Re: LF: its back - maybe -

[Andy Talbot <andy.g4jnt@googlemail.com>]

Usually impulsive interference, carriers  and the laws of probability.

 

Given a data stream with strong source coding and all original redundancy removed, then any data that passes the error correction tests on receive must then be decoded as something that looks like valid source data.

 

The false decodes you are getting are those bursts of interference that get interpreted as valid-ish tones, and by chance fall into a sequence that the convolutional decoder accepts as a reasonable contribution to the decoding.  

 

WSPR codes callsigns assuming a letter or number,  followed by a number, followed by three letters or numbers (or spaces). The locator is encoded as a 15 bit value representing any of the 32400 possible locators it could ever be, and power is coded as a 6 bit number showing any of 60 possible power levels.   

 

So if the error decoder decides the rubbish/QRN  it has received may have enough symbols in the correct position to be a valid WSPR sequence, then whatever pattern of 72  '1's and '0' it lets through to the decoder will always look like a callsign + locator + power level.

 

The error detection circuitry is pretty immune against noise, but impulsive interference and spurious carriers may confuse it

 

You do seem to be getting more than most people though. 

Andy
www.g4jnt.com

On 21 September 2010 15:14, Dave Pick <davepick@btinternet.com> wrote:

Andy

Yes, some very interesting callsigns coming up! Super DX... some of them appear to be in the middle of the Pacific ocean!
What causes that?

Dave.

---

From: Andy Talbot <andy.g4jnt@googlemail.com>
To: rsgb_lf_group@blacksheep.org

Sent: Tuesday, 21 September, 2010 14:45:07
Subject: Re: LF: its back - maybe -

Several decodes of the QRP transmission at G3YXM, all at the -25 to -27dB S/N level, so right on the threshold.

 

Dave, you're getting quite a number of false decodes accordng to the WSPR database.  Inpulsive QRM perhaps ?

 

Andy

www.g4jnt.com

 

 

On 21 September 2010 11:47, Dave Pick <davepick@btinternet.com> wrote:

Andy

I've put my WSPR back on on 137.5 but as I was setting it up I heard a "G8IMR" CW ID so I'll be surprised if I don't get a copy!

Dave

By the way, I had a similar shed conflagration a few years back, it was better than yours - the fire brigade attended!

---

From: Andy Talbot <andy.g4jnt@googlemail.com>
To: rsgb_lf_group@blacksheep.org
Sent: Tuesday, 21 September, 2010 11:10:10
Subject: LF: its back - maybe -

Hope I'm not typing this too soon, but monitor 137.53kHz (USB dial 136kHz) WSPR.

600 Watts again

 

A bucket of water is lying close by this time.  

The coil is now in a brand new plastic dustbin with the variometer control brought to the outside world on a plastic spindle.

 

Nary a bit of aluminium tape in sight

Andy
www.g4jnt.com

ps.  the WSPR message is still untrue.  Its nearer 17dBm ERp than the 13dBm is says.