Tell me about sampling interpolation in DSPs

Sampling is used to convert an analog signal into a digital signal to facilitate digital signal processing.

Interpolation is generally because the previous decimation is done to reduce the digital signal rate to facilitate DSP processing, and after the extraction is done in order to recover the original signal, it is necessary to do interpolation.

Sampling is done to analyze power problems.

porcupine^o^ 22:02:11

Isn't the white spectrum convolution arbitrary signal still the white spectrum?

The process of sampling is actually the process of converting the electrical signal of a normal analog audio signal into binary codes 0 and 1, which form a digital audio file. The higher the frequency of the sample, the more *** the sound quality. Since the sampling frequency must be twice as high as the highest frequency of the recording to avoid distortion, and the human hearing range is 20Hz 20kHz, the sampling frequency must be at least 20K 2=40kHz to ensure that there is no low-frequency distortion, which is why the CD sound quality is slightly higher than 40kHz to leave room for it).

DSP: Digital Signal Processing. DSP technology has a wide range of applications in analog electronics such as tone control, distortion effects, WAH WAH pedals, etc.

At the same time, DSP can also be used for a variety of effects such as analog equalization and reverb.

Usually the analog value between two points of the sampling point is set to a constant value, so the faster the sampling frequency, the closer the sampled value is to the simulated value! And the sample clock frequency is the clock cycle of the processor! The sampling window is the selected sampling interval!

The sampling frequency and the sampling clock frequency are the same thing. Because the sampling time cannot be infinitely long, it can only be taken for a short period of time. It's like looking through a window of time.

The key to realize the sampling rate conversion lies in the design of an efficient digital blind filter to filter the input sequence in a low-pass manner, so the selection of a filter with simple structure and efficient processing is the primary problem for the design of the system. An integral comb filter is sufficient.

If possible, you can use interpolation to make up points. Because as long as the three points can be taken, they can be well restored after its interpolation operation. Generally, you can use the method of adding a filter loop (adding capacitance, resistance or inductance) to achieve good results.

That is to say, if your sampling rate is greater than 20K, obviously, the waveform you see is not like a sine wave at all, but if you do FFT, you will find that all the information of the 10K signal is not lost. If you want, you can spread the data and use a higher rate to enter the low pass, and you can restore it to a smooth sine wave that you like. If digital signal processing is not deep, DSP is actually no different from MCU in your eyes

This depends on the real-time requirements, the buffer method is suitable for low-speed processing of slow-changing signals, if the signal changes very quickly, the buffer capacity may not be enough, and there are real-time requirements when you must use high-speed processors such as DSP.