- More (clueless) FFT tweakings
Replaced the Hanning Window
Applied dumb automatic scaling to the bars (based on max frequency)
220 ac from wall it very very dirt source ))
because of many consumers like flyback power source, gas discharge lamp, etc
…right but compared to the awful “sine” of the quad, it is almost perfect
Yes, but sampling it with 8 bits will get it just as awful
Wonder if you can look at this. I tried looking at the code, but can’t figure out where it is going wrong.
See this video that shows the GCC version in SCAN mode. Here you will see the waveform is printed in 3 distinct segments that repeat themselves.
This is another video with the AP1_100 version in SCAN mode with buffer set to 360(smallest setting). This problem does not happen here. Sad we don’t have the sources for AP1_100
Please let me know if I there is a way I can help. Right now, I cannot figure out the whole code, so, not sure where to look. But, I am guessing SYNCHRO and PROCESS are the right places to look.
…that means you can/will contribute to the project?
All the best,
I’m sorry, but I have my own project.
Picture order was to compare. In 220 no loud noise.
Do you have your source code available to the public?
About the internal generator
It is far from perfect. At first - the signal from the DAC is not filtered. Second - the output buffer amplifier goes to a limiting on the signal levels close to GND and Vss, it is noticeable even visually, and produce harmonic components.
There are other cases were the FFT seems to behave badly
(my guess, I’m really not an expert. This FFT stuff is out of my league)
, but under those same settings, the results seem similar:
Now, chip is probably using a higher FFT sample count (1024 perhaps), so the results in his firware will be more accurate.
Besides the bugs I’m sure my code has
I don’t understand.
Did you acquire a licence to use the original firmware source code?
I see only 2 options for the original source code licencing:
Yes, I use 1024 values of the signal and get the 256 points of the spectrum
I do not know what kind of license is needed to open source, but I know that with this code is a bad oscilloscope
Well, not just the code: The hardware also has some big flaws.
I don’t understand the behaviour of Pedro’s FFT.
I use a 10kHz “sine” from the internal DAC, channel A with 0.5V/div, 100uS/div as an example.
Maybe this has to do with the fact that only 256 samples are used, i.e. only about 10 periods and this is not enough - does the FFT only give good results when there is an integer number of periods in the 256 samples? Could this explain why the sines have so many harmonics?
I had another problem, which looks like a bug: when moving the “navigator B”, changing between the “trigger” and the “auto/norm…” menus - but not changing anything at the settings - sometimes the FFT result change completely!!!
Two more questions:
The number of samples influences the resolution (frequency / x scale) of the FFT.
Can’t increase the number of samples at the moment because of the increase in program size/storage needs: