glias
Full Member level 2
Hello all,
I permit to "re"post and try again to explain my problem :
I use a IR sensor that provides a voltage response on the output. This output is pertubate by a regulation loop from a compressor (temperature regulation with a diode placed in the sensor). A sort of “glitch” is present on the output of the sensor (come from the PWM command of the compressor motor). The bandwidth of the sensor is 314kHz and the PWM is about 60kHz.
My question is : is it possible to filter this perturbation with a digital filter ?
I don't think it is possible because if I use a notch filter for example to filter the perturbation I will loose information of my sensor ...
consider the 2sd proposition :
I can predict the apparition of the perturbation because the PWM is stable in frequency and I can capture frequencies components of it by a FFT, then, if I consider that I can have the FFT of this perturbation, is it possible to substract it to the signal at the PWM frequency (period of the apparition of the perturbation) ?
what do you think of this method ?
I hope that you could help me...
thanks a lot.
PS : I have a ADC @10MHz abnd a FPGA to do the processing.
I permit to "re"post and try again to explain my problem :
I use a IR sensor that provides a voltage response on the output. This output is pertubate by a regulation loop from a compressor (temperature regulation with a diode placed in the sensor). A sort of “glitch” is present on the output of the sensor (come from the PWM command of the compressor motor). The bandwidth of the sensor is 314kHz and the PWM is about 60kHz.
My question is : is it possible to filter this perturbation with a digital filter ?
I don't think it is possible because if I use a notch filter for example to filter the perturbation I will loose information of my sensor ...
consider the 2sd proposition :
I can predict the apparition of the perturbation because the PWM is stable in frequency and I can capture frequencies components of it by a FFT, then, if I consider that I can have the FFT of this perturbation, is it possible to substract it to the signal at the PWM frequency (period of the apparition of the perturbation) ?
what do you think of this method ?
I hope that you could help me...
thanks a lot.
PS : I have a ADC @10MHz abnd a FPGA to do the processing.