137kHz 30-40W TX transverter schematic

As promised, here is the schematic of the current 30-40W TX transverter for 137kHz. I have used a 10MHz crystal but a lower local oscillator frequency would be slightly better from a stability viewpoint when using narrowband modes like QRSS30 or WSP15. One choice is to use a low cost 1.843kHz crystal mixing with the driver transmitter's output at 1.979kHz. When a stable external LO source is possible e.g. a GPS disciplined oscillator, remove C7 and feed the LO into the double balanced mixer. Although an SBL-1 mixer was used an alternative, available from eBay is the ADE-1. As balance is not super critical, a home made double balanced mixer would also be possible.

For PA heatsink I just used a couple of small TO220 heatsinks in the prototype. With optimal PA matching into 50 ohms there is little heat from the PA device. A larger heatsink could prevent overheating when the antenna is mismatched. My output network was optimised for my slightly higher than 50 ohms output load presented by my earth electrode antenna.

ERP is around 30uW with my earth-electrode antenna.

Some may prefer to use capacitive coupling into the FET gate. If TR3 failed open circuit, the gate voltage would rise leading to the IRF640 failing. Hasn't been a problem so far though.

A dialogue with OFCOM

At last, and only 3 months late, OFCOM have renewed my NoV for 8.7-9.1kHz VLF operation.

Currently I am "in negotiation" with OFCOM on another matter mentioned a few days ago.

In the next few months I want to do some QRPp radiated tests with the earth-electrode antennas at some frequencies between 9kHz and 137kHz. Ideally this would be around 35-45kHz and 70-80kHz, where I can find a quiet spot without any activity from primary users like the Ministry of Defence.  With QRSS, continuous carrier or WSPR and with an EIRP of around 10uW it should be interesting. Tests on 137kHz suggest this power is enough to see a radiated signal out to at least 25km locally with an E-field probe on the car, which would allow me to check coverage and polar plot. On the lower frequencies this may be limited to just 5-10km, but that would be OK.

But, OFCOM say they will not grant me licence-exempt permission to operate in this part of the spectrum, even at the miniscule powers and bandwidth (uWs and a few Hz).

They say they would HAVE to consult with the primary users and this would take months. They suggest instead I apply for a Non Operational (Test and Development) Licence, but does that not need primary user approval too?  Problem is this licence costs £50 a year, which sounds like total overkill when the risk of me interfering with anyone is about as likely as me winning the lottery or landing on the moon when I jump.

Now, I could just go ahead and do the tests anyway knowing I will not cause anyone a problem.  Really this is not the way it should be and anyway I'd like to publish the results later. Hardly possible if I don't do things with OFCOM's approval.  I am hoping that the good people at OFCOM will see that this is worthwhile amateur research, it won't cause ANY issues and they will soon say, "go ahead on the strict understanding that if you cause any harmful interference to primary users you close down immediately, but at that power, no licence needed."   I have written back to OFCOM again this evening making this case and we'll what transpires. I also suggested issuing me a one-off NoV, but they say they won't!

Many of you will be saying, "why not just do it". If the powers that be really cannot see sense, then maybe I may have no other option.

Website working again

The main www.g3xbm.co.uk website is now working properly. I contacted my domain providers last night who confirmed they'd updated some systems that screwed up web forwarding. If you spot any more problems, typos or link errors please let me know.

Currently I am drawing out the schematic of my 30W 137kHz TX transverter. This should be on the website tomorrow. Still deciding whether to box it, and add the RX preamp section, or stop at this stage.

Simple low cost test equipment

There was an interesting post in the last day on the GQRP Yahoo group asking about portable workbenches that led on to talk about test equipment. This got me thinking about simple pieces of test equipment and what one needs to do basic QRP design work.

In my own case the most useful pieces of kit are (1) a signal generator, (2) an inductance/capacitance meter, (3) an oscilloscope,(4) a basic power meter, (5) a wide coverage all mode RX, (6) a multimeter. I keep thinking about building a dead basic spectrum analyser too. Test probes too of course. Also a good lot of cables with BNC plugs one end and croc clips on the other.

Just wondering what other people use? Certainly for HF QRP work only very basic test gear is needed, which is why I enjoy it so much.

In a way it is a little surprising that no-one sells a low cost "all in one" test box for amateurs with many of the above in one unit. A bit like a CB test box?

Not the best of weeks

Well, apart from the excitement over the ferrite rod on TX, this has not been a too brilliant week.

As you saw in the last post, my website domain name provider seems to have messed up web forwarding thereby screwing up the G3XBM website, for now at least. This is annoying as it has also caused issues on the www.cambschoral.org.uk website that I set up and maintain for the Cambridgeshire Choral Society.  And then there is 137kHz where a week of trying has netted me spots from just 4 unique reporters. Even the effort to get WSPR15 working was all but scuppered by the 4Hz drift over 15 minutes in the FT817.

In the grand scheme of things none of this much matters. Perhaps next week will be more fruitful, HI. In the meantime, I'm watching TV and having a whisky!

Website woes

In the last few days I've done a major overhaul of my main website www.g3xbm.co.uk . All was working fine, but for some reason the web forwarding is not working correctly. Landing on the www.g3xbm.co.uk site you should automatically be forwarded to https://sites.google.com/site/g3xbmqrp3/ which is where the site is now located on Google Sites. So far so good, and this happens, although the address bar, which should change to the new site address based on how I've got my forwarding set up does not. Not only that, but clicking any images does NOT bring up larger, clearer images as it should although this works fine when you start with https://sites.google.com/site/g3xbmqrp3/ . So, I think it is a problem with the web forwarding, which I will try to fix.

In the meantime, you may want to go directly to https://sites.google.com/site/g3xbmqrp3/ if you want to explore the revised site.

First WSPR15 transmission decoded successfully

After a LOT of effort, G8HUH managed to decode my WSPR15 transmission on 137.607kHz at 1815 this evening. Even with a very low (1.7MHz) LO, the drift was still 4Hz in the 15 minute TX period, mainly due to the FT817 I think. At only -32dB S/N decode this was actually not as good a result as with WSPR2 over the 250km path. Short of force cooling the FT817 or using a rig with a TCXO reference, there is nothing more I can do.

My WSPR15 transmissions were also copied and decoded by G4FFC and G4FEV.

CONCLUSION: WSPR15 is a much more tricky mode to use successfully unless the transmission is very stable. This is the screenshot Tom G8HUH sent me.

30uW ERP 137.607kHz WSPR15 signal received by G8HUH 

137.5kHz WSPR15 tests

Using a very stable external signal generator as my 10.0000MHz LO 5dBm injection, I'm now in a position to try WSPR15 with a stable signal. After a quick frequency check with WSPR2 I propose to switch to WSPR15 with its greatly improved weak signal capability for this evening and overnight. WSPR15 activity takes place in the 25Hz slot just above the WSPR2 activity.

Please let me have any reports.

The first WSPR15 transmission will occur at 1600gmt.

Update: Tom G8HUH reported an upward drift of 5Hz during the 15 minute TX burst, so I've moved the LO to 1.7000MHz and the FT817 to 1.836kHz. Drift (all due to the FT817) is now 1Hz. Awaiting decodes. 

137.5kHz ...it's getting repetitive

Well, I've had the 137.5kHz WSPR2 beacon running for about 18 hours now, almost continuously on a 25% TX cycle. I've tried to drum up reporters by posting on the WSPRnet chat page, RSGB LF Yahoo Group and the LF Blacksheep reflector, plus a few direct emails to stations likely to copy my signal.

Consistent 137.5kHz WSPR spots from G8HUH at 250km, but no others!

All I've managed are the very consistent reports from good old G8HUH at 250km. The signal is obviously spanning this distance without any issues so why no others reporting?

Later I am going to replace the 10MHz LO in my transverter with the stable source from my synthesised signal generator and have a go tonight at WSPR15 (15 minutes TX periods) and see if that gets me a few more reports. It should be several dB more sensitive. If that fails to get anywhere, then I am calling it a day on 137kHz. In terms of returns for effort, this is not good value.

Internet Time Servers

For WSPR, and probably quite a few other programmes, it is important to have your PC's clock accurate to within a second or so. With WSPR, if you are more than a few seconds out the software will be unable to correctly decode signals. There are several internet time servers around which sync the PC clock to very accurate time standards around the word.

The one I use is http://www.worldtimeserver.com/atomic-clock/ . This allows you to download a program called atomic.exe which syncs your PC clock at an interval you choose to an internet time server. It is a great help to keep an old PC in time for WSPR. Without this my old Dell Inspiron would be all over the place with time. Once installed there is nothing to do as long as you have an internet connection.