[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

Re: P3D test transmissions

Dquagliana@aol.com wrote:
> In a message dated 98-11-06 12:11:58 EST, you write:
> > By the way, i 've written a P3D pskdemodulator for sound blaster that
> >  works very well. !, No need of a DSP to do the job.
> >  --
> >  ----
> Well, yes.  This is also what I am working on.  My modem
> also uses a soundcard and the "DSP" is done by a Pentium.
> I would be *very* interested in hearing more about how you
> implemented your modem.
> What techniques did you use? DPSK? Phase locked loops?
> How well does it perform?
> Douglas KA2UPW
> dquagliana@aol.com
> Robson Gomes wrote:
> > 
> > From: Robson Gomes@MLX on 09/11/98 10:22
> > 
> > >Just go to amsat ftp site and doxnload p3dtest.zip. This will be just
> > >fine for your requirements.
> > >By the way, i 've written a P3D pskdemodulator for sound blaster that
> > >works very well. !, No need of a DSP to do the job.
> > >--
> > 
> > Dear Ruy
> > 
> > I have looked on your page and could not find
> > the P3D pskdemodulator for sound blaster. I could only
> > find a test set to use the sound blaster to generate
> > psk modulated audio input to a psk modem.
> > Is this set up that you refered as a P3D pskdemodulator ?
> > 
> > Thanks for your information.
> > 
> > 73 de Robson - PY1DGV
> Hello Robson,
> thank for your interest. There are in fact two programs :
> - the first one is p3dtst.zip (or wathever name that kb5mu gave it on the
> amsat ftp). This pgm generates fully compatible p3d telemetry psk signal
> using your SB card. 
> - the second, which is currently NOT available is pskdec.exe. This
> program enables p3d telemetry psk signal demodulation without using a
> G3RUH demod. and uses the SB  CARD as an A/D converter. However, it is
> NOT real time for the moment i.e. it works only with recorded files
> (using the window$ recorder for instance). In other words, you could
> demodulate p3d tlm signals by first recording them on your HD, and
> process the resulting .wav file with pskdec.exe and get the TLM bytes.
> The main reason for this fact is that i merely wrote the programs for other
> purposes than p3d. (I am working on the DSP of the amsat_f
> satellite). 
> Hence you can now understand why there are two pgm : a 'tx' and a 'rx'.
> Tests i have made showed good results with the program 'as is'.
> Second : i have no time to make it real time, i.e. using all the PC
> interupt routines ... make a beautiful display, file management and my
> inability (and repulsion) to pgm under win$ completes the reasons why it
> stands as is. (i.e. dirty)
> However, i joined the pgm pskdec.exe to this mail so that you can play
> with it.
> To use it : 
> download the ZIP file on the amsat ftp site.
> Use it to generate a file EXACTLY as INDICATED in the readme text in the
> zip file.
> Then, you have two options :
> 1. run pskdec.exe nameofthefilegenerated.wav and input gain:20 f:1260.
> If all is OK, you should see on screen the txt you have input in the
> generating process. (Choose a txt of less than 1024 bytes to keep a
> reasonable size for the generated file)
> 2. You are a St Thomas and need more, then :
> Play the generated .wav file with windows mediaplayer and record it on a
> K7 recorder.
> Play back the tape and record it with the windows recorder and save the
> file .(PCM,8bits, mono, unsigned)
> Run pskdec nameofthefileyousaved.wav using the same input as in 1.
> It should work. (At least it worked with f6bvp, db2os and me !!!).
> For your tech information : pskdec.exe is made up of a quadrature digital
> vco, a Costas loop, and is featured with full doppler tracking and a +- 40Hz
> lock range. For bit clock recovery, the modulus of the complex signal is computed
> afterwhat it is fed to a 200 coef bandpass FIR
> filter centered on 800Hz with a -3dB bandpass of 7Hz. The data filters
> are made of a total of 37 coef, with 1/sinc compensation followed by S&H.
> The purpose of pskdec.exe is mainly experimentation, so DO NOT expect
> fantastic results.
> You may distibute it but i ask you TO MENTION THE ORIGIN OF THE SOFTWARE
> and to join that mail with it. 
> Kenavo.
> -- 

Hello Douglas,

You will find a copy of a mail i sent to PY1DGV yesterday. I send you a
copy of the program in an another mail. I added a copy of the mail to
amsat-bb because it seems to be of general interest.
To evaluate the performance vs noise level, you can alter the noise
level input in psk.exe, but WITHOUT EXCEEDING 53 for the sum of the
signal and the noise. Not doing so may exceed the allowed values of the
samples (0-255). Try s=28, n=20, and listen to the signal using your
sound card. You may use use COOLEDIT to alter signal strenght and
simulate fading for example. However, with n=0 you can specify a signal
of 128. This is because of the probability of exceeding the max value
due to the added gaussian noise.
Pskdec.exe shows you an 'approximation' of Eb/No and correlated BER.
If you are a good window$ programmer (C, not visual basic), we could
arrange a cooperation in order to provide a general interest program to
Ham sat community.
/*Ghislain et Nelly RUY 
/*F1HDD/ON1RG  - AMSAT_F #26
/*19 rue EDGAR FAURE 75015 PARIS

Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org