Homebrew radio construction projects

The amateur license is a privilege that comes with a purpose - ‘self training in radio communication’ as the old UK license used to say, and maybe still does. Electronics construction is an enjoyable part of the hobby for me and fun, though I never have enough free time for it. 

By the way, I’m told you really need only two tools in life: WD-40 and duct tape. If it doesn't move and it should, use WD-40. If it moves and shouldn't, use duct tape. Duct tape is like The Force: it has a light and a dark side and holds the universe together.

Anyway, here are some of my more successful projects ...

QRO baluns

In the US, it’s relatively easy to obtain QRO baluns but not so down here on The Far Side. So, rather than pay an arm and a leg to obtain a commercial balun from W-land via airmail, I decided to have a go at making my own heavy -duty baluns capable of matching my amp’s output to various antennas. The wonderful book “Understanding, building, and using baluns and ununs - theory and practical designs for the experimenter” by Jerry Sevick W2FMI, (US$20 from CQ Magazine) has been my inspiration and source of most of the ideas.  It’s a bit short on practical construction advice though and presents a confusing range of options including many designs lifted from the amateur literature and (mostly) systematically dismantled as a result of Jerry building and testing them. Lacking the ability readily to measure losses, impedance transformations, frequency response etc., I’m simply sticking to W2FMI’s preferred designs.

The next step was to find a source of toroids. Again, small toroids are sold in ZL but not the large ones suitable for QRO power levels. On a recommendation I went to Amidon’s site and found all I need right there. The site has limited technical information - it took some Googling to confirm the specs for toroids mentioned in W2FMI’s book (the “K” material ferrite evidently has a permeability of 290) - but the prices are reasonable and delivery is quick and charged at their cost price.

The first project was a wide-range 4:1 balun using a man-sized T400A2 iron powder toroid, designed to match the balanced open wire feed to my antenna up the hill down to something closer to 50 ohms unbalanced, close enough for a normal unbalanced QRO ATU to handle anyway. I didn’t fancy paying extra for the copper wire and Teflon tubing suggested by W2FMI so I improvised, using the core from a spare length of RG58 coax which is more or less the same size and hence probably about the same impedance and voltage rating. 21 bifilar turns took the inner from about 5.6m of coax and the discarded coax shield will make earth straps for other projects. The toroid costs US$30 ea + post from Amidon. The nice grey gasket -sealed polycarbonate box came from Dick Smith, our local high street electronics supplier, for about US$15. It’s better than a metal box in this application as there is no need for difficult-to-get HV feedthrough insulators on the balanced side. Two ordinary 5mm bolts do nicely, in stainless steel if the box will be used outside. At left are the parts laid out in kit form and the photo at right shows construction in progress. Notice the 12” ruler under the kit: this toroid is 4” in diameter and needs a box at least 5” across to leave room for the windings.

Next I built two 1:1 baluns using FT-240-K cores (US $18.50 ea + post) and more RG58 offcuts. These were even easier to make with no need to strip off the coax outer cover and shield. I use them to feed HF dipoles using RG58. Ten turns of RG58 took about 88cm of coax.

I’m currently using one of the 1:1 baluns for a 160m inverted vee dipole and another for a separate 80/30/17m multiband dipole, made simply by connecting the 3 normal dipole wires in parallel on a common feed point i.e. the balun. It seems to match quite well, at least the SWR is low enough and I’m getting out OK on those bands. The hardest part of making these baluns was making a mechanically strong centre for the antennas: I used ceramic open wire spreaders, attached to the balun box using cable ties (two on the Mark I, four on the new improved Mk II) held in the boxes by jamming the cable tie heads into the box mounting holes from behind. I have since replaced the antenna connectors with 5mm stainless steel bolts and wingnuts. The bolts and SO239 are sealed with superglue which is liquid enough to flow into the tiny gaps and goes rock hard, preventing them from moving even under intense pressure. The same goes for any fingers that stray into the glue.

The hardest part of the balun projects was making neat holes in the boxes for SO239 sockets. Polycarbonate is quite easy to work to size with a drill then a tapered hand reamer, finishing it off with a circular file and (for the round sockets) a small flat file to flatten one side of the hole to match the flat on the socket. This flat, along with the superglue, stops the socket from turning while tightening the PL259 on the coax downlead. By the way, I always cover plugs used outside with a layer of self-amalgamating tape (NOT insulating tape!) to avoid corrosion and stop water getting into the coax.

The toroids are held in place in the boxes using offcuts of high density foam cut to size from an old gardening kneeling pad. The rated efficiency of these toroids (around 98-99%) shouldn’t cause any problems with overheating even at QRO levels, but I’ll check them for heat damage and moisture ingress during the next few months. One advantage of constantly experimenting with new antennas is that there are plenty of opportunities to check the connections!

K2 suitcase set

My most ambitious project to date is an Elecraft K2 QRP radio. It came as an excellent kit with comprehensive instructions and very helpful email reflector for on-line support. It took me about a month of evenings to complete.

My ‘suitcase set’ consists of the K2, some antenna wire and string, a mini paddle, pen and paper, and a small switched-mode PSU to recharge the rig’s internal 12V lead-acid battery, carefully packed into the foam insert and ready to go in an emergency (living in an earthquake zone, that’s no joke).

K2 #1772 worked first time and is a joy to use. The thrill of hearing the first callsigns on a receiver you’ve just finished is hard to beat - working DX with a homebrew QRP transceiver is superb (the K2 comes with built -in mojo)! Mine doesn’t actually deliver 10 watts, more like 6. Most likely, I mis-counted the number of turns on the PA toroids. On the other hand, 6 watts is usually enough and is that bit closer to the “official” QRP 5 watt level. 

The K2 has an impressive specification including 9 band coverage, built-in auto ATU, 10 watts maximum output and an excellent contest-grade superhet receiver. As well as in the shack on decent antennas, I’ve used it portable stylie on picnics and hols in France, Brussels, and New Plymouth feeding wires thrown over convenient lighthouses etc.

Other QRP radios

I've got various homebrew QRP "rigs" (more like patchboards really!) in various states of dis/repair about the place. There’s an OXO somewhere, naturally, and a TEME from Practically Witless magazine. Most of the fun is in building them and after the thrill of making the first few QSOs, I soon lose interest, except for the K2 that is. SPRAT, the G-QRP-Club magazine and Pat Hawker’s Technical Topics column in RadCom are the inspiration for most of those projects.

Station accessories

I have keyers, ATUs, filters, computers and all sorts of measuring gizmos, some of it homebrew, some not. I built a crude wooden box (right) for an AC voltmeter to keep an eye on the mains and discover exactly why the lights flicker when I key the amp and go very dim every so often (left).

The boring grey diecast box below is my version of an IF232C Kenwood radio to PC interface. It uses a MAX232 chip to convert between 12V RS232 and 5V Kenwood standards. I didn’t bother with opto-isolators - no need.  That construction project was finally completed in 2005 but started about 5 to 10 years ago previous. The RS232 connector on the right is a commercial USB to RS232 convertor because the PC doesn’t have an RS232 serial port. At last I can log frequencies directly from the radio and avoid some of my contest logging screwups, I hope !

Remote control ATU

I picked up a pair of big motorised vacuum capacitors at a radio rally in North London a few years ago, and a couple of big motorised roller-coasters too. I'm hoping to finish a remote controlled (manually-tuned) QRO ATU .... if only I can find the time to build it! So far, I've managed to get the roller-coaster stepper motor drive working, and I've started to figure out how to control the vacuum cap (10-1000pF variable plus 3 x 500pF fixed caps with RF relays in series). Unfortunately, the vacuum cap units are earthed through the cap chassis so for a balanced feeder I’ll probably need to leave the chassis floating (a technique I've used for years). I may go for the dual-balanced ATU design published by AG6K in QST Mar 1990, using an air-core balun instead of lossy ferrite. Anyone out there fancy designing an auto ATU with PICs?

Ouch! Large 4WD meets small Daiwa rotator control box

OK, it’s not exactly a construction project - more of a reconstruction.

This mess is the result of someone who shall remain nameless deciding to drive the car across the field when I was assembling a new beam and rotator ready for CQ WW CW 2006. The rotator control box was “hidden” in the long grass under a tree and while I worked on the pole 20m away, that ‘someone’ drove out to the field to bring me a beer. Very nice. But then the ‘someone’ drove off the track and back directly across the grass to the house with a sickening crunch as she passed the tree. She didn’t even notice the new speed bump in the lawn, or the irate radio ham jumping up and down in her ‘lipstick mirror’.

The wheel crossed the front left hand side of the box, crushing the indicator panel and case. First job after we’d collected the pieces and I’d calmed down was to disassemble the box and assess the internal damage....

The rotational indicator is basically clockwork: a series of gears, driven by a small electric motor, turns the pointer and a potentiometer gives positional feedback to the electronics, a simple Wheatstone bridge circuit. 

With all the main bits out of the box, the front panel was buckled and above you can clearly see the bent drive shaft for the indicator/pointer. So, with pointy-nose pliers in hand and nothing to lose by having a go, I disassembled the gear box, straightened the shaft and reassembled it. The motor didn’t work due to its wires being sheared inside the end cap near the commutator so some precision shaping with a scalpel and soldering iron was called for. Amazingly enough, it worked!

After reassembling the innards, I painstakingly rebuilt the plastic box using superglue and a hot glue gun to tack the fragments back together, and turned it upside-down so the worst bits don’t show quite so much. I used clear sticky film to hold the cracked bits of the perspex front panel in place rather than replacing it because it would be too hard to re-draw the compass bearing outer circle by hand. Anyway, the cracked panel tells a story.

The repair cost me a day’s labour in the shack and workshop, and with no rotator there was no way to turn the HF beam for CQ WW CW 2006, so I had a go at 40m single-band instead ... and made a record Oceania score.

During 2007, I bought a shiny new control box for a Yaesu rotator that I’m planning to convert to work the Daiwa, hopefully with a bit more oomph too since there is a fair bit of voltage drop on the long rotator control cable at present (even using heavy duty “trailer cable”). That’s another “one day project” (as in I’ll get around to it one day).

Meanwhile, I’ve fitted 4 x 12V white LEDs (a dollar a piece from Dick Smiths, NZ’s equivalent of Radio Shack or Maplin) behind the display of the old control box in place of the temporary incandescent bulb which burnt a big brown blob into the Pacific Ocean (and they say global warming is a myth!). The LEDs are powered off the 12 -0-12v transformer taps with 2 diodes and no smoothing capacitor. The old control box really is a junker but still it refuses to die!

Now all I have to do is remember to fit the beam pointing the right way ... yup, you guessed it ...

Next “one day” project - a new HF amplifier

I’ve been collecting bits to build an HF amp to replace the LK550 - one day.  Thanks to another ZL ham, I have a chassis, a GS35b valve and a little pile of circuit boards and components, most of which are usable. 

I’m looking for a better ‘centrifugal’ fan to keep the valve cool - the one I have is too weedy according to my linear Elmer Phil.

I bought this meaty HV transformer from someone who bought and dismantled several 5kW commercial transmitters from an airport installation: it weighs about 90 kg and now sits on its own furniture-removal trolley base, capable of carrying 200 kg allegedly.  Should be good for a few kV at an amp or more. 

The capacitors sitting next to it will do a few µF at a few kV and will probably be built in the transformer box along with the rectifiers, soft start, control relays and meter for a self-contained PSU.

The bars currently laid on top will be needed to lift this beast: I’m making a strong steel frame from angle-iron.