For a back-of-an-envelope calculation, you could use the Carson Bandwidth Rule which says that total bandwidth of an FM signal is twice the sum of the frequency deviation and the modulation frequency.
I don't know how applicable it would be for a linear sweep rather than sinusoidal modulation, but I guess it should be in the right sort of area.
sine sweep max speed?
-
- KVRian
- 799 posts since 25 Apr, 2011
-
- KVRist
- Topic Starter
- 64 posts since 31 May, 2016
hey,I was doing more research on this question and I am 99% percent sure that there isnt any "pure" sinewave sweep,becose the first moment we deviate from what is flat constant amplitude & freqency sine,its starts getting sidebands
I reed every form of modulation creates sidebands,sine sweep,be it amplitude sweep or freqency sweep I believe is considered modulation
well I have new question,Wavelets..... its like sinewave that that have fade-in at start (increasing amplitude ) and fadeout at end ( decreasing amplitude ),you can do wavelet transform with it to do spectral ( time&freqency ) analysis of signal just like short time-fft.
my question is this,since FFT correlates signal against sines and cosines of constant freqency and amplitude ( so they are "pure",no sidebands... ) and Wavelet transform correlates signal against wavelets,which are like short sinewave bursts with changing non-constant amplitude,so I asume they must be non "pure",they must have sidebands,so the wavelet transform then must have less precision when it comes to freqency?
I expect that if we try to measure freqency by correlation,pure sine would offer superior resolution than if we correlated with wavelet that is sine + another quieter sinewave of lower and higher freqency aka sidebands
I reed every form of modulation creates sidebands,sine sweep,be it amplitude sweep or freqency sweep I believe is considered modulation
well I have new question,Wavelets..... its like sinewave that that have fade-in at start (increasing amplitude ) and fadeout at end ( decreasing amplitude ),you can do wavelet transform with it to do spectral ( time&freqency ) analysis of signal just like short time-fft.
my question is this,since FFT correlates signal against sines and cosines of constant freqency and amplitude ( so they are "pure",no sidebands... ) and Wavelet transform correlates signal against wavelets,which are like short sinewave bursts with changing non-constant amplitude,so I asume they must be non "pure",they must have sidebands,so the wavelet transform then must have less precision when it comes to freqency?
I expect that if we try to measure freqency by correlation,pure sine would offer superior resolution than if we correlated with wavelet that is sine + another quieter sinewave of lower and higher freqency aka sidebands
