I received a few reasonable images early this morning between 5 and 6am. Hopefully later passes will yield better images. These were received on a 2m FM set with a loft mounted collinear, and decoded by qsstv on Linux.
So… having an Allstar node I wanted to configure it for Echolink. It seemed this was just a matter of editing the pre-prepared configuration file echolink.xxx and renaming it to echolink.conf. Edits in place, this I did. Oh yes, and I set our broadband router up to forward the relevant ports to the hub. On restarting asterisk it gave numerous errors of the form ‘Error in parsing header on servers.echolink.org’. Hmmm.
Ok, scratching around the web I found that Echolink has a firewall test service at https://secure.echolink.org/pingTest.jsp. It failed. Ugh.
Then it dawned on me (meaning I read the documentation a little better!). I had set a callsign with ‘-L’ at the end which appeared to be the way to go. But this needed separate validation! Once that was done it all sprang into life.
Simple, and also obvious when I realised. Old age?
http://km6uso.net/index.php/2021/02/27/adding-echolink-to-your-allstar-hub/ is an excellent guide – there are others of course but this one pointed out clearly the need to register the -L or -R callsign.
Up until now I’ve used generic QSL cards rather than an actual design. These served well as an M6 and a 2E but I always planned to design a card once I passed the full exam. So, one year on (no sense in rushing!) here it is:
The basic design was done by one of my boys with the final layout and production carried out by fdsQSL who also produced my previous generic cards. The photo is my own, the logo as explained previously was basically why I chose RVB because the Morse fits the wheels. The rear has all the usual stuff including an area for ‘Via’ above the target callsign, both top right to hopefully make it easy for the bureau.
I have only just got these and have yet to try one in the laser printer – which would make filling the details in easier. But given I hardly send any cards that is a lot of faff when a pen does just as well!
I only managed one almost decent image from the 1st December ISS SSTV event, the others being partial or very noisy.
I am not bothering with the diplomas this time. When I checked most of them use a Google form that required me to log in (twice in one case), failed to submit and, in one case set 40 (!) cookies for the trouble.
Slowly coming together. Yesterday I decided to attack the front panel with a drill and mount the rotary encoder, switches and the little Arduino board which I programmed earlier. This is for tuning the Langstone. For some reason my drill press insists on making triangular holes – if I wanted a triangular hole I’d never manage of course. So I’ve resorted to making a smaller hole and using a round file. Anyway, everything went into place, although the Arduino board has no mounting holes so I’ve tie-wrapped a piece of plastic under it as an insulator and used a decent (hopefully!) sticky pad to secure the board inside the front panel.
So far, so good. Here it is receiving the Allstar microHub…
I need to sort the microphone out. The USB audio dongles seem to be constructed for stereo input so I wired the same to the front panel. Plugging the headset in gives no audio, presumably its all shorting out. I can make it work by ‘adjusting’ the plug (pulling it out until it works!) so I need to re-wire or make a little adapter.
There are a couple of fans in the case and so far it seems to be keeping nice and cool. Next steps include making the GPIO breakout board and possibly the band switching stuff.
Edit: of course I have a mono to stereo adapter plug, I just never looked in the junk box! Mic input now working fine…
After temporarily giving up making my own Allstar microhub I opted to purchase a ready made one from G7RPG. It arrived today and simply works – plugged into 12V and it fired up and announced the IP via radio (subsequently I grabbed the MAC address and gave it the IP I wanted it to have via DHCP).
I ran an audio test on it fully expecting it to be perfect and it was indeed. So whatever I did when making mine is nothing by comparison. I shall investigate that at a later date but having a Pi in a box doing nothing has advantages as I am looking to make a GPS referenced NTP server.
I finally got round to putting my Allstarlink node together after looking at it sitting almost completed for a week. Poor thing. Anyway, it’s now functioning but I want to add the LEDs.
Software-wise it turned out a bit of a faff. I had already signed up and got a node number and set a password etc. My first attempt was via the Raspberry Pi image downloaded from the Allstar wiki. That seemed to go in just fine with a fairly easy setup and well scripted information on the wiki. All seemed ok except for when I wanted to install Allmon… the instructions for which began with the need to install git. That failed and so I did the usual update / upgrade cycle – which I really ought to have done right away as the image is quite old. After that, nothing worked. The USB interface was not working and so there was no radio functions. Power cycling did nothing.
So I installed the hamvoip image. One nice thing about the Pi and similar thing is you change SD cards and this changes the o/s and everything. Hamvoip went in fine with a fairly automatic installation and after a couple of loops where I’d missed something it now works fine. It comes with Allmon and Supermon installed. (the Allstar image does not have a web server and so the Allmon installation would have failed anyway even if I had not already given up).
Not picking on the Allstar image – had I taken more time I would have set up a Raspbian system and manually installed the Allmon software onto this, as detailed in their wiki. But, having got hamvoip working I will stick with that for now.
So… now to go off and figure out exactly why I built this thing anyway!!
Edit: Ok, I spoke too soon. Having connected to a test server the audio that comes back, allegedly the same as what I sent, is awful with a tone and buzzing in the background which the audio only just makes it over. Something is wrong… audio from the box is fine with the generated voice announcements perfectly clear, as well as a few calls heard when connected to hubnet receive-only. The HT is fine as I can hear good quality audio via a scanner. Hmmm…
Edit 2: I’ve experimented with every parameter and made no difference. I have increased the mic gain on the HT as it seemed low even with a high setting in the simpleusb setup. But the noise remains. It seems that whenever I transmit anywhere near the node it happens, and simply moving to the other side of the shack cures it to a great extent. Currently the radio is connected to a dummy load but swapping for a small antenna makes no difference. Moving the dummy load well away from the node with an extension coax makes no difference to the noise if transmitting near the node. So the transmission is clearly getting in somewhere that it should not. I have verified the HT is ok by monitoring on a scanner, so no issue there.
Edit 3: Adding a ferrite around the cables between the radio and the CM108 has helped a little in that I can get closer to the node and not get this noise played back in parrot mode. Unfortunately I have put all my small clip-on ferrites in A Safe Place, ever to be seen again, so I can’t add more right now.
I have carried out a tactical withdrawal i.e. the node is sitting under the desk in disgrace. No amount of ferrite’ing, re-routing wires or even running the node from a lab supply makes any difference to this noise. The only thing I can be reasonably sure of is it is some interaction between RF from the HT and the audio chain, so the CM108 I guess. Holding the HT next to the node while monitoring it on a scanner shows no audio issues but playback in parrot mode and it is there every time. Moving away from the node cures it to a great extent but no amount of shielding works. I could (should!) investigate further but I have other projects to take care of plus a bunch of DIY stuff.
It really is taking me far too long to get this sorted out. But, finally some progress has been made… at least I now have a decent case and the screen mounted. So far, there is very little else in the case other than the Lime and Pluto, the screen, the Pi and a 12V to 5V converter. But at least there’s plenty of space to add all the other necessary bits and pieces.
The screen was the most awkward thing to mount as it is designed to fit easily into a plastic surround with the Pi directly behind. For that, the 4 mounting screws fit directly to the case. But here I want it in the front panel. My rather less than elegant solution was to get some 1/4″ square aluminium bar, make holes for 4 M3 nuts and bolts and tighten these to become studs – these are in the horizontal pieces as shown in the pic. These were then epoxy glued to the front panel. Two more bars then take the screen, with spacers so the screen isn’t pulled back too much when tightening.
Anyway, elegant or not you can’t see it from the front! Still to go on yet are the three push buttons and encoder / tuning wheel for the Langstone transceiver part of the kit. As for the rest, well, one thing at a time! But anyway it works… as shown by yet another successful 6″ transmission to the Winterhill box sat on top!
The EMF talk on yesterday’s BATC CAT21 event was extremely useful, as indeed have other talks at previous events. It has all slowly become a lot more clear, in particular the current (and hopefully future) standard setups. What was not clear to me, and probably should have been (!), is that ‘ground’ is actually the level where someone may be rather than the actual dirt. So the first floor in a house for example. At least that’s how I understand it.
Having grabbed the latest (v0.1.2a) RSGB spreadsheet I have done the calculations for my kit from 4m upwards and all is well. The two transverters feeding the big wheel and dipole in the loft both pass as compliant due to their low power, and the colinear on the FTM100DE at 50W is just high enough to be fine.
What I have done by way of recording is to save each completed spreadsheet to PDF, cut off everything except the first page, and then paste it into a document along with information explaining where the antennas are etc. The only more complicated one is my QO100 dish as that is basically at head height behind the garage. As I can see the dish from the shack and as the area is not generally accessible anyway I simply added the note that transmission will cease should anyone enter the area.
As I use a Mac and neither have nor want MS Office the documents are all in ‘Pages’ and stored locally, backed up to Time Machine and in iCloud. I reckon that’s sufficient!
The Mark 2 bit of wet string seems to be going ok with 5W. Well, mostly. Concentrating on FT8, which is my usual mode of operation I can reach the east coast of the US with decent-ish reports, and all across Europe, all depending on conditions of course.
However, it won’t tune too well on 30m – two LED flashes on the YT-1200 – but I’ve made several contacts just fine. It does seem deaf higher up than 15. And no spots yet on 80. Plenty on 60 though, a band I could not get on to at all with the Mark 1 wet string affair.
It picks up major noise below 30m but FT8 is still doing it’s thing.