• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 47 subscribers
  • Views 405 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

ADCPro FFT with dBFS scale: Bug with the vertical position of the diagram?

ADCPRO: any known bug with dBFS scaling with ADS1282EVM?

Martí Bassas
Prodigy 140 points

Part Number: ADCPRO
Other Parts Discussed in Thread: ADS1282

Hello,

I've been gathering some data using ADCpro and a ADS1282EVM board.

I also process data with matLab to ease comparing spectra. I think I'm finding some lack of coherence. The two spectra below are supposed to be from the same data :

To scale according to dBFS I just divided all data by « Max Code » included in the data text file generated by ADCPro.

I’ve tried a few different PSD (power spectral density) calculation techniques in MatLab and the dBFS noise floor is always at the same place, like 30dB lower than what is indicated by ADCPro.

Reading older posts related to this kind of problems I found :

https://e2e.ti.com/support/data-converters/f/73/p/258714/904881?tisearch=e2e-sitesearch&keymatch=ads1282evm%25252520adcpro%25252520dbfs#904881

and

https://e2e.ti.com/support/data-converters/f/73/t/94449?tisearch=e2e-sitesearch&keymatch=%2520user%253A65187

The problem those users found could be the same I see.

In both cases there seems to be a possible bug in Y-scaling recognized from TI side, but in both cases the thread ends without a clear solution or work-around (as far as I can understand). Both are quite old.

Do you have updates on these topics ?

Would it be possible to have a detailed description on how dBFS values are calculated ?

  • 0
    TI__Mastermind 44780 points

    Hi Martí,

    Unfortunately, the images you shared did not appear correctly in your post. Would it be possible to share them and the ADCPro data file via email? You can send them to them to pa_deltasigma_apps@ti.com.

    I'll see if we can find information on how ADCPro scales the FFT. However, since the original developer is not longer in our group, we may only be able to confirm if the FFT results should match your MATLAB results.

    For the ADS1282, I know the output data format can change based on whether you are using the FIR filter or SINC filter. Which filter mode are you currently using?