| |
EB2DVY > PSK31 07.07.99 02:58l 166 Lines 7080 Bytes #-9975 (0) @ WW.SYD.NSW.
BID : 119-EB2DVY
Read: OE1RRB GUEST
Subj: PSK31 Digest V1 #91
Path: OE1XAB<OE3XPR<HA1VH<HG1PSO<OE3XKR<OE1XIB<OE3XBS<DB0WGS<DB0PV<DB0AAB<
DB0SIF<DB0GV<DB0GE<F6KFG<DB0CZ<HB9EAS<HB9OS<HB9H<HB9OK<IW2FPO<IK1MVX<
TK5KP<F6KED<F5KPO<F5KAT<EA3D<EA3B<ED3ZAF<EA3RKT<EA3ARR<EA5VDR<EA2BBS<
EA2C<EA2N<EA1URR<EA1URM<EA2G<EA2URV
Sent: 990629/2153Z @:EA2URV.EABI.ESP.EU #:59283 [URV-ABRA Bilbao] $:119-EB2DVY
From: EB2DVY@EA2URV.EABI.ESP.EU
To : PSK31@WW.SYD.NSW.AUS.OC
De: Psk31 Digest <owner-psk31@aintel.bi.ehu.es>
Para: <psk31-digest@aintel.bi.ehu.es>
Asunto: Psk31 Digest V1 #91
Fecha: lunes 28 de junio de 1999 4:22
Psk31 Digest Monday, June 28 1999 Volume 01 : Number 091
In this issue:
[psk31] FD/PSK
[psk31] FSK31 explained
[psk31] R/S Gnd loop isolator
See the end of the digest for information about psk31-digest
----------------------------------------------------------------------
Date: Sat, 26 Jun 1999 21:26:56 -0700
From: Bill Brannick <WmLB@prefer.net>
Subject: [psk31] FD/PSK
Well I have managed to pick up 3 or 4 stations operating PSK on 20M.
Lots of traffic in other modes.. CW all over the digi allocation (they
always have very short callsigns) and of course RTTY from end to end which
is pretty much the norm on contest weekends.
I have not heard anything on the other bands.. but it is very NOISY here in
SoCal.
=====
Alternate Email Addr: W6WLB@arrl.net
ICQ No. 8393042
------------------------------
Date: Sun, 27 Jun 1999 10:46:33 +0000
From: Peter Martinez <Peter.Martinez@btinternet.com>
Subject: [psk31] FSK31 explained
From Peter Martinez G3PLX
My apologies for this long message. If you don't want to read it now,
save it for later!
I now have more information on the structure of Nick's FSK31 mode,
thanks to Nick himself and some observations of off-air signals.
FSK31 is in fact minimum-shift keying (MSK), but with a twist to it.
Let me first explain MSK by starting from PSK so you can see how to
compare the two. This is a little different from the way most
textbooks describe MSK by starting from standard FSK.
BPSK works by making one of two possible phase-shifts for each bit,
either zero or 180 degrees. QPSK works by making one of four shifts,
either zero, 90, 180, or 270 degrees. MSK is just QPSK with ONLY the
90 and 270 degree shifts allowed, so it sends the same amount of data
as BPSK but using the other pair of phase-shifts. I could call this
90/270 QPSK.
Looking at it this way we can see that the weak signal performance of
MSK will be the same as that of BPSK. After all, the BPSK
demodulator needs to know only whether the phase scope trace is in
the top or the bottom half of the circle, and the MSK demodulator
needs to know if the trace is in the left or right half, and the
noise is the same both ways.
Actually, it's not quite right to say that MSK is like 90/270 QPSK,
because with QPSK, there is some amplitude modulation as the phase
changes, whereas in MSK it doesn't. However, performance-wise there
is no real difference so it's quite valid to use this way of
explaining MSK when trying to decide how well it will work. Notice
that MSK will suffer the same polar flutter problem as BPSK, so my
previous remark on this reflector that FSK31 might be better than
PSK31 for polar flutter should be ignored: this would only be true if
the FSK shift value was larger than the baudrate. The difference
between the amplitude modulation envelope of MSK and 90/270 QPSK also
means a difference in the spectrum width: MSK is a little bit wider
than PSK. However, I am sure that MSK fans would point out that PSK
gets wider if the transmitter is not linear but MSK stays the same,
so it's not so easy to say which mode would be narrower in practice.
The frequency shift of MSK is half the baudrate, or 15.625Hz. If you
think of this as a shift of baudrate/4 either side of the centre,
it's possible to see that in one symbol-interval of 32mS, a mark or
space tone 7.8Hz off-centre will have drifted one quarter of a cycle
(90 degrees) out of phase with the centre-frequency. This is the clue
to why we can treat MSK as 90/270 QPSK. Incidently, if we tuned-in an
MSK signal on a PSK31 receiver, the phase scope will look like a
horizontal flipping line. In theory we could actually copy MSK by
just retuning 7.8Hz one way so that the line was flipping vertically
like BPSK does. In practice it doesn't work very well, and FSK31 is
not simple MSK anyway. This is where the "twist" comes in that I
mentioned before.
With BPSK and QPSK, the amplitude of the signal varies as the signal
is keyed. This "free" amplitude-modulation is the key to keeping the
receiver timing in sync with the transmitter. With MSK there is no
amplitude modulation, and if we transmitted a continuous mark for the
idle signal like in RTTY, there would be no way to get in sync at
all, so some other technique has to be used to keep in sync. With
FSK31 - and this is the "twist" - this is done by
differerentially-encoding the transmitted data. So, instead of a
continuous mark for the idle, the signal is switching between mark
and space all the time, and the receiver can recover the timing from
this, in the same way as is done on many other digital modes like
packet. Like packet, the differential encoding also has the advantage
that it doesn't matter which sideband you are on.
This differentially-encoded data is decoded in the receiver by
comparing one demodulated data bit with the previous one. This is not
quite the same thing as the differential demodulation process that is
already used in BPSK, where the phase of one symbol is subtracted
from the phase in the previous one. In fact if we think of MSK as
90/270 QPSK, we can see that decoding differentially-encoded MSK
involves having BOTH a differential phase demodulation process AND a
differential decoding process. The significance of this is that if
there is a corrupt demodulated bit in an FSK31 receiver, TWO corrupt
bits will be fed to the character decoder, whereas if there is a
corrupt demodulated bit in a BPSK receiver, only one corrupt bit is
fed to the decoder. The decoded character error-rate of FSK31 is
therefore likely to be a little worse than BPSK.
All this is theory of course, but I worked through all this 4 years
ago before I chose the PSK31 format. I decided then that the
advantages of BPSK for symbol timing outweighed the advantages of MSK
for non-linear transmitters. It will be very interesting to see how
FSK31 performs on the air.
73
Peter
------------------------------
Date: Sun, 27 Jun 1999 10:14:18 -0700
From: dave_epps@juno.com
Subject: [psk31] R/S Gnd loop isolator
Being that it is setup for stereo.
Is there 2 xfmrs? Could it be used for the soundcard input and output?
dave ab5pc morro bay ca
___________________________________________________________________
Get the Internet just the way you want it.
Free software, free e-mail, and free Internet access for a month!
Try Juno Web: http://dl.www.juno.com/dynoget/tagj.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
to unsuscribe from the list send to majordomo@aintel.bi.ehu.es a message
with a text line as follow: unsubscribe psk31 or unsubscribe psk31-digest
More instructions on PSK31 Webpage: http://aintel.bi.ehu.es/psk31.html
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
------------------------------
End of Psk31 Digest V1 #91
**************************
Read previous mail | Read next mail
| |
