WSPR on VHF

[VK5PJ]

Having recently stumbled across the WSPR digital mode after hearing a rweference to it on the WIA broadcast, I have been leaving my HF radio on 30mx running the WSPR software to see what is heard. It had occured to me that we could do the same on the VHF bands when not listening to beacons etc.

The first band that came to mind was 50 or 52 MHz, maybe 52 might be better to let more people get involved.

WSPR will receive by default then randomly tx for a one minute sequence with a very narrow freq s**t to put data on what seems like a straight carrrier... so we coulsd create a network of beacons

Any thoughts?

[VK2OMD]

Any thoughts?

Peter,

I thought WSPR might be interesting and looked at it briefly some months ago.

The dearth of information on WSPR makes it difficult to understand.

As I understand it, it monitors a band 200Hz wide for signals from another WSPR station.

Thinking about two stations using stock IC910s on 144MHz, their rated frequency stability is +/- 3ppm fo -10° - 60°C, that is a spread of 860Hz between the two stations. Clearly that would reduce the probability of capturing band openings because the actual receiver frequency might be far enough from the actual transmitter frequency that signals do not fall within the detection band.

With the hi-stab option, things are better. Rated frequency stability is +/- 0.5ppm for 0° - 60°C, so that is a worst case difference of 144Hz, so probabilities have improved.

That makes me think it was probably intended more for HF than for VHF, much less even higher frequencies.

I have never quite understood WSJT's indicated S/N ratio. The documentation isn't clear on how it is calculated, and leaves me with a sense that bigger is better, but that although it is reported in dB, it might not be linear, and absolute reference is uncertain. I guess that the ease with which discussion in the documentation seemed to swap between S/N and (S+N)/N left me unsure whether reported S/N is truly S/N, the effective measurement bandwidth, and integration time. I was left quite unsure of the 'instrument' strength of WSJT, and I think WSPR inherits that attribute.

IIRC, it has provision to encode the tx power, is it EIRP, I don't recall, but that is of more value for a radiator that does not have high directivity. VHF antennas used might by normal stations (ie not beacon stations) might have high directivity.

Again, makes me think it is for a HF experiment rather than VHF and above.

Interestingly, WSPR should fall within LCD 9(1)(d) permitting unattended operation by an Advanced station (subject to some further conditions).

Perhaps you could make contact with JT and get his comment on WSPR applicability on VHF and above.

A survey of a VHF path should provide a great deal of interesting information, the question is whether WSJT is a good tool for the purpose, or whether it might be adapted to improve its suitability.

Owen

[VK6ADF]

A friend of mine in Perth asked me the same thing a few weeks ago. He also has been monitoring 30 metres and asked if I had heard of anything of 6m activity. I started listening on 30m myself on and off but yet to receive anything.

Have you had any luck sofar on HF Peter?

Also this
http://sv8cs.blogspot.com/2008/04/oz-de ... -mode.html

73 Phil...VK6ADF

[VK5PJ]

Hi Dave,

A friend of mine in Perth asked me the same thing a few weeks ago. He also has been monitoring 30 metres and asked if I had heard of anything of 6m activity. I started listening on 30m myself on and off but yet to receive anything.
Have you had any luck sofar on HF Peter?
Also this
http://sv8cs.blogspot.com/2008/04/oz-de ... -mode.html
73 Phil...VK6ADF

I am also a newbie to WSPR and have just spent the last few days looking at it on and off. I must say that 30mx has been the biggest eye opener for me today (Saturday) with signals from EU and NA all being detected on my shoddy old Windom. So I would try 10.138.7 USB for a few hours and let it decode... remember to enable it into RX (do not leave on idle like I did) and then at the start of an even minute it kicks off... one proviso I think the PC's time settings need to be within a second of UTC.

As Owen pointed out the RX bandwith window is a challenge for 144 and above so in its current form it might only be useful to 50 MHz. Any way my station is cooking away on 10 MHz for all of the weekend. It does not help that there is a number of OTHR transmitters there for a considerable amount of time today running various waveforms from their collections....

Regards,

[VK2OMD]

...
As Owen pointed out the RX bandwith window is a challenge for 144 and above so in its current form it might only be useful to 50 MHz....

Peter, have a look at the WSPR spots database. There are entries up to 2m, and it is informing to look at the reported frequencies between different station pairs.

Owen

[VK2OMD]

Though I can see issues with use of WSPR on VHF and above, I would be happy to participate in a trial.

Let me propose 2m horizontally polarised point to point from Canberra from say 0700 to 1000 local time on selected days. Melbourne is a good prospect during those times for AE, if station or stations to my south west were interested in participating.

If others in Canberra were interested, an interesting trial would be to transmit 100% of the time from Canberra on one day, and from Melbourne on another.

Such a trial would provide interesting information on frequency accuracy, and interesting observations on AE windows for nearby stations.

If one or more GPS stabilised stations participated, it would provide an absolute reference for some of the frequency accuracy observations.

My thoughts are that WSPR has shortcomings for this application, but identifying the shortcomings with a practical experiment would provide useful information to feed into the development process.

Anyone up for it?

Owen

[VK4YEH]

I have just started to play with WSPR, mainly on 30m but some brief forays into other bands. The ability of the software to pick up weak signals is quite extraordinary. A couple of sites that I have found useful:

1) get the software from : http://www.physics.princeton.edu/pulsar/K1JT/

2) a good site for basic setup : http://www.g4ilo.com/wspr.html

3) the WSPR.org site - has a logger and other useful stuff : http://wsprnet.org/meptspots.php

The suggested frequencies that I have dug up on the web for 2m are: Dial - 144.4885 , TX between 144.489900 and 144.490100

Yes Owen I'd like to give it a shot on 2m. Not sure how I'll go heading south as I'm on the northern side of a hill, and have a modest 2m rotatable beam.

Any other takers?

Tim Roberts
VK4YEH

[VK2OMD]

...
The suggested frequencies that I have dug up on the web for 2m are: Dial - 144.4885 , TX between 144.489900 and 144.490100

Yes Owen I'd like to give it a shot on 2m. Not sure how I'll go heading south as I'm on the northern side of a hill, and have a modest 2m rotatable beam.

Any other takers?

Tim Roberts
VK4YEH

Tim, you are around 930km @24° from here. That would be interesting.

4FNQ is around 1710km @350° from here... that is almost the optimal distance for extreme Es for 2m according to Mike Willis... so that would be an interesting prospect. Unfortunately there is 35° difference between those, a bit of loss in gain if I middle the antenna. I will contact John and see if he is interested.

I looked at 144.490 on the band plan, it is in the middle of the beacon segment, but it looks like the nearest station is 5kHz lower.

Perhaps to try it on some days during the Es season when the Hepburn charts suggest high probability of ducting. I would probably think of 10W, and the antenna could have gain up to 18dBi including the effects of ground reflection.

Some of the best Es on the Brisbane ionosonde can be before daylight!

My WSPR already talks to the logging database, so all that side is working, I have heard and been heard on 30m, and things work.

Owen

(Above updated 21/12.)

[VK4YEH]

Thanks Owen

Despite the potential difficulties, still worth a try. I have a couple of locals in mind may be interested before we try dx - I'll report back with any news.

Tried my rotator this arvo and it didn't work, so had it down on the bench. Fortunately very easy fix, just a frayed wire from rubbing.

Please feel free to contact me direct when you are ready to give it a try - pre dawn is good for me as I am usually up before 5.00am anyway.

I'll look out for you on the WSPR waterfall!

Tim R
VK4YEH

[VK3HZ]

I looked at 144.500 on the band plan, it is smack in the middle of the beacon segment, but it looks like the nearest station is 15kHz lower... so that should be ok.

Apart from not complying with the band plan, operating in the middle of the beacon segment is not such a great idea, particularly if someone nearby is looking for VK8RAS

The segments 144.220 - 144.240 and 144.320 - 144.340 are specifically designated for Digital Modes. At the moment, the convention is to use 225/325 for the JT modes, and 230/330 for FSK (meteor scatter). Perhaps 235/335 should be designated for WSPR?

Regards,
Dave.

[VK4YEH]

2m WSPR experiments :

This evening Ray VK4ZW and Paul VK4ZBV and I had a play with 2m WSPR - in note form here is what we found:

Ray and I were doing TX/RX and Paul was monitoring and reporting back on our mistakes and successes, as well as coaching in some aspects that we were very unsure about.

- In the WSPR software window set the the dial and transmit as if you are going to use 30m. We used Dial: 10.1387, TX 10.140200
- Set the dial on your txvr to 144.488500 . The offset will automagically be calculated to match the offset on the WSPR pane - in our case 144.4900
- Ray and I both use txvrs that have minimum output of 5W. Better local results would almost certainly be obtained with lower power.
- We both used vertical antennas
- Proximity was a problem for us as Ray and I are no more than 1km apart and Paul is less than 10km from us. This makes turning the rx gain down low enough quite problematic. There were multiple paths from both of us, but I think this is as a result of rx "overload" due to proximity.
- On the logger http://wsprnet.org/meptspots.php our final success was recorded as a 30m contact rather than a 2m contact - I have yet to find out how to change this to reflect exactly where we are.

All in all a very successful trial - all three of us learned a lot.

Based on this I think that WSPR on VHF is very achievable but it is worth spending some time working out how to get it working properly and getting an understanding of what is and is not effective. Local trials are well worth the time and effort I think, before leaping into dx.

Experiment, experiment, experiment!

I hope this is of some use to other WSPR'ers. please feel free to contradict, change or add to these observations as you see fit.

Tim R
VK4YEH

[VK2OMD]

I looked at 144.500 on the band plan, it is smack in the middle of the beacon segment, but it looks like the nearest station is 15kHz lower... so that should be ok.

Apart from not complying with the band plan, operating in the middle of the beacon segment is not such a great idea, particularly if someone nearby is looking for VK8RAS

The segments 144.220 - 144.240 and 144.320 - 144.340 are specifically designated for Digital Modes. At the moment, the convention is to use 225/325 for the JT modes, and 230/330 for FSK (meteor scatter). Perhaps 235/335 should be designated for WSPR?

Regards,
Dave.

Dave,

WSPR is new, and quite unlike any other application. It is most similar in purpose to propagation beacons, but uses a more sophisticated modulation type than traditional propagation beacons and incorporates a decoder and automated logger than can be integrated with a global propagation logging database.

There appears to be an ad-hoc convention internationally that WSPR uses 144.489900 - 144.490100 MHz. As a propagation logger, use of a common segment is important to obtaining best utility.

The Nov 2008 band plan does not mention WSPR. As we might expect, the bureaucracy of band plans cannot anticipate new applications well, and are unlikely to be sufficiently flexible to incorporate new modes without this kind of grief. We worked hard to unbind ourselves from mode restrictions in the LCD that prevented experimentation, but since then the WIA Band Plans have grown.

In the 4MHz wide 2m band, we are having a discussion about the appropriate place for a 'mode' that has a necessary bandwidth of 200Hz to contain all stations using it, that is 0.0025% of the band. In fairness, such operations need a guard band from substantial interference sources. It seems that 20kHz is barely enough distance from local stations running high power.

I note that existing beacons appear to have been allocated channels as close as 1kHz (eg VK3RGI, VK3RRU according to the VKLOGGER database).

What is the meaning of the term 'beacon', and is WSPR more a propagation beacon than anything else?

Is the beacon segment only for use by formal 'beacon licences'? Is new technology permitted, or are only traditional beacons permitted?

Certainly, the WIA lays claim to control of the beacon segment in the band plan when they say "Beacons give an indication of band conditions and provide a warning of DX openings. They also serve as test signals for receiver calibration and testing. There should be no other transmissions within the beacon segments or on their band edges. This applies even if you are hundreds of kilometres away from the nearest beacon! On the VHF/UHF bands, beacon frequencies are allocated according to a geographic allocation plan with a frequency spacing of 2 kHz. Further details on beacon frequency allocations are available from the Technical Advisory Committee."

I have never had a timely response from the TAC, so I would not waste the time asking.

Re the classification under the 'digital modes' umbrella, I could argue that FSK441 (sometimes QRO+) and WSPR (usually QRP) are not good candidates for sharing a narrow band segment.

Tim, I think I will leave this until the WIA has made an explicit recommendation in the band plan. To be useful from VK, they will need to coordinate with all of our RF neighbors, and they might even address WSPR on all bands.

Owen

[VK3HZ]

WSPR is new, and quite unlike any other application.

It's a development of WSJT JT65, and both are classed as "Digital Modes".

There appears to be an ad-hoc convention internationally that WSPR uses 144.489900 - 144.490100 MHz.

There are many international ad-hoc conventions, but the 2m band is essentially a local thing (EME excepted). We have band plans for a reason, and they try to cater for these conventioons, as far as they affect us.

The Nov 2008 band plan does not mention WSPR.

The band plan only explicitly mentions JT and FSK in "e.g." because of their popularity. However, all Digital Modes, like WSPR, are already incorporated in the plan.

I have never had a timely response from the TAC, so I would not waste the time asking.

My experience differs - John Martin has always been very timely in his responses to me. I have asked the question, and we'll see how long the answer takes

Regards,
Dave.

[VK2OMD]

WSPR is new, and quite unlike any other application.

It's a development of WSJT JT65, and both are classed as "Digital Modes".
...

Oh well, it is done and dusted... part of the general classification of 'Digital Modes'. Simple rules for simple folk.

The two small segments allocated to 'Digital Modes' on 2m are both active here on weekend mornings, and IMHO WSPR is incompatible with high power nearby FSK441.

Personally, I think that the general classification of 'Digital Modes' does not serve us well. It is a perspective that doesn't consider the differing modulation schemes, applications, compatibility etc... but I know it is popular. It makes as much sense as the band plan statement that "modes such as SSTV are usually sent as SSB signals, so these modes can be used in the SSB band"... the dumbed down explanation implied is that SSTV is SSB because it is fed into the mic jack. The reality in this case is that there are SSTV repeaters that will relay a pic irrespective of whether they cause interference, and for that reason these rule ignorant robots and their feeds are incompatible with ad-hoc SSB telephony.

For those who may not have heard a WSPR signal, it at first sounds like a steady (unmodulated) carrier... but it is 4 level FSK with a tone separation of about 1.5Hz and a baud rate of about 1.5Bd, necessary bandwidth is about 6Hz... but the receiver monitors a 200Hz segment for multiple transmitters. Because of the location of the 200Hz segment at baseband frequencies of 1400-1500Hz, it will not be possible to use CW filters in many receivers, so they must run with voice filtering which requires a guard band of about 2kHz on both sides.

Owen

[VK3HZ]

Well, I did get a provisional reply to this within hours of emailing John, but some clarifications took a little longer.

I think that as new digi modes come along (and others become superseded), we should be able to accommodate them in the existing band plan framework. See the footnote re the digital segments on 2 metres (and the same pattern applies on all other VHF-UHF bands). We have set aside several focus frequencies based on bandwidth, so the slow/narrow ones are kept apart from the faster/wider ones, with medium bandwidths in the middle:

The following spot frequencies are recommended for digital DX operation using SSB-based modes:
144.220 / .320 Weak signal modes with bandwidths below 100 Hz, e.g. PSK and slow CW
144.225 / .325 Weak signal modes with bandwidths up to 500 Hz, e.g. MFSK, JT44 and similar
144.230 / .330 High speed meteor scatter modes with bandwidths up to 3 kHz, e.g. FSK441

One of the factors to bear in mind is that a new digital mode doesn't of itself create more digimode stations needing to fit in the same space - it is more likely that stations using WSPR might be doing so instead of another comparable mode. But it is true that some stations might be experimenting with WSPR at the same time as others using JT65.

Then some investigation into the bandwidth occupied by WSPR produced:

For modes like WSPR the correct spot in the band plan is in the .220 / .320 channels, which are specifically for the modes with the narrowest bandwidths.

Regards,
Dave

[VK2OMD]

For modes like WSPR the correct spot in the band plan is in the .220 / .320 channels, which are specifically for the modes with the narrowest bandwidths.

Thanks Dave, it is the response you predicted based on the JT heritage, and I am not surprised at the traditional thinking.

However, as I reasoned earlier, it is unlike other existing applications (even if a derivative of WSJT) and my view is that it is incompatible with nearby FSK441 activity, so I shan't waste any more time on trials.

73
Owen

[VK5PJ]

For those who may have some spare time on their hands during the cooler times, I have a WSPR beacon running on 50.293 (the international WSPR freq. for 50 MHZ).

It is currently 1W (+30dBm) into my 50 MHz yagi pointing out towards VK3 from here. I can swing the yagi if wanted to different directions for a few days.

Both vk3ot and vk7jg are hearing the 1W on a regular basis, would you like to have a listen and report what you hear? If so the latest WSJT package can decode WSPR or you can join the WSPR group by downloading the dedicated WSPR package.

[VK7HDX]

I too have downloaded WSPR. Its a nice neat bit of software to use.

73..Karl
VK7HDX

[VK2JDS]

hi Peter, I too am aimed over your way listenning on 50.293, i get bursts of data saying off timing by 1.5 seconds and give a garbled decode, but they are visible and audible too, they are drifting up and down out of the noise regularly.
also the melbourne beacon is visible a lot of the time too. the problem i have which you might be able to assist with is on 2 laptops i have tried to run it on i get it crashing and sending error report off to microsoft after a few minutes.
i have found that by opening wsjt7 as well as wspr it seems to stay running for many hours , but the processor load is a problem as the laptop is getting gaps in the received data.wspr mode in wsjt7 isnt giving as much info on the decoded(misdecoded) data.
73's from Dave vk2jds

[VK2BLS]

Hi Dave,

New to wspr here, I am up and running on HF but, similar to the 2M question just answered, I am seeking clarification of the correct frequency(ies) to run wspr on 50Megs within the VK band plan.
The international std seems to be dial freq 50.293 / 50.2945 tx.
VK3RMV is also on 50.293 but, is listed as a wspr beacon.
Can I/we also run wspr, tx'ing on this freq? or, what is the correct spot/s to 'tx & rx' wspr on 50 meg?
I understand that many wspr stations can operate simultaneously on the same freq. but, it seems that the carriers could qrm anyone trying to listen to 3RMV audibly?

Regards,
Darrell vk2bls