Incorrect Q value in HPF?

matthew fortin

In the FilterPro Desktop version, I am designing a 3rd order Chebyshev 0.5dB 135kHz High Pass Filter with a stopband freq of 10kHz. For the first stage, it lists a Q = 0.5. Doesn't Q=0.5 mean 2 repeated real poles at Fo? Is this mistake or is there a reasoning behind this? Shouldn't this stage be a single pole HPF?

Thanks for the support!

over 13 years ago

0 Geovid Kali over 13 years ago

TI__Intellectual 1245 points

Hello Matthew,

I recreated your filter in filter pro and I think I found the source of some confusion. In the topology I used (salen key chebyshev) the first stage of the filter is a unity gain buffer with a passive filter. Since the buffer is part of the first stage, the Q factor is affected. If the first stage was just the passive filter then the Q factor would be 1/sqrt(1) since there is one pole and the relationship between Q and poles is Q = 1/sqrt(# of poles). http://en.wikipedia.org/wiki/Elliptic_filter. Check under the BOM tab in filter pro and see what is included in "stage 1" for your filter.

Hope this helps,

Geovid Kali

0 matthew fortin over 13 years ago in reply to Geovid Kali

Prodigy 20 points

Hello Geovid,

Thanks for your reply. So if Q = 1/sqrt(# of poles), and the RC filter is 1 pole, that would mean the unity gain buffer has 3 poles? Is that correct?

Thanks,

Matt

0 Geovid Kali over 13 years ago in reply to matthew fortin

TI__Intellectual 1245 points

Hey Matt,

Yes, there is one pole in the RC filter. But the Q value for the unity gain buffer is not based on poles. The relationship for poles and filters is only for elliptical filters. For the Q value of the unity gain filter you have to look at energy stored vs powered loss http://en.wikipedia.org/wiki/Q_factor.

Regards,

Geovid

0 Geovid Kali over 13 years ago in reply to Geovid Kali

TI__Intellectual 1245 points

Sorry for the typo, poles and Q-values*

Simulation, hardware & system design tools

Simulation, hardware & system design tools forum

Incorrect Q value in HPF?