Is that "drastically wrong"? Perhaps, but some tuning is often needed and that's where TINA-TI can be a big help.