Is there a way to save the analog signals (the waveform) from the serial plotter to create my own library of signals from my sensor. If there is a way, what format will it be saved? Can I use that format for signal analysis in Matlab?
You can use Serial Monitor. Just use cut and paste the data into whatever You like.
Save to an SD card
Use a sspreadsheet To make graphs.
Analog signals are converted to digital so the serial outtput is just a cotrivance and not the actual thing.
If you want higher resolution then google sound card oscilloscope
While the serial monitor in the Arduino IDE doesn't have the ability to log output to a file there are any number of alternative serial monitors that do. You could do a web search for "serial terminal emulator ". I've used "minicom" on Linux to do this sort of thing and "realterm" on Windows.
Alternatively one can write a program in their preferred programming language to receive data from the PC serial port. There's a Python example in this thread: Demo of PC-Arduino comms using Python - Interfacing w/ Software on the Computer - Arduino Forum. I'm pretty sure that this can be coded directly in Matlab, though I've never done so.
PLX-DAQ version 2 - now with 64 bit support! (and further new features) - Interfacing w/ Software on the Computer - Arduino Forum Works pretty good.
What's the speed/frequency of the analog signals? (The Arduino may not be fast enough and serial transmission may not be fast enough if you need to capture & transmit in real-time.)
[u]Here[/u] is a little tutorial about how analog signals are digitized. It's about digitizing audio but the same concepts apply when digitizing any analog signals.
If you need frequency information, the most important thing is the sample rate (well, one of the most important things).
The sample rate has to be known if you want to "connect the dots" to re-construct the continuous waveform (with a digital-to-analog converter) or to analyze the frequency content.
And, the sample rate must be at least twice the signal frequency (the "Nyquist limit.) That's easy to understand that because you need at least one sample for the top-half of the cycle and at least one sample for the bottom half. If you exceed the Nyquist limit you get aliasing (false frequencies). If there's a chance of that happening you need an analog filter in front of the ADC. (You can't filter digitally because the data is screwed-up immediately when it's being digitized.)
If you just need amplitude information the sample rate is not as critical, but of course you have to read fast-enough to read the changes in amplitude.