WSPR has restored my faith in AR

[VK4DU]

After a mate told me of his experiments with WSPR, I decided to give it a shot myself.

This cynical radio professional is impressed. Mighty impressed.

You can watch HF propagation shift with the sun and the sunspot cycle on the google map presentation of WSPR stations (at http://wsprnet.org/drupal/wsprnet/map).

I’m on 40m at the moment (0930Z), and am connected to a load of US stations. When I check the PC in the morning, the Europeans are connected up to me. All with 5W and a dipole. On the higher bands (with better sig/noise), even lower power works fine.

My mate runs 100mW on 20, and regularly connects to dozens of stations with an antenna described as “a piece of wire lying across my roof”….

It is fascinating watching the performance of individual stations – the VK7 stations, for example, really work well on 40 to Europe. There is one VK7 who must have a wonderful QTH…

The WSPR algorithm is stunningly good – I can watch my system decode signals that I simply can not hear, at all.

For those of you who haven’t tried WSPR, give it a go.

73
Glenn VK4DU

[VK3ZAZ]

wspr doesn't connect it decodes
and yes it is a nice way of seeing propagation
and its hands free
no clap trap
no waffle
just pure science
suits me to the ground as its seamless
and you get some incredible hits
I even set up wspr as a beacon at work qth
That's was interesting
and tricky
enjoy glad to see you have rekindled your interest
Ive lost my interest in the same old same old and all the posted selfies on the loggers.
WSPR does it all for you uploads data downloads data and tells the whole story.

[VK4DU]

Yes, I meant decode rather than connect, per se.

All the old beacons should be replaced with WSPR ones, IMHO.

So many advantages; duty cycle, power requirements, instant automated reporting, etc, etc.

[VK4APN]

I have just finished building the Ultimate 3 WSPR (and other modes) kit. See http://www.qrp-labs.com/index.php?route ... duct_id=50

This kit is an elegant piece of design, with use of modern "smarts" DDS gen, Atmega MCU, LCD, a couple of fets and some smart firmware. Add a cheap GPS engine with 1 pps and standard NMEA string output and you have very reasonable precision timing and frequency accuracy as well - and this for a $30 kit!!

It is fascinating to see your signal being decoded thousands of kilometers away - emanating from such a simple and very low power device.

If you are looking for a simple fun kit to build, you wont be disappointed in the U3 Ultimate.

Good luck

[VK2GOM]

I have often looked at the Hans Summers kit, but am holding out to see if they ever develop a WSPR tansceiver along the same lines. That would be more than double the fun of just a transmitter/beacon.

73 - Rob VK2GOM / GW0MOH

[VK2OMD]

After a mate told me of his experiments with WSPR, I decided to give it a shot myself.

This cynical radio professional is impressed. Mighty impressed.
...

I also experimented with WSPR, both on air and extensive data mining of my own data, and all data.

I was particularly interested in evaluating it as a tool for comparison of stations overall, or station components (an A/B antenna test if you like).

There were a bunch of reasons for which I abandoned WSPR as a serious tool.

WSPR has been appropriated by a bunch of train spotters who 'rule' on the maximum tx power, irrespective of whether it is sufficient to be likely to be heard over a path of interest. These people live for the once in a lifetime 'spot' and are vocal in condemning people who run sufficient power to get a few spots an hour over a working 'circuit'. Leading this brigade are the guys running 500mW or less to an attic dipole in residential precincts, they can't hear much but claim kudos when there signal is heard by a good low noise station on the other side of the world (read UK attic dipoles heard in rural Australia on 40m at 500mW when they never spot the opposite path at 20W).

At the time of my experiments, the on-line database archives were not maintained adequately. Several days in a month were not consolidated to the archives and this despite emails to the maintainer prior to permanent loss of the data. This turned out to ruin some experiments coordinated with another station outside the regular WSPR segment so as to avoid interference problems.

Errored records are common place as for instance operators change bands without changing the software settings.

Sheer lack of care on the part of operators winds up with too many HF spots in the middle of oceans and deserts etc and when you look at them, they are most unlikely positions / paths for the call sign used.

Now since so many WSPR bunnies hop from band to band, you have no idea how many times they transmitted on any band in any interval, so you cannot evaluate even how many times you spotted them compared to how many times they transmitted... and for some of these bunnies, you might only receive a couple of spots in an hour... meaningless statistically but yes, a thrill to train spotters.

When you get through all the hurdles and have a data set of valid records, the next problem is that the SNR metric reported is not a linear indicator of S/N for two reasons, at the high end it suffers from non-linearity in the measurement system, and at the low end, SNR below -20dB fall off because low probability effects (whereas if you looked at the laboratory measurement you would expect increasing frequency of spots as SNR approaches the decoder's 'brick wall'). The uncertainty of the measurement system is huge. (By comparison, some co-ordinated QRSS / grabber tests that I conducted had undertainty in the tenths of a dB range and did permit comparison of two antennas based on several thousand S/N measurements over an hour usgin a QRSS keyer that alternated the antennas on one minute intervals).

So, my own experience was that it is a waste of time for serious propagation studies or for attempts to compare stations or station components.

It is however, a great way to run up your QRP hours total, WSPR and QRSS tests are the main reason why I can claim to have used QRP for more than 90% of my transmitting hours.

Owen

[VK2OMD]

I have often looked at the Hans Summers kit, but am holding out to see if they ever develop a WSPR tansceiver along the same lines. That would be more than double the fun of just a transmitter/beacon.

See my comment above about attic dipoles, tx only, either intentionally or effectively as a result of poor station rx performance is very common.

Owen

[VK4DU]

[VK4DU]

But seriously, yes, I know it has shortcomings, but it is interesting nevertheless.

It is AR, afterall....

[VK2OMD]

But seriously, yes, I know it has shortcomings, but it is interesting nevertheless.

It is AR, afterall....

Yes, yes, it is interesting and worth exploring... just for the understanding if nothing else.

Owen

PS:, love your portrait... you have changed since we last met!

[VK4DU]

Ha.. Yes.

[VK3ZTE]

Some comments on my experiences with WSPR: Having come back to AR after a long absence and attempting to do it homebrew, I have found that despite its limitations, WSPR tells me far more about the inadequacies of my homebrew gear than the invariably kind, well meaning reports of the 'yea sounds good to me 5x9' type when in many instances it has turned out I had all kinds of issues incl poor freq stability, lack of linearity, rf feedback and just about every other rf malady you can imagine. WSPR on 2m has provided huge insights into what has been going wrong and equally importantly, a reliable means of evaluating the fixes. For me, on receive, it has complemented the valuable contribution of the 2m beacons. I should also mention that VK3YT's Web SDR site has provided similar, enormously useful assistance to my 40m homebrew efforts.

[VK3BQ]

Ive been tinkering with my raspberryPi and wspr on 30m, with my ~5 milliwatts getting me a decode into vk6 (2800km). the low pass/ band pass filter, some wire and the pi is a beacon!.

and no, it does not listen (yet)

http://www.vk3bq.com/2014/08/12/raspber ... -wsprrypi/

its interesting to see the propagation move around and see just how man people can hear my 5mW now thats MAX POWER.. imagine how much better it would work with unicorn wire.

[VK3ZAZ]

Compared with the carbon life forms who infest the BORG LOGGER day after day posting selfies about the same old paths I find wspr a refreshing hands off approach to AR prop studies.
I did it on 80m a while back with a few watts and found stuff I didn't know about.

Sadly the only thing that wspr didn't do was open up the pacific 6M path we still had to do it on our station merits.

But it did show a 6M path to UA 0 Russia

You cant pick a fight with a robot.

[VK2AVR]

WSPR sucks
The people who use WSPR suck
Their antennas suck too
The WSPR database sucks
Everybody sucks

I think you're asking it to be something it's not, Owen. WSPR is a fascinating tool for seeing what low-powered propagation options are possible. It's not a scientific measurement tool, I've seen your graphs on QRSS between different antenna types and the result was fascinating. But to do real propagation analysis like that you need to control all the variables. It is impossible to do that with a community-based database.. but that should not take away from the excitement of knowing that a small fraction of 500mW of RF made it halfway around the world.

[VK4APN]

Ditto

Paul vk4apn

[VK3HJ]

I found it an interesting challenge setting up WSPR and getting it working. Timing is critical.
However, when I did get it working, I set up on 160 m with 10 watts to see what was happening. I was spotted in VK2, VK3 and VK6. Big deal...
My interest lasted two nights. It was, however, an interesting exercise in setting up some applications on the computer.

I know of at least one station using up to full legal power (1500 watts) on JT65 in a desperate attempt to make QSO on 160 m, under poor mid-year band conditions.

By all means, keep on having fun with WSPR and other clever digital modes.

You'll find me on CW.

73,
Luke VK3HJ

[VK2OMD]

VK2AVR, I expect that you know I take offence at people falsely representing that I have said something that I didn't, and your post could be taken as intended to cause offence.

I did not say what you claimed I said in the following quote, they are your words entirely and a fraud.

WSPR sucks
The people who use WSPR suck
Their antennas suck too
The WSPR database sucks
Everybody sucks

Owen

[VK4DU]

You'll find me on CW.

73,
Luke VK3HJ

Ha ha ha!

So, only CW operators are real amateurs, eh?

Oh, teh irony....

[VK2KRR]

I found it an interesting challenge setting up WSPR and getting it working. Timing is critical.
However, when I did get it working, I set up on 160 m with 10 watts to see what was happening. I was spotted in VK2, VK3 and VK6. Big deal...
My interest lasted two nights.

Luke, Ive been hearing K7FL, KJ4YBS, KH6UL, PA0O, PA0A, nearly every night past week. Might be worth another look, though its starting off quiet this evening. But I notice there has been a lot of solar activity today which could have capped things off.

They do say people expect everything NOW. You cant expect miracles everyday, they would no longer be miracles if they happened every day. Yes if you want to push a signal through where it doesnt want to go you can run up bulk power and give yourself a better chance, mostly I find its either open or closed and no matter what you throw at it, it will not work. There is no point trying to force signals, they will travel where they want to easily enough when a band opens up. And really thats what WSPR is mainly for, looking for good openings. Your best to get used to whats going with bands on WSPR over a long period of time and you will then more easily understand when a band is open and good enough for voice etc.

All the old beacons should be replaced with WSPR ones, IMHO.
So many advantages; duty cycle, power requirements, instant automated reporting, etc, etc.

I agree and have mentioned this long ago. The other advantage is, they only need to use a single frequency. But they would all need to be very accurate in freq and timing. They can also have CW ID at the end of transmissions for those who need it.
As Ive said in the past, if it was tried on 144 MHz band in Australia, you would know which beacon was transmitting in any transmission because of its location in the WSPR bandpass, eg VK4RTT could be on 144.490440, VK5VF on 144.490450, VK6RPH 144.490460, VK3RGL 144.490530, VK5RSE 144.490550 etc.