Spectrum analyzer

[Bob]

A spectrum analyzer breaks a complex signal into frequency components and displays the results graphically, to give a familiar representation in the frequency domain of a time-domain signal. Audio spectrum analyzers work in the range of audible sound while Radio frequency spectrum analyzers work at radio frequencies but costs much more. At home I use one for free which use the computer and the PC sound card. Fully tested in my small notebook to have a graphical idea about clicks generations from my RTX, by the the audio from a second receiver tuned on the same frequency. It can't beat the latest professional analyzer, like Aaronia and/or many others, b.t.w. can be used for free at home as the licence stands. The name is RMAA, RightMark(R) Audio Analyzer version 6.xx available for free at http://audio.rightmark.org/ , as far as I know it offers an advanced WAV spectral analyzer in several WAV sampling speed/bit, scale LOG/LINEAR/MEL, different FFT sampling etc. It's very easy to use by recorded WAV audio, and I got an idea about what happens, as different CW speed as about the CW rise and fall time effect analysis, and so on. Mmhhh, it's time to put hand again on the old iron solder, I leaved it unused to much time...
73 Bob

[Chuck aa0hw]

Excellent program Bob !
Thanks for bringing this to our attention and it puts out a great graph of the cw frequency spectrum on your Yaesu rig.

I was wondering if you might tell us how you set this up to help us step by step to do this too...
Also, how did you put your rigs data and info on the picture you provide of your Yaesu FT 1000 spectrum photo.

Thanks again for posting this.
Looks like a "must have" free program for the shack.

Chuck

[Bob]

Hi Chuck,
I'm pretty sure that an Audio with both RF Spectrum Analyzer is a dream in many ham activity. Could be a trick to use another Receiver as the front end and so work around in the Audio Frequency spectrum only?

To do that trick, limited in frequency width, simply tune the second receiver at the same frequency of the transmitter, the output audio of receiver put in the line or mic input of the personal PC.
Tune it very close and 1000 Hz pitch for example; 20 minutes of warm up both rig, or even more, is better.
Use the wide, flat filter, available i.e. 2.3-2.5 KHz, then shut down any other DSP or NR; reduce RF gain and if possible select AGC off.

The RF TX is closed to a dummy load, the receiver near by may need some piece of antenna. I used a screwdriver to get a reliable signal since the MFJ-986 internal dummy load is shielded enough at 3.5 MHz.

I do believe the audio spectrum can not be better than the RF real one, may be worse due the non linearity and distorsion introduced by the receiver and propagation media.
Anyway nothing better inside the receiver bandwidth, or I hope so, this is the functional idea.

I dunno how much is representative about the reality, as could be enough for ham purposes, mainly to got qualitative idea about what happens, some doubt about dynamic range or something not foreseen.

B.t.w. almost amazing since all is working simple, even too much, easy and fast in a small Eee-PC netbook like with WinXP OS.
Step by step:

- Install RightMark Audio Analyzer 6.x.x

Have a previous recorded audio file to analyze, or make one with 44.1 KHz 16 bit sample data, as described above, I used a preformatted text as " the quick brown fox jumps over a lazy dog 12345 67890 " by your device or the keyboard, with the speed you like for test, at faster speed retry message twice or three time to get a file longer than one minute.
For the wav recording I used an old Wave Editor, I guess any recorder is the same, the one inside RMAA also, but I didn't try with it.

- Launch RMAA program;
- Select the Icon "Spectrum Analyzer" from the RMAA panel;
- Program ask for file to open, then select your own [.wav] to analyze;
- Spectrum analysis options asks: I select FFT size 262144 then go ok;

- RMAA program start FFT analysis and a window with audio spectrum is presented;
- Select "linear" presentation instead the "log" by default (MEL is also possible);
- Tune the scale of frequency cursor to got your pitch frequency at centrum, play with +/- button to have the selected bandwidth presentation;
- Adjust the level also with +/- buttons to keep full spectrum inside the selected bandwidth and level scale,
- You may adjust better with the offset graph adding a bias in dB, by spectrum windows options now available, just below the graph, third icon.
- Adjust the size of spectrum window before select the save as button (floppy icon), then a spectrum.png is done.

May further add data on the .png spectrum by Paint or any other graphical editor, I add rise and fall time text as reminder.

In this manner I got my FT-1000MP spectrum by the 2nd FT-847 as receiver (I used a capture area with another viewer editor since I didn't yet know that print facility!).
For what I got and saw, I finally decide to put hands inside radio; too many clicks, this is a well known problem in the original 1000MP.

I got a funny plaisure as compare different CW speed and realize as the whole clicks spectrum is about the same, since it mainly depends from the rise an fall time of the CW envelope. Only a very small speed interaction, due to circuitry inside the transmitter, and needs to further investigate.

One thing is to known about technical readings, all another thing is to see it in the reality or near that, into a quite graphical tangible image, made by you self.

Wow a lot of words, QRU now.

Hope helpfull to begin, any way QRV here for the bad grammar or anything I've wrong and missed to say.

God Bless, Keep the good work and have a lot of fun in QRQcw.
73 Bob

[Chuck aa0hw]

Thanks Bob,
Very good instructions.
I was able to setup FLdigi to send a stream of dits at 100wpm using the BLACKMAN raised cosine waveform, and thanks to your step by step tutorial, I was able to get this graph of the FLdigi wave file: