[600MRG] WSPR Flavors

[Hans Summers]

Hi Frank

My question - On the receiving end, in order to maximize the ability to
decode a WSPR signal, does the receiving station have to use the same
"flavor" WSPR as the sending station?  By flavor I mean:  (1) The same
transmit / receive duration in  the 10 minute cycle?  (2) The same
revision, e.g., WSPR-X vs another version such as an early issue of WSPR
that is a number of years old? (3) And on occasion, I have seen / heard
stations referring to things such as WSPR-1, WSPR-1.2, WSPR-15, or
whatever.  What are these and their impact on successfully decoding a
signal if the receiving station is not using that same version (flavor) as
the transmitting station?


There is only one WSPR encoding scheme. The only two variants are the 2
minute cycle WSPR (actually just under 1 minute 52 seconds transmission
time), and the 15 minute cycle. The use of the 2 minute WSPR version is
almost universal and is supported by all versions of WSPR encoding/decoding
software.

The 15 minute WSPR (called WSPR-15) is only possible with some of the WSPR
software packages. Not all. WSPR-15 was intended for LF operation but does
not seem to have become popular. Standard WSPR has a tone spacing of 1.46Hz
(which is 375/256) and the symbol duration is 256/375 seconds. There are
162 symbols in the transmission. In WSPR-15 the sequence of tones is
exactly the same, but the tone spacing is 1/8th and the tone duration is 8
times longer. Hence it takes almost 15 minutes to send.

The WSPR encoding protocol has been unchanged since the beginning. So it
makes no difference what software version is used by transmitting/receiving
station - except of course that they both need to be on either 2 minute or
15 minute (LF) transmissions. When people referred of WSPR-1 or 1.2 or
whatever they were probably talking about some specific features of a
particular version of the PC software.

The only exception to that is the "extended" mode WSPR which sends TWO
2-minute transmissions, and allows sending callsign prefix/suffix (with
certain constraints) or 6-character Maidenhead Locator precision instead of
the usual 4. This variation in the encoding wasn't available in the early
software versions, it was added later. If an early software version heard
one of these "extended" transmissions it would only be able to decode the
first of the two parts. You do see the "extended" system used sometimes. In
my opinion it's a bit of an abomination except in certain limited
situations (I won't bore you with my reasons, this email is long enough
already) so I consider it fortunate that "extended" mode is not used much.

There were also improvements to the decoder in the last couple of years.
Much of that was innovations by K9AN. The latest software versions use the
K9AN decoder. It has deeper search capability, and successfully decodes a
larger number of weak signals than the earlier decoder. This is not a
change in the WSPR protocol, as I said that is unchanged... it is just an
improvement in the digital signal processing to enable weaker signals to be
recovered better. Even K9AN's decoder has room for improvement... but
everything is a trade-off and the law of diminishing returns applies...
beating K9AN will take a lot of work and probably only marginally improve
the decode capability.

FYI my Ultimate3S transmitter kit http://qrp-labs.com/ultimate/u3s can send
both the most common 2 minute WSPR, and the (fortunately) rarely used
Extended type  (prefix/suffix/6-char locator) and WSPR-15 below 1MHz. It
generates RF output directly on any band from 2200m to 2m. Power output
depends on frequency. 250mW mid HF. More power lower frequencies, less
power higher frequencies. An optional 5W HF PA kit is available. GPS can be
connected for frequency and time discipline (important for WSPR). It's a
standalone system requiring 5V power and an antenna only (no PC to control
it or configure it). A lot of fun starting from $33.

Since this is a 600m group I should mention a number of features that were
added specifically for 600m users:
1) WSPR-15 mode
2) configurable second frequency output from the Synthesizer with inverted
phase, for driving push-pull amplifiers
3) x2 configuration for generating double-frequency output, for 600m amps
that require this double frequency input
4) PTT delay feature, to key a Power Amplifier a configurable number of
milliseconds before the RF transmission starts (and after it ends) to avoid
hot-keying big amplifiers.

73 Hans G0UPL
http://qrp-labs.com