I built a thing! This is a GPSDO using a PCB and kit of parts supplied by G8CUL and a OXCO from G1OGY. It uses a Jupiter GPS module which provides the PPS signal and a 10kHz output and the completed module provides 2x 10MHz and 1x 1MHz outputs. Although there are a number of such designs this one is nice in that it also has a display and shows the current date and time as UTC.

This was, I think my third SMD construction and certainly the Mose SMD devices including multi-legged chips. No issues in construction especially given the quality of the PCB that G8CUL had made.

The backup battery is a CR2 3.3V type and helps with warm starting. As the regulator gets hot I managed to fit a heatsink between it and the rear of the case and hopefully this will sort out heat transfer, otherwise I may need to bolt another heatsink on the rear. Construction in a die cast box would have been better maybe but the blue/white box fits in with others in the shack, plus I had it already! The bezel is cut down from a 3D printed one from Printables.com designed specifically for the 2×16 LCD displays. The button – which is not the best but I had one etc. – selects the various displays which include date and time, satellites seen, latitude, longitude, altitude and QRA locator.

So this is my day job – well, not exactly a job and only one day a week – volunteering at a heritage railway. And here we are in the middle of nowhere… good job it isn’t raining this time.

I had a number of Molex pins to wire up recently. To make things easier I decided to use some 4-core signal cable I had but found that the insulation is so poor at resisting heat that soldering the Molex pins was a non-starter as it always ended up with bare wires. Of course, Molex pins are designed to be crimped… so off to eBay.

The latest addition to my toolbox arrived in a couple of days and made the job a lot easier.

When I first got my foundation licence I was lucky in that a friend sold me a used FT450D at a seriously good price. Since then it has been in regular use. But now I also have the TS2000X I am somewhat torn. Currently the TS2000X is plugged into three antennas, the loft wire + tuner, and the loft 2m and 70cm big wheels. I had been using an FT817 for 2m and 70cm but that is now sat on the floor all disconnected. Poor thing.

But what to do with the FT450D… I do like the radio, it’s very easy to use, nice display etc. and quite compact. Ok, not compact like the ‘817 but it’s a 100W rig. But the TS2000X does everything I need in just one set, so do I sell the ‘450D? Decisions…

I mean, I don’t need both, I won’t use both together… or will I? Well, yes actually, I can run the ‘450D on HF and the ‘2000X on VHF, especially once I finally set up my external wire (the long one not the current short one). And the ‘450D is connected to the Linux box with the ‘2000X connected to Windows. So I can run both.

And I am reminded of those photos – you know the ones, someone in the US with a 100 foot workbench and several thousand transceivers (ok, a bit of an exaggeration there).

So I am keeping both. So there.

Now… the ‘2000X does FM on 2m and 70cm so another question is do I sell the FTM100DE? I mean, I rarely use C4FM. Nah, who started this selling question anyway?!

I am surprised that I actually managed to do something useful today. For a while now I have been meaning to put a Cross Country Wireless HF/VHF/UHF antenna splitter to use and today was the day. It is now sat upside down on the top shelf above the radios, fed from the discone in the loft and feeding two little LoRa modules, one for TinyGS and one receiving radiosondes. The TinyGS receiver had been running for some time sat in the loft and was previously the only thing connected to the discone. The radiosonde receiver had a 70cm ground plane in the shack and never received anything. Since reorganising the feed it has burst into life, rather surprisingly finding a balloon quite close to this QTH which was apparently launched from somewhere to the west of the Lake District but the data does not show the launch site. Looking at the altitude figure in the data I suspect it is already sat on the ground. Pity I cannot go out right now to see if I can find it.

Now, the cabling is RG58 and so not particularly good at UHF. The next step is to put a pre-amp next to the discone, and then to run some Ultraflex 10 instead of RG58. It’s only a few metres but every little helps.

TinyGS info can be found at https://tinygs.com

The software on the radiosonde receiver is rdzTTGOsonde and information can be found at https://sondehub.org/

Some of you may remember that I used to collect valves. I started collecting when I was around 6 years old, although back then it was more to impress friends than collect. An old directly heated valve plus a Lego battery box lit my desk up at primary school. I did not start collecting in earnest until the 1990’s and launched my first online valve museum in 1999. Since then the collection grew in several directions at once, including German WW2 types, Russian Cold War types and British military and civilian types. There were specials from all over the world as well including a few Japanese WW2 ones. Valves ranged from tiny little things to a RD150YB that had to live in the garage, and a 6-anode mercury arc rectifier that was equally not allowed in the house, and for good reason too. The main collection grew to over 3,000 types, many of which had duplicates, so probably 4,000 in total. And then there were boxes of valves that did not warrant adding to the collection.

And so the collection continued to expand. While on holiday in the US friend in the US was discussing collecting trends with me and another collecting friend and said he collected US antique types, others collected microwave types and, pointing at me he said I collected everything and there is nothing wrong in that. But it made me think what exactly is my interest. And so I decided to concentrate on what I found most interesting – British military types, mainly in the CV, and A, N and V military series. The collection included a number of CRTs as well and eventually took over the whole garage.

I decided then to concentrate solely on CV types and trimmed the collection to 1,500 types, again with duplicates taking the collection to over 2,000 valves. Of the remainder many were sold and many hundreds went to the National Valve Museum which was nearly as old as my own.

Eventually though three things happened. First, it was becoming increasingly difficult to find new additions. Second, the website was now seeing fewer and fewer actual hits (as opposed to search engine spiders), and, most importantly I realised it had become an obsession. Time to quit. I also came to the realisation that I had an awful lot of valves in lots of boxes and I never even looked at any once they went into a box.

So I decided to close the website and sell off anything I could, donating the remainder to the National Valve Museum. The website was essentially converted to flat HTML files with none of the database behind them and taken over by a member of the BVWS. Of course, all praise to them for doing that, but none to me for all my years of work. Par for the course. In the past 20 years I received just a handful comments thanking me for providing the photographs and information about the collection. I was somewhat surprises at the screams when I announced the website was to close. Of course, I did not make the website for that, I did it because I thought people might actually be interested, and they clearly were back at the turn of the millennium but times change.

I was fortunate that someone local took many of the CRTs and a bunch of valves as well. Of the rest, a few hundred are destined for friends in Australia, the Netherlands, Belgium and Italy but the logistics are going to be a nightmare. Several hundred have found new homes here in the UK. Of the thousands left these went to the National Valve Museum with the more mundane radio and TV types being scrapped as no-one wanted to come and get them. I am keeping many of the early magnetrons for later sale, and some of the more decorative valves for, well, decoration!

There are still several boxes and a cupboard full of valves and they are destined to be scrapped. Selling on eBay as an individual has become more and more complex over the years so I will rarely sell there. As no-one was interested in paying me a visit to take them away they will end up in the dump.

At its height the collection took over half the workshop and half the garage. Once trimmed down to the CV types it was still half the workshop. Now it is all under one bench and I have more space to set up the various tools that have been sidelined for years and actually get back to working on the house.

Time to replace several lengths of coax and run new… 100m ought to do it!

I finally got some wire in the air. Well, not very much and not very high, but at least it is outside. One end is attached to an eye bolt outside the shack window positioned so I can reach it with the window open wide. From there the wire runs down to the workshop. I am not exactly sure of the length, Google Maps suggests just short of 30 feet. So pretty short! But it works, and the FT450D internal tuner can cope with it from 30m to 6m, except for 12m. Anything lower than 30m has SWR approaching 3 but otherwise it seems to be ok. Pity the sun is being naughty right now but it gets a good spread across much of Europe on 20m.

The wire feeds a 9:1 (I think!) unun and a 1:1 choke, then RG58 (*) to the patch panel. The cable needs to be routed yet as it is currently just running through the open window.

Signals received are definitely stronger than with the loft wire and Z11 tuner but it is hard to say exactly how much because of the time taken to swap leads, disable or enable the internal tuner, and tune the wire during which time the watched FT8 QSOs have ended and moved etc. However, with the TS2000X on and connected to the loft wire a random Italian FT8 signal was -17 on the TS2000X + loft wire and -9 on the FT450D + new external wire. Plus the waterfall on the two rigs definitely shows the advantage of the external wire. In fact, the wire length is probably similar to the existing loft wire.

So now I really must get round to running a wire up the garden which will be a lot longer.

(* Typical. I put RB58 instead of RG58, didn’t proof read, turned everything off for work and just got back to edit it now!)

New (to me) arrival in the shack… something I’ve been toying with getting for some time now. This is a TS2000X and has the 1296MHz option installed. It will take me a little while to get acquainted with it but it brings more capability on 2m and 70cm and adds 1296. I’ve been using an FT817 on VHF for FT8 etc. up until now and generally running it at 2.5W – fun, but I wanted a bit more juice up there.

The Signalink from the ‘817 is ready to go, just waiting for a CAT cable and some coax plugs as I’ve used all the ones I need. That will at least give me time to read about the rig before playing…

I dug my poor dusty MD380 out yesterday and charged it up. For a while now I have had Fusion and POCSAG on the pi-star but I rarely use the FT2D and when I do it’s only for APRS. So I thought why not get DMR back into the pi-star. I have a dual DVMEGA HAT with pi-star set to duplex from when I was fiddling with the new HAT.

Anyway, since I last used the MD380 I changed the rx and tx frequencies in pi-star to the ‘designated’ hotspot frequencies so the MD380 needed reprogramming.

That’s where it all started to go south…

First off, since I rebuilt the Windows PC I did not, for some reason copy across the MD380 programming software. Ok, found it on the web and installed it. Can it see the TYT programming lead? Nah. This particular lead is basically a wire, it does not have a chip built in apparently and so needs a specific driver. Oh yes, Windows will see the lead and knows what the device is but has no clue about the driver.

Ok. Found a driver. Installed it – apparently – it does not give an error but the rapidly disappearing screen on installation did not fill me with confidence. And no, it doesn’t work.

Tried another programming software suite called editcp (https://www.farnsworth.org/dale/codeplug/editcp/) in case that included the driver. Nope. The disk that came with the MD380, which took some searching as it is one of those small CDs and has nothing at all written on to indicate it is from TYT, also has the driver but it would not install either.

Right. Over to Linux. It saw the device and lsusb shows a reasonable text string. The same editcp software has a Linux version. And, typical of Linux it just works. It read the radio, allowed me change the frequencies and programmed the radio again, no issues (*). I did have to remember just how to set up the codeplug as I wanted to use both TS1 and TS2… that took a bit of head scratching as it’s been years since I fiddled with this. But a couple of useful websites (there are many others) helped sort my brain out and I now have pi-star and the MD380 set up for TS1 and TS2 with different talkgroups on each via the Brandmeister self service.

(* OK I’ll admit it took me five attempts to program the frequencies as I kept reading it wrong! FOUR times entered wrong before I got it typed in correctly)

(edited 27/7/24 15:34)

I recently came across Meshcom (https://icssw.org/en/meshcom/) which broadly speaking is an amateur radio off grid messaging network like Meshtastic.

I had a 433MHz Heltec V3 with Meshtastic installed but as there is no 433MHz activity nearby it has just been sat in the abandoned project pile aka the corner of the desk. So, I thought, here is an interesting use even if there are no nodes nearby… yeah, that’s always a bad start.

Anyway, as yet the instructions are not as polished as those for Meshtastic and I struggled to get anywhere, routinely managing to lock Windows up completely when using the esptool. When I finally got the thing to look like it had connected the downloader instantly gave an error and packed up. It was then I noticed there is a web flasher (url) so I tried that.

Off to a bad start, the flasher reported that the serial port (the device is directly USB connected) was not ready, and no amount of whinging and gnashing of teeth fixed it. Unplugging and re-plugging the Heltec in made no difference. Three times, same thing. I connected it to the Mac and checked that the esptool.py there could see the board in case I had killed it. It found it and gave the correct output. Back to Windows and the web flasher, and on the fourth failure I pressed the reset button on the Heltec and then the installation succeeded. 5th time lucky. [Subsequently I flashed a TBeam device first time and with no issues, so the Heltec may be a bit odd]

Some configuration is needed which can be done by commands typed in via the web flasher which also has a console function. After setting the location and wifi access the node appeared on the home LAN and also announced itself to the world via the dashboard at  https://srv08.oevsv.at/meshcom/# and after a while it sent all the location information etc via the Internet.

The Heltect only has its tiny spring antenna right now as the 433MHz collinear has and rx LoRa APRS gateway connected. As this is really just yet another experiment it may get repurposed again anyway, but it is an interesting project. Time to read.

27/7/24 I repurposed the TBeam unit that was my LoRa APRS rx gate and was basically doing nothing. Meshcom uses the same spot frequency of 439.9125MHz so it needed to go. This has GPS built in and appears to be communicating – it has its location and altitude etc. plus time The node will go back in the oft on a collinear but once there I have no way to communicate with it because for some reason the onboard web server will not work and the Meshcom iPhone app is beta and on Testdrive which I do not want to install on my phone. I may need to run a USB lead down to one of the Pi systems assuming I have serial comms enabled on it. [15:34 rebooting fixed this – I was sure it had rebooted since I set some parameters but rebooting via –reboot cured the missing web server!]

Certain websites are becoming terrible these days and for no reason other, I suspect than poor or non-existent testing.

I am actually trying to buy a cooker hood. I made a shortlist of 4, one poor, three ok. I went with one from a company but when I got to the ordering screen it wanted to add nearly £15 for delivery. This had not been made clear before. So, the second choice was from Amazon, free delivery. I tried to order this in Safari on the Mac but, after logging in and adding the OTP code I got a blank screen. So I tried via the DuckDuckGo browser – same thing, a blank screen. Plenty of code, pages of scripts and obfuscated stuff, just nothing rendered. I resorted to the Amazon app and this time Amazon told me the product cannot be delivered to our address. No explanation.

Anyway, goodbye Amazon. Again. I actually ditched Amazon when as a Prime user paying annually they decided they were going to inject adverts into films even though I was paying for Prime. I ended up paying the delivery charge with the other company.

KLM is another infuriating site. I can do everything but select a flight when I use Safari or Brave on the Mac. If I want to actually fly with KLM I need to use the DuckDuckGo browser. Go figure.

I used to work with web designers and project managers, good ones, and so I know a bit about user testing and stuff. The fact that this appears severely lacking in some quite major websites is amusing, but with no way to feed back to said sites exactly how useless their sites are it will not improve. Oh well.

Anyway, none of this is radio related!

After our trip to Europe and Japan I really want to do FT8 or other digimodes – my preferred mode because of naff hearing etc – rather than just take handhelds. But the weight of the FT818 + tuner + PSU + a laptop is prohibitive. True, the PSU and tuner would go in the checked bag but I feel that the FT818 would really need to come as carry-on. And anyway I do not own a laptop, plus the fact that we tend to travel light so there is very little room for my stuff anyway.

So it got me thinking. At home I use one PC for Linux, one for Windows 10, the Mac mini, and a Pi. Four systems, four screens, and great flexibility as all the audio is interconnected by a mixer. But the Windows PC is now playing up and is too old for Windows 11, not that I want that but I guess it is inevitable as running Windows without security patches is just about the worst thing I can imagine. Well, ok, my imagination runs a lot deeper than that, but you know what I mean.

The QRP Labs QMX+ looks most interesting, small and portable, and with the QRP Labs heritage ought to be a really useful bit of kit. So it got me thinking, is it time I invested in a small Windows laptop, junk the current Windows box and use the laptop connected to the screen for here plus simply unplug it and go portable? I mean, it’s not like I’ve not done that before and it is a pretty standard model. It’s just that, all my life I’ve used old hardware, secondhand or chucked out, cobbled together to make it do what I want. I’ve never had a new system other than the Mac Mini and its MacBook predecessor, the latter not eve being mine anyway. It is, therefore rather uncharted territory, purchasing a new or new-ish laptop which will not see a great deal of use as I really prefer the Mac. Or do I find an older MacBook and cobble that together? I have an older 11″ model with a dead battery – I changed the battery before but the replacement didn’t last. That was going to be sold but I no longer trust eBay as suitable for someone that sells very little.

So, the choice, should I go this route seems to be a new(-ish) Windows laptop that will take Windows 11, or a cobbled together old MacBook which I already have but could potentially sell to offset the cost of the Windows one.

Either way I will investigate the QMX+ first because the weight of my other HF transceivers is the killer.

We just had a short trip to Europe and then Japan and so I took a couple of handheld radios along. For Europe the CEPT licence was fine but for Japan I arranged a short term one via JARL who were very helpful.

So, licences and HTs in hand, or, rather, in carry-on bag, off we went. I had broken the radios down into battery, radio and antenna and put each in a plastic bag to keep everything clean. Big mistake! My bag was searched at the UK airport and in Schiphol. The bags were causing confusion, especially as, at Schiphol they asked how the battery related to whatever the Mouser code was on the old plastic bag I had used. Very quick though, it only added a couple of minutes each time and they were happy.

So I left the radios intact for the onward trip from Schiphol and into Japan. No issues at all at either end or on the return.

So… note to self, leave the radios intact.

I wanted to get a programmer for the ID51. So, off to RT Systems as I have some of their programmers already. Got the Mac version of the package for the ID51 but it does not allow programming via SD card. That caught me out because their package for the FT2D does and I wrongly assumed this would too.

Of course, it needs a cable! Off to eBay… an allegedly suitable cable arrived today but the RT Systems program will not find it. Typical, it needs one of their enfangled cables that are incompatible with the rest of the world. No way that will arrive in time for our trip.

So… Chirp then. Downloaded the Mac version but it, too will not see the eBay cable. I know the Mac sees it so why won’t the software? RT Systems has no option to select a USB port, Chirp does but offers no help as to what it is.

Try Linux? Hmmm…  I’m really not into fiddling with flatpack or any other enfangled package managers – apt is enough, no apt, no use.

Windows then! Got Chirp, and it can find the cable, and Chirp will see it and will download from the ID51. All good then? Nope!

This exercise was to enable me to program up all the foreign repeaters ready for the trip and also to figure out why our local repeater was in the list incorrectly. So let’s take a step backwards now…

I had tried to update the list of repeaters by downloading from Icom. This method works fine by loading the relevant file onto an SD card and telling the radio to load it. RT Systems take note – why does your FT2D software allow this but the ID51 software does not? This should be standard!

But the resultant list was missing the local repeater. I tried to add it but then the radio tells me there are no repeaters at all. I tried downloading the file from repeaterbook which at least knows about the local repeater. No joy. I tried editing the files already on the radio but each time it somehow messed up. After many attempts and permutations I had to give up which is why I though getting an actual programmer would help.

Enter Chirp (on Windows, yuk…). It downloaded from the radio fine but when I looked at the repeater list it seemed broken – all it had was a list of callsigns and no other information. Ugh! But then I remembered that the radio had whinged that there were no repeaters. So, I loaded a known good repeater list file already omitted the SD card, checked the radio would see repeaters when DR is pressed, and downloaded that from the radio into Chirp. The repeater list is just as broken.

Ok… so I exported everything from Chirp as CSV and opened in LibreOffice. But it does not export the repeater list so I am no further towards understanding how the data works. Time to hit the ‘net. Meanwhile, I checked the current repeater list to see if the ‘Near Repeater’ option was working… nope… 20 minutes in and the radio still has not got a GPS location, just like the FT2D!

Fiddling further with Chirp I found someone had kindly created a list of repeaters that does include our local one. I downloaded this and tried to load it into Chirp but it just says the file format is invalid. I copied it to the SD card and tried to import it into the ID51 – data error. Ok, this has gone far enough, this is not rocket science but it seems that it is made so.

Let’s try Icom’s own software, apparently called cs-51plus2. This is called cloning software and, of course is for Windows. After installing the software it read the ID51 ok. I found the repeater list under ‘Digital’ and, indeed, the local repeater is missing. The big question is, can I add it? Ok, I added the  local repeater to the end of the list of repeaters, adding the frequency and offset, plus coordinates. Coordinates are degrees, minutes and seconds but I had digital so had to convert – not a biggie. Wrote the data back to the radio and rebooted. And… the repeater is there! And it keys! Just had to wait for the radio to find the GPS – I reckon it does this by carrier pigeon – but finally the ‘Near Repeater’ function works and also finds the newly-added local repeater. Now to add the foreign ones…

So what is the takeaway from all of this? I wasted money on RT Systems software by assuming it would work the same as the FT2D software i.e. use the SD card. Never mind. I tried Chirp but that just confused the issue. And the manufacturer’s own software, generally looked down on, worked fine! (Apparently the software comes with the radio, I don’t remember it having a CD and there is no sign of it but I downloaded it from Icom just fine). I must remind myself if I ever get another radio to first try the manufacturer’s software as my needs are generally lightweight, otherwise try Chirp, and last get the RT Systems software after making sure (by which I mean email them as it is very muddled reading the literature) whether it needs their special cable – and, personally, if it does then I won’t buy the software. I mean, come on, I pay for the software and I am forced to buy a lead I’ll probably only use once or twice?

“Zilog has called time on the Z80 CPU.” (https://www.theregister.com/2024/04/29/opinion_z80/) Wow. Actually I had no idea (through never having checked) that it was still being produced.

And a fine chip it was too. I never built a system from wires up using the Z80 though. My first system, designed, built from chips and wire-wrap was an 8080 system, hand programmed to control al x-ray diffractometer. This was decades ago now but I still remember it, although I have no photos unfortunately. The system had a timer chip for a 1-second count and was interface to a Nuclear Engineering (I think it was!) counter that used nixies.

But I did at least use Z80s, just they came as boards. The first was a Transom Triton computer and by then I was programming in Turbo pascal – back then this was really neat as one could have procedures full of assembler code which made interfacing easy. Later I used Gemini boards and that also gave the ability to have a graphics card. By then my interfacing to the diffractometer included a stepper motor and shaft encoder to control the arc motor.

In the end there were two sets of Gemini Z80 boards, one for the x-ray diffractometer and one for an optical microdensitometer. Both gathered data and were interfaced to a mainframe computer for the processing using a suite of Algol 60 programs. Good old days…

Personally my first system was a 6502 Newbear single board, followed by the ubiquitous Nascom 1 which was, of course, Z80 based.

Farewell, Z80…

I now have a Raspberry Pi set up on one of the four monitors in the shack. The original layout was two screens at the top on Linux, then bottom left on Windows and bottom right – central to where I sit on the Mac as the main screen. But that layout had two major issues…

I use a program called Barrier to basically act as a KVM switch for the three systems with the Mac as server. That way the Mac mouse and keyboard controls any of the systems, although it can be awkward sometimes where Windows expects keys which Apple doesn’t have. But Barrier does not understand dual monitors and so moving the mouse up from the Mac got to the Linux box fine, but moving it down from the left hand screen would not get to Windows as the program does not see it being physically there. I could live with that, except for issue two…

The main issue was with the Linux screens being at the top and thus making me sit back or crank my neck upwards, not a good position.

So…

I got to realising that although I use both screens on the Linux box for radio stuff this tends to be with wsjt-x on the right screen and pskreporter on the left.

The solution, which somehow never occurred to me, was simple. Move all the wiring about so that the Mac is right and central, Linux is to the left at eye level so no neck ache, Windows is top right because I rarely use it anyway, and that left a dead screen top left. Enter a Pi 4B. So now I can arrange the four screens with pskreporter top right, Hamclock top left, wsjt-x bottom left and logging bottom right. QED.

Another new toy, a Quansheng UK-K5(8). Size-wise it’s taller and heavier than the Baofeng UV5R and has a nicer display. The reason I got this, apart from the price is there is firmware available to enable a spectrum display among other things.

I’ve loaded this version: https://github.com/egzumer/uv-k5-firmware-custom/wiki. Loading was simple enough. There have been mentions that one needs to insert the programming cable very firmly but I found that my aftermarket Baofeng USB cable went in quite easily with a click. Once I had remembered you need to turn the radio on in program mode the firmware went in via the Edge browser in just a few seconds.

Other than that I have yet to use it on air and it has not yet been near a spectrum analyser so no idea how good or bad it is RF wise. There’s plenty of information out there already anyway. For me, like the UV5R it’s a handy little radio that won’t hurt too much if lost or damaged, but it is the available firmware that seals the deal.

I may have the chance to visit Japan for a couple of weeks this year. I was hoping to take a couple of handhelds with me and to apply for a license. The information how to acquire a license as a non-resident foreigner seemed straightforward until I got to the requirement that any radios need a ‘technical standard certification number’ without which each one needs to be certified at a cost of around £30. It took some searching but I eventually found a document about this including the mark applied to equipment.

Neither my Yaesu FT2D or Icom ID-51 carries that mark. The Yaesu came with a printed booklet which discusses various conformances, the Icom has a PDF which says very little on that score. Presumably this is because of the market – I remember reading that some radios which can be wide banded cannot be if they are Japanese spec ones.

Both the handhelds are, of course radios which are made in Japan, and say so, but it seems I cannot bring them to Japan without incurring costs to have them certified. Both carry the CE mark. I am not (yet) convinced that I read everything correctly but that is what seems to be the case.

So, a radio-less holiday then! At least with an upcoming trip to Amsterdam I can just bring any radio and use it under the CEPT agreement. I am determined to use one outside the UK at least once!