single or double precision filter

Sorry.

Whether it is 1701 or 1452 have a similar problem.

Thanks,

Anthony

Hi,

It could be a bit related to this problem too:

ADAU1452 T-Connection, Readback, Level Detector

Best regards,

Simon

Hi Anthony,

I've replicated the issue and will be talking to our software team about it. I'll update you with our findings.

Thanks,

Tyler

Hi Anthony,

It looks like the issue is related to quantization error introduced for low-frequency signals. This is an inherent issue with using a single-precision filter, and one of the reasons why the double-precision is also available. A good rule of thumb is that for filters near or below 0.5% of Fs, you should use a double-precision filter.

Regards,

Tyler

Hi Tyler,

I'm using the double-precision filter now. I just want to know more about the reasons for the problem.

Thanks,

Anthony

HI Anthony:

The reason is the quantization of intermediate calculations due to the 5.23 fixed point format used inside the DSP. Using double precision doubles your fractional precision to 46 bits.

Hope this helps.

Miguel

Hi,

Thanks Tyler and Miguel.

Thanks,

Anthony