Matlab Filter Design vs Stimulus/Probe show different results

Hello Andrew,

Sorry that no one has helped with your question. I'm not a Matlab user so I was hoping someone else would jump in here. Since that didn't happen I'll just share what I found and hope that helps.

Miguel who works with SigmaStudio development at ADI has commented on a similar issue here:

http://ez.analog.com/message/113428#113428

The ADAU1701's parameters can have 23 bits in their fractional parts, equivalent to approx. 6 decimal digits. Thus I suggest you have Matlab print your filter coefficients and gains, and enter them into your project, with a full six decimal places. Four places is plenty enough for typical parametric EQ filters, but your requirement appears to be a very sharp 8-pole filter with four high-Q sections. I'm not sure if the added resolution will help, but it's easy enough to try.

Best regards,

Bob

Hi all:

i'm sure this has to do with the spacing of the Frequencies being calculated by the Tool in simulation. Try slightly moving the center frequency and you will see that the peak happens at a completely different gain value. This doesn't mean that the gains are different, simply that the chosen (Simulation) points are not calculated at the exact frequency of interest.

Please let me know if this is not clear.

Miguel

Hi Miguel,

Thanks for the clarification!

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