From rec.radio.amateur.digital.misc 687
Path: ucbvax!agate!spool.mu.edu!sol.ctr.columbia.edu!math.ohio-state.edu!news.acns.nwu.edu!network.ucsd.edu!news-mail-gateway
From: ka7oei@uugate.wa7slg.ampr.ORG
Newsgroups: rec.radio.amateur.digital.misc
Subject: Detecting PSK... an easy way...
Message-ID: <34404@uugate.wa7slg.ampr.org>
Date: 20 Nov 93 16:57:16 GMT
Organization: ucsd usenet gateway
Lines: 22
NNTP-Posting-Host: ucsd.edu
Originator: daemon@ucsd.edu
The simplest way to detect PSK with the phase detector is to get rid
of the phase transistion. With BPSK, double the frequency: QPSK, quadruple
it, etc.
A dirt-cheap circuit to do that would be to build an amplifier for
your input audio (use an LM386) and get, say, an 8-1k audio transformer
with a centertapped 1k "secondary". On the secondary, just put a
full-wave rectifier (the 2 diode kind... The 4 diode kind will work,
too, if you want... You have suddenly doubled the frequency... and multiplied
the voltage, too, so you will probably have to attenuate it.
An advantage to this scheme is that it also rejects even-order multiplication
projects. It might be reasonable to have a 1st or 2nd order highpass filter
on this thing, but avoid too much filtering, since that will slow your
dcd...
All that is necessary (to have this work) is to have enough audio voltage to
drive the diodes well, hence the '386 amp.
You *could* build something with a costas loop, but...
<Clint>
From rec.radio.amateur.digital.misc 688
Path: ucbvax!agate!library.ucla.edu!europa.eng.gtefsd.com!uunet!news.sprintlink.net!demon!llondel.demon.co.uk!dave
From: dave@llondel.demon.co.uk (David Hough)
Newsgroups: rec.radio.amateur.digital.misc
Subject: Re: How to detect 2400bps 4DPSK?
Message-ID: <753814370snx@llondel.demon.co.uk>
Date: 20 Nov 93 09:52:50 GMT
References: <1993Nov20.023628.9192@frc.maf.govt.nz>
Sender: usenet@demon.co.uk
Reply-To: dave@llondel.demon.co.uk
Organization: last!
Lines: 40
X-Mailer: cppnews $Revision: 1.38 $
In article <1993Nov20.023628.9192@frc.maf.govt.nz> wk@frc.maf.govt.nz (Wilbert Knol) writes:
>I was wondering if some Knowledgeable Person on the Net could point me
>in the right direction. The HT I use with 2400 bps (terrestrial PSK)
>packet has a very slow squelch, and I have attempted to build a
>carrier detection circuit so as to run the radio unsquelched. This
>should cure the problem of loosing the first part of a packet.
>
>The circuit I tried is a 567 PLL, centred on 1800 Hz. It works
>fine...as long as the signals contains no phase jumps. The loop flies
>out of lock as soon as phase transitions occur, creating spikes on the
>PLL output.
>
>I have played with the loop band with and gain, all to no avail. Also,
>I tried feeding the PLL output through a low pass filter into a
>Schmitt trigger, which worked, but the DCD circuit became slower then
>my squelch...
>
>Rather than trying to re-invent the wheel, I was wondering how this is
>done in commercial PSK packet modems. Any pointers appreciated.
>
Based on similar problems at RF, you could try putting a x2 or x4 frequency
multiplier in the audio circuit and running your PLL at the higher
frequency.
The other thing to try if you are using a TAPR state machine eprom in your
TNC is to build the state-machine DCD-mod and see if that works. You might
need to adjust a few time constants from the 1200baud version but it ought
to work.
Note that I haven't tried either of the above, but they do seem like good
ideas :-)
Dave
--
*****************************************************************************
* G4WRW @ GB7WRW.#41.GBR.EU AX25 * Start at the beginning. Go on *
* dave@llondel.demon.co.uk Internet * until the end. Then stop. *
* g4wrw@g4wrw.ampr.org Amprnet * (the king to the white rabbit) *
*****************************************************************************