GNURadio fun…

I have a had a real hard time trying to get GNURadio installed. The distro failed every time when running gnuradio-companion. This isn’t GNURadio’s fault, it may well be because my Ubuntu desktop has been upgraded a few times via dist-upgrade without it being a fresh install. Perhaps there is some old crud left in there even though I tried my best to remove all old versions of GNURadio and associated files and folders. After constant errors no matter what I tried it seems that my Ubuntu 18.04 installation somehow grabs GNURadio compiled against a previous boost library 1.58. So I resorted to the pybombs method which has installed a functioning GNURadio-companion despite errors with apache-thrift. Sadly, when installing gr-iio via pybombs it failed, again seemingly trying to refer to boost 1.58, rather than the 1.65 that the 18.04 distro has.

So… I purged libboost and grabbed the 1.58 code from sourceforge and set it off building. There are lots and lots of warnings but it did compile and install. Trying again to install gr-iio failed – it seemed to try to install boost itself and then whinged that the version is wrong. All deleted and purged again, still using the pybombs method and making it compile everything rather than installing the binaries. This time the uhd cod went in with no errors and GNURadio-companion runs ok except it has no rtl-sdr. Installing RTL-SDR did not help as a source file is missing… Installing the gr-iio package for the pluto also worked fine, and the FMComms block is there along with others such as PlutoSDR al categorised under Industrial IO.

After downloading an example .grc file that some kind person put on the web it works! I can see 70cm and see the result of a cq test call. Much to learn yet but at least the pluto is working in receive with GNURadio.

As an aside, while waiting for pybombs to sort everything out – and it took ages on uhd – I installed sdrangel which has a .deb package that works with Ubuntu 18.04. After figuring out that it installs itself into /opt rather than /usr I fired it up with the pluto attached and I can make the pluto receive on 70cm.

I’m sure the lesson here is sort out your Linux installation rather than continually fiddling with it like I have. When I get some time I plan a fresh installation of 18.04 and getting GNURadio from the distro to see if it works that way.

Project 444…

…the slowly ongoing saga. Typical. I wired up the 50-way D connector and connected it all up and … nothing. On checking voltages half was missing, i.e. -80 was there but no +80. The barretter was open circuit on one side. Huh.

A new barretter arrived via eBay and so I now have the PSU fully working. A quick check of the 444 showed it to be no better than before, which I’d expected but I really wanted the correct power supply to verify that. So I connected up the TDMS to send data to the 444 and with the keys set to SMSMS it reliably prints Y – pity that should be R! Fiddling with switch combinations is equally confusing.

So, progress, of sorts. However, the TDMS itself needs work yet because the circular trace is far from a circle, or indeed not even close to anything which might be imagined to be circular.

Sleeping teleprinter

It’s beginning to look serious now…

I just need the time to solder up a 50-way D connector to connect the interface unit on the right to the 444. The interface has the +/- 80V supplies inside as well as all the interface logic for the various teleprinter functions. if I nick some volts out of it I should then be able to feed in the test signals into and out of the TDMS and use those to align the 444.

New toy… (updated)

My latest toy arrived on Monday, a TDMS5BV telegraph distortion measuring set (TDMS). This guy (hopefully!) can generate a test pattern and all sorts of other magic to help test circuits (not particularly relevant) and also teleprinters (definitely relevant).

It tests polar relays too, thus the socket on the front. I have a PDF of the manual and schematic for it. I now also have a variac so I carefully brought the TDMS up to voltage while watching the current and all seemed well – no exploding electrolytic. But it did nothing. After bothering to read the manual it gave instructions on how to prepare the instrument. After a fiddle with a control at the back it produced the required circular(-ish!) trace and the polar relay started to cycle. So far, so good…

Weather… and all that sort of stuff

Well so much for me tinkering with the satellite dish to receive QO-100, the dish arrived and the weather arrived too! We had days of very strong wind and horizontal rain and then when the weather got better I had railway stuff on. Sorting the valves out just made it worse as the dish is now buried in the workshop. Anyway, the valves are leaving next week so I will have space (Space for space, get it? Ok, I’ll get my coat…)

This is the pile of valves to go so you see the issue facing me workspace-wise. There must be 1,000. There is half as much again behind the boxes. This finally rids me of any non-CV marked valve and all are going to a better place (no, not the dump!). And yes, those are 5 of the 21 VT98-type valves I imported. The guys at the depot managed to get all 21 crates into my (then) Astra hatchback – they were a lot better at packing that I am and I’m pretty good!

Projects, projects…

I seem to be collecting projects but making no progress on any of them! Well, ok I have five on the go, of which one is started and one nearly, and three are in boxes still.

Since Es’Hail-2 (QO-100) went up I’ve been planning to set up a receiver for it, and later on hopefully a transmitter as well. One thing at a time (or in my case it seems no things at any time!) So, I have an acceptable LNB, not a good one but should suffice until I get a more decent one. I now have a bias tee, some relevant connectors, I have a roll of CT125 satellite cable somewhere and I am now waiting for a 120cm satellite dish and pole to arrive. If this arrives today I may even get it working by tomorrow, otherwise next week.

The Creed 444 is the one project that has at least had some progress. For that, I now just need to wire up the 50-way D connector to the signalling unit and then it should type to itself, otherwise it will need adjusting as that’s all that is left if the text is still garbled.

Then there is a box full of bits to construct a Minitiouner receiver which will receive ATV (and will hopefully receive it from QO-100 too). All the bits are there ready to solder up.

And there is a box of bits to make a low power 5.6GHz ATV setup. All that needs is a box!

And finally, a box of bits to make up the QRP Labs 40MHz QRP transceiver. I’ve only had that since the National Hamfest – of 2017!


More rattling about…

Well I fiddled with the teleprinter some more, wondering if I could at least get it to print with just 30-0-30VDC – it needs 80-0-80. Nothing. I tried connecting directly to the relevant pins for the receiving relay but could not feel any resistance when this was applied. Odd, because I could measure its resistance with a multimeter, so what’s up.

Strangely the next time I tried I could no longer measure the resistance. Might have been me. Then it occurred to me that on the 444 everything comes out on the 50-way D connector and you can get directly at the receiver coil, and I can see the relevant ohms. So, re-wiring the lash-up and still with 30-0-30V and with a resistor in series just in case and it burst into life. Almost.

It did print, or rather tried. It got most of the letters wrong, but is at least trying. Most likely the voltage is just not quite enough, or my lash-up resistor is denying it sufficient current but I’d rather not risk the coil right now.

However, and more of a concern was that although it went through the motions, the carriage did not move. It was stuck fast and no amount of fiddling would move it. After pouring over the workshop manual there seemed to be a lever in the wrong position. Moving this freed it up and it does now move with keypresses, even though the letters are wrong.

This lever was in the wrong place and could not move. Freeing it fixed the non-moving carriage issue.

Nearly there then… the ribbon is not rotating but there is at least one broken spring so hopefully that’s an easy fix. All I need now is a proper TU.

The Creed lives

Getting to grips with the Creed 444… everything seemed fine and it all rotated ok by hand so with the help of eBay I got the proper power socket and wired it up today. Things did not go entirely to plan…

1st attempt. Lots of loud noise from the motor. Well, at least it spins… the motor has a height adjustment at the rear which engages the worm with the drive gear at the front and this was completely undone. Adjusted…

2nd attempt. The lights went out! Hmmm. So, a tiny strand of copper wire had linked between the earth and live on the plug. Yes, the same one I had just soldered up carefully. Grrr.

3rd attempt and it runs. Key presses work the sequence, the WRU works too. As yet no print as I need to put a ribbon on and some paper in, but more importantly sort out the data side of things. I wonder what the WRU says… I could work it out by hand but I’ll leave that until I get it to print. Equally, I wonder if I can find a blank one to put my callsign into… now that would be an ideal candidate for 3D printing! (can I use this as a reason to buy a 3D printer…)

The motor and gear makes a particular whirring sound and a useful Youtube video of someone testing one of these beasts sounds the same, so that’s good. There is a rattle which I need to locate but I still need to take the unit apart anyway and make sure everything is properly oiled.

I wonder if the kids will believe that we used to use similar machines as input/output terminals on the mainframe at work in the 1980s!

Good old days…

So I now have a Creed 444 teleprinter. I’ve been after a teleprinter for about 2 years now. When I was still at school I used to buy, strip, repair and sell Creed 7E’s – maybe 7 or 8 passed through my hands and on to others for pocket money. Back then I had two 7Es for myself, as well as a tape reader, tape punch and a valved TU hanging off a B40 / B41 setup. Weight was not an issue in my basement workshop… But I’ve never had one of the more modern units like the 444 (yes, modern is a relative term here!)

Having manhandled the beast from the car to the workshop I now need to learn how it is put together so I can be sure it’s not going to fly into bits when I power it up. However, hopefully I’ll get it running and then get an interface set up. I doubt it will be allowed in the shack (aka the little bedroom) so it may well have to run in the workshop – but that does give the excuse to acquire a nice old comms receiver to go with…