Pix, thanks for your answer; yes, indeed 90% of samples below 2% sounds good but my question is why if we have such an amount of good RxQual samples (this is measured after signal demodulation) later we got this different in the FER distribution(measured after de de-interleaved and Error Correction).
For me sounds logical to have, lets say, a % of RxQual samples below 3 of about 90% and have a % of FER samples below 2% of 98% since the Error correction mechanism could “recover” the information of the frames.
Regards,