Ah; we now need a transmitter to go

with Rick Campbell’s great DC receiver.



The prototype of the VFO for 7.000 – 7.040 MHz; built from KØIYE’s design.


26th January 2016. Yes, I’m afraid so. We managed a couple of QSOs with our existing 4-transistor crystal-controlled 5W QRP tx, but it only covers about 7.0275 to 7.0340 MHz. This little tx was fine when we were using our TS-590 as a receiver; but trying to use an unfamiliar receiver – especially a DC rx that has signals both sides of the basic frequency – was tedious rather than difficult in view of the narrow xtal coverage. We’re actually quite used to using a DC rx, for we’ve possessed a lovely Ten-Tec Century 22 for many years. However, that offsets its VFO automatically on receive so that you can choose the side-band you prefer, without having your VFO idling in the background on one sideband. And in any case the C22 will tune 6 whole bands – not just a few KHz on one of them!


So: the die was cast. A 5 Watt QRP transmitter for 7 MHz, VFO controlled. That’s the project. Otherwise, how can we exploit the virtues of Rick, KK7B’s superb DC rx?


Where to start? Well, the VFO is the obvious point. It’s crucial; if it isn’t stable you’re sunk before you set sail, as it were. 8^)   So where are the good VFO designs, carefully worked out by experts over many years? We’re not into digital electronics, so we need a good analogue VFO – which means vintage designs. So we also need to have all the original components available. There are many places to go, and I did so; but no secret is made of my admiration for the work of Frank Harris KØIYE, whose fine book ‘Crystal Sets to Sideband’ is freely available as a .pdf, thanks to his generosity in having it published on-line. You can find it at:




If you have the slightest interest in short-wave (= Real Radio), listening-in on the short waves to broadcast or amateur stations, home-brewing radio gear, perhaps even becoming a radio ham yourself with your very own unique call sign (mine is G4LQF), you’ll thoroughly enjoy this eminently readable and well-informed book. IMHO, you should download it immediately. It’s full of hints, tips, examples, explanations, practical circuits, how to become a radio amateur: not to mention the typical wry humour of the radio ham. We have learned much from it already. So Frank’s general design of a 3-stage VFO (oscillator, buffer and amplifier) was adopted.


Of course, you aren’t supposed to make an analogue VFO running anything above 5 MHz, and a much lower frequency is to be preferred. But what the heck, we thought. Even if it didn’t work very well, it would be valuable experience.


I can’t be bothered to transcribe all of Frank’s circuitry into TINY-CAD, and hesitate to crib the diagram from his book – but you will find the circuit in Chapter 10, page 260 of his book. It uses a 2N5484 FET as an oscillator, another as a buffer, and the common, cheap transistor 2N3094 as an amplifier. We had tried some of his VFO circuits before with success; but now, this is to be the Real Thing, so the prototype was built very large (I have big, clumsy hands).



It looks awful, but it works. It’s all pretty simple really. This is a scan of the oscillator. The FET is a 2N5458. There’s a 5.1 V Zener diode at the top as a voltage stabiliser. There’s a 470 pF between gate and source, and a 270 pF between source and deck. These two capacitors are usually equal, but I was attracted by this arrangement, found in a G3RJV circuit, and was used as the original VFO for the KK7B receiver, in which application it was just fine. The coil has 23 turns on a T-50-6 toroid. The wire gauge is relatively unimportant. I used a reel of wire that was to hand: it was 0.48mm diameter, which would equate to 26 SWG (British Standard Wire Gauge), or 25 AWG (American Wire Gauge), so what the heck. SWG & AWG are pretty much the same at this level!


The series capacitor in the gate of the FET is 100 pF, with another 100 pF going down to deck, plus a 25 pF air-spaced variable bandset trimmer. The caps are all Russian surplus silver micas, available on ebay very reasonable, especially if you buy more than 5 at a time. It you buy 10 at a time, they’re even better value for money! The solder blobs at bottom left are where other capacitors &c have been put in for trials. The final version will be much neater! 8^)    Mind you, the above lash-up works fine. Of course, the coil will be epoxied down, &c., &c. Oh yes: Frank advises the use of single-sided PCB. The effect of heat on double-sided PCB is to make a generally-distributed universal coupling capacitor of the whole bottom of the board! Stand-offs should also be used to space the VFO board from the metal box in which it will – ABSOLUTELY MUST – ultimately reside. There must be no VFOs lying en plein air on your bench. Except when you’re prototyping them, naturally. 8^)



This board is joined to the oscillator by a couple of soldered ties. The first transistor is another 2N 5458, the buffer. It’s followed by a common 2N3904 transistor as an amplifier for the now-buffered frequency. In the VFO we made for the KK7B DC rx, the VFO didn’t need a buffer – though the purist would have surely included one. We didn’t, because in a receiver, the VFO only has to provide a ‘steady state’ signal. When we tuned up & down our narrow band of 7.025 – 7.040 KHz with our receiver, the load on the VFO hardly changed – if at all. So the single transistor VFO we used was fine for that undemanding job. However, this VFO is for a transmitter; specifically, a CW transmitter. So when our key is up, the VFO will just be happily idling. But when we press the key, the VFO will be called upon to provide some power to drive our transmitter. Unless our VFO is buffered, the drain of power will de-stabilise the VFO every time we press the key, and the result will be very bad. Our transmitted signal will change in frequency with each dot & dash. It will ‘chirp’, as the old timers used to call it. It would sound like Rudy Wiedoeft’s Laughing Saxophone, if anyone who reads this stuff has ever heard of that. Which I very much doubt. Right-click the link anyway, & open in a new tab.




Well, we wouldn’t want our Morse code transmissions to sound like than would we? Or at least, I hope not. Of course, many signals used to sound like that. Even when I first started on the bands (which was not all that long ago – the early 1980s) there were many signals like that, and they not only chirped, they also gradually changed frequency as well, so that you had to follow them down or up, whichever was the case. In some cases, our QSO would gradually drift across another, better-regulated QSO, descending from above, and finally exiting below them. Or above them: whichever.


No. We really do need to keep our transmissions a spot-on frequency as we can manage. Hence the prototype we made today. It seems very promising; so we will do more work on it. It will only have to tune 7.000 MHz to 7.040 MHz, so the use of varicap diodes will probably be the best way forward. Also, it will need a small offset (or maybe an onset?) so that we can receive signals without hearing the VFO in the background. It will also need to be made smaller, so that it will fit in the die-cast box we used for the earlier version. Above all, it will need to drive the KK7B DC receiver as well as our new transmitter!


Crikey – there is so much to be done…




Page started 26th January 2016.