Tone Generator - Multiple Tones and Variable Amplitude

[percy]

still haven't tried LTspice or Scilab but ended up writing a script to generate my own sine wave! I even wrote my own function to calculate sine based on taylor series. Its interesting how much precision (atleast 7 digits after the decimal point for 16 bit depth) is required to avoid rounding errors and take the precision upto or less than 1LSB.

I have yet to add the file creation part to write to a pcm wav file. Once that is done I will know how much distortion there is in the produced wave, or lack thereoff.

In v2.0, complex waves - ability to add harmonics with variable amplitude!

I wonder if these packages, or the languages they are written in, have the kind of precision for these type of operations.

[AndrewT]

Originally posted by Pan
it's the other way around. A square wave is made up of lots of sinewaves.

not when the system (DAC) is trying to reconstruct an analogue signal (sinewave) from a digital datastream.

If you do breakdown a squarewave into it's components one finds that the harmonics go out to infinity. That is the problem. The digital datastream runs out of bits in the range of 44Ks/S to 192ks/S. I have seen it estimated that a good approximation to a squarewave can be reconstructed if all the harmonics up to the tenth are included in the signal. To reconstruct a reasonable 96kHz squarewave would require a bit rate of 1Ms/S or looked at from CD datarates it should be possible to get a 4kHz squarewave from the DAC. I suspect that even a 4kHz sinewave coming out of a CD DAC will show significant distortion.
Does anyone have any figures?

[percy]

6 digits of precision(i.e. after the decimal point) and processing the sine series(taylor polynomial) up to the 9th degree is what you need in the least to create a sine wave with distortion buried in the noise floor.
Note that I did not say "without distortion". It is there, just not discernable.

I had been fooling around with creating a wav file of a pure tone and observing what kind of loss of precision generates what kind of distortion and thought I'd share some results with you.

It is a common belief that computing upto the 7th degree of the Taylor series for a sine function is a good enough approximation for a sine function.
Wrong. You need alteast one more degree i.e. upto 9th degree to achieve a sine wave which has very little distortion.

Here's a screen shot of the FFT of a 500hz sine wave produced with the sine function computing only upto the 7th degree -

[percy]

and here's a screen shot of the FFT of a 500hz sine wave produced with the same sine function, but computing upto the 9th degree -


FFTs were done with the spectrum analyzer provided in RMAA 6 with a Kaiser window of atleast 15.

[Pjotr]

Originally posted by percy
I am searching for PC based tone generator that will allow adding multiple tones with variable amplitude. For example, 1Khz at -3db, 2Khz at -40db, etc...

Also it has to be capable of creating tones at 92Khz or higher also.

Right now I have the NCH Tone Generator that has multiple tones and variable amplitude but unfortunately its fixed at 44.1Khz so can't add any frequency above Fs/2.

Any help greatly appreciated!

TiA!

Hi,

FYI, have also a look here: http://www.fesb.hr/~mateljan/arta/

It works with 192 kHz soundcards as well. The build-in sine generator can do dual tones with adjustable ratio quite accurate. Attached a picture of a loop through measurement of my USB soundcard at 44.1 kHz.

[percy]

I don't think it can save it to a wav file, does it ?
I want a "generated" wav file, not a "recorded" wav file (to avoid introducing deficiences of the playback + recording chain).