ΣSigmaStudio Ask AI Русский Log in

Q value of filter

SigmaDSP & SigmaStudio Philip Mar 30, 2011 00:00 5 replies Original on forum ↗
Subscribe
Philip Question
Mar 30, 2011 00:00

I am playing with the second order general filter in my adau1701.

I generate a noice input from PC, passing though the 1701 and back to the PC. Then I plot the freq response in the PC

I use the sigma studio to calculate two set of filter coeff

1) freq=15000, gain=-16, q=1.5 (as sigma_studio setting.jpg, correct me is I am wrong)

2) freq=1000, gain=-16, q=1.5

And I have two question about that:

1) By looking at the 1k filter freq response, the peak and the freq cutoff relation is not matching with the normal filter relation.

I use the calculator(http://www.sengpielaudio.com/calculator-cutoffFrequencies.htm) and get the high freq cut off should be at around 1.3kHz while I end up with the high freq cut off at 1.1kHz

2) By comparing the 1k filter and 15k filter with the same q, it seem the 15k filter is more narrow then the 1k filter. Is that a reason behind?

I am very new to those filter's, correct me if I doing somethings that didn't make sence

Audio Signal Processors

SigmaDSP Audio Processors

adau1701

ADIApproved Reply
Mar 30, 2011 00:00

I will answer question (1) tomorrow or Friday.

For question (2): Since the filter is constant-Q, that means it is variable bandwidth. The bandwidth is dependent on the center frequency. As the center frequency increases, the bandwidth will decrease. You can verify this using the calculator you referred to on your external link.

Philip Reply
Apr 8, 2011 00:00

Looking forward for your answer, thanks

ADIApproved Reply
Apr 13, 2011 00:00

Hi Philip,

I gave this a try today. I used filters at 1 kHz, 2 kHz, 4 kHz, 8 kHz, and 16 kHz with a Q of 1.41.

I tested on an evaluation board with S/PDIF input and output.

You can see from my results that the bandwiths match up perfectly for 1 kHz through 8 kHz. The one at 16 kHz seems a little off, but I don't have time to look into that anymore at the moment. Let me know what you think.

6aab84c445795c589a67883b544af63f.jpg
6aab84c445795c589a67883b544af63f.jpg
6aab84c445795c589a67883b544af63f.jpg-450x267.jpg
6aab84c445795c589a67883b544af63f.jpg-450x267.jpg
344bc054c96df88305c0b4c6b3e674e0.jpg
344bc054c96df88305c0b4c6b3e674e0.jpg
344bc054c96df88305c0b4c6b3e674e0.jpg-450x267.jpg
344bc054c96df88305c0b4c6b3e674e0.jpg-450x267.jpg
8b855fd19aec461389305debf2ec9903.jpg
8b855fd19aec461389305debf2ec9903.jpg
8b855fd19aec461389305debf2ec9903.jpg-450x267.jpg
8b855fd19aec461389305debf2ec9903.jpg-450x267.jpg
f670bee01cdd92e3804d710e3e544d37.jpg
f670bee01cdd92e3804d710e3e544d37.jpg
f670bee01cdd92e3804d710e3e544d37.jpg-450x266.jpg
f670bee01cdd92e3804d710e3e544d37.jpg-450x266.jpg
81bdd6436ec560acd7e5d9238fcb7635.jpg
81bdd6436ec560acd7e5d9238fcb7635.jpg
81bdd6436ec560acd7e5d9238fcb7635.jpg-450x267.jpg
81bdd6436ec560acd7e5d9238fcb7635.jpg-450x267.jpg
ADIApproved Reply
Apr 13, 2011 00:00

Philip, actually I think I know why the 16 kHz filter is off. According to the calculation, the upper limit of the bandwidth should be above the Nyquist rate of 22 kHz, which is impossible. I guess for that reason, the BW gets shrunken a little bit near the high end of the spectrum.

EZAdministrator Reply
Aug 2, 2018 00:00

This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

Thank you,
EZ Admin