FIR Filters on the STM32G4 Part 1

[@ytmoshte]

The only bad thing about this video is that it ends. Great job!

[@fano72]

I wonder about the performance boost of using CMSIS against a native convolution loop?

[@hamidk4772]

Outstanding Job.

[@ArsalanAhmadi-c8x]

Excellent, thank you. 🙏🙏

[@seyedsajjadmirbakht278]

It's a very great video, however, I'm having trouble understanding why you filter the first and second half of the ADC data separately? Why can't we receive data, filter it, and push it to the DAC?

[@Drizzt1343]

It's a way to double-buffer the data, so that the filter code is only ever operating on data that isn't being changed by the DMA.

[@L2.Lagrange]

I am very interested in this topic. Thank you

[@JorgeSilva-em8pf]

Dear Sir. I would suggest you to show the mechanics of generator & Spectrum Analyser (extra video) ; this last made around one digital oscilloscope.

[@markday3145]

I'm curious why you set up the DMA to transfer words instead of half words from ADC and to DAC. Seems like half words is already more precision than the ADC/DAC can give you (or the same as ADC with oversampling). Would it make sense to subtract the DC offset from the input data, so that you have signed values (with no offset) being passed to the filter? Would that change the behavior of the filter?

[@Han-EEE3]

Dear sir. Since the FIR introduce the group delay, may I ask how the DMA call back function solve this delay. Why it still works

[@AmitSharma-bm2qj]

Why don't you use FMAC of STM32G4?

[@RunDong-b4u]

it's too beneficial !

[@sergbo3114]

I don't think filtering in the interrupt handler is a good idea. Better would be to set some flag or use rtos events and tasks.