Other Parts Discussed in Thread: TAS5711,
Hello, engineers!
I am designing a 2.1 digital audio amplifier. By now I had used TAS5711/31 GDE to calculate the BQ4a coefficients. Now I need to calculate it on my own. I read the treads here and came to a solution. However, there is a "slight" difference between what I had calculated and the GDE values.
The filter is Low-pass, Fc=800Hz, Fs=48kHz, Butterworth 2-order. Values are as follows:
GDE: 00 00 53 93 00 00 A7 26 00 00 53 93 00 ED 12 81 0F 91 9F 32
My value: 00 00 53 93 00 00 A7 26 00 00 53 93 00 ED 12 81 03 91 9F 31
| b0 | 0.002551 |
| b1 | 0.005101 |
| b2 | 0.002551 |
| a1 | -1.85213 |
| a2 | 0.862329 |
I used the formulas of "Bi-quad filters realization on AIC Codecs" and "Cookbook formulae for audio EQ biquad filter coefficients" - results for an and bn are equal.
For bn values I'm using this in Excel: DEC2HEX(IF(bn <0,(bn *2^23+2^26),bn *2^23),8)
For an values I'm using this in Excel:DEC2HEX(IF(-an <0,(-an *2^23+2^26),-an *2^23),8)
I found this expressions in this forum in a tread describing a notch filter.
How is it possible to have almost all values equal except two bytes about a2? Where is the mistake?
Best regards,
Svetoslav Valkov
