engrMunna
Advanced Member level 4
Hi,
I am building a matlab model of a RF Transmitter. I upsample the baseband IQ data using the matlab function 'interpft", by a factor of (carrierFrequency / IQbandwidth ).
And then modulate the carrier using this upsampled data.
For deomdulation I first downconvert the the modulated signal by multiplying it with exp(-j2π*carrierFrequency*T). And then low pass filter it using the comand "filtfilt" for zero phase filtering.
But this low pass filtering causes some errors in the demodulated Amplitude of the carrier (probably due to the ringing step response of the low pass filter)
and this causes increase in my EVM.
This is an ideal system, yet still the EVM is around -40dB . the test signal I am using is a 20MHz 64-QAM signal. In my view the problem is due to the ringing in the step response.
Any ideas for ideal - demoulation??
I am building a matlab model of a RF Transmitter. I upsample the baseband IQ data using the matlab function 'interpft", by a factor of (carrierFrequency / IQbandwidth ).
And then modulate the carrier using this upsampled data.
For deomdulation I first downconvert the the modulated signal by multiplying it with exp(-j2π*carrierFrequency*T). And then low pass filter it using the comand "filtfilt" for zero phase filtering.
But this low pass filtering causes some errors in the demodulated Amplitude of the carrier (probably due to the ringing step response of the low pass filter)
and this causes increase in my EVM.
This is an ideal system, yet still the EVM is around -40dB . the test signal I am using is a 20MHz 64-QAM signal. In my view the problem is due to the ringing in the step response.
Any ideas for ideal - demoulation??
