A 'Modern Vintage" 3.5MHz ATU.
 

I'm gradually getting things together with my AEL3030 transceiver - sourcing some crystals, a 6-pin Jones socket for the power-lead, rewiring the mic socket to match the Pye standard, all with the assistance of a manual supplied by forum member G3KVM_Roger (thanks!).

To put it on-air I needed an ATU to match its 50-Ohm output to my random-length end-fed. Theory, confirmed with a borrowed VNA, said the antenna appeared capacitive, so should be matchable by a series-inductor. I hooked-up an old 'roller coaster' coil to the VNA and found that something like 25 turns was needed.

So, off to the 'vintage RF parts' junkbox - which yielded a nice ceramic 3 inch diameter Eddystone ribbed coil-former of a type I recall from 1930s RSGB Handbook ads - with an existing thick copper winding, along with a 6-position heavy-duty ceramic rotary-switch, made by General Electric, whose first career was probably aboard one of Uncle Sam's larger WWII aircraft. The switch would select taps on the coil for fine-tuning.

Putting it all into a plastic box (plastic so as not to detune the coil - I've seen loads of commercial ATUs where the metal case is indecently close to the coils) I connected it to the antenna, and supplied a whiff of RF through my trusty Bird, the idea being to probe the coil with a cric-clip-lead to find the resonance-points between 3.5 and 3.8MHz, then make more-permanent connections with leads soldered between the coil and the tap-select switch.

Ah... with the full coil in circuit it was resonant spot on 3.8MHz! My calculations were out - I needed more inductance to get down to 3.5MHz! But how.... I pondered shoving a few lengths of ferrite-rod into the centre of the coil but started to fret about hysteresis-heating, Curie-points etc. Then I had an idea - the antenna being a capacitive impedance, what about making it a bit more-capacitive? So I fished-out a Jackson "C803" wide-spaced variable capacitor and hung it between antenna and earth terminals - I was delighted to find that adding 45pF got the thing to resonate nicely at 3.55MHz - but, alas, the Jackson capacitor was too big to fit inside the box!

There _was_ room though for a smaller 50pF variable capacitor. One with much smaller plate-spacing, intended for receiver applications. I worried about flashover, but then remembered that in plenty of transmitter designs shown in ARRL/RSGB handbooks a 'receiver type' variable-capacitor was used at the output of pi-networks in transmitters running hundreds of Watts.

Try it and see... Yes, it tuned-up fine, so I tried it with 100 Watts of continuous carrier. No flashover! I'm happy!

Just to prove a point, I left the transmitter running [on an empty frequency] for five minutes of continuous carrier, then powered down and felt all over the ATU for any signs of heating. No problem...

If only I'd known at the start I wouldn't have drilled the box for the big switch, and would have used the space for the larger Jackson capacitor.

[Photo shows unit in part-built, test state]

Re: A 'Modern Vintage" 3.5MHz ATU.
 

Interesting bit of "cut and try" design there!

Just a thought about the possible flashover at the antenna side of the ATU. That would obviously depend on there being enough volts to produce the flashover, and in turn that would only occur if the resistive part of the impedance were to be high. You note the VNA told you the antenna is capacitive, but you don't mention what the resistive value was? Given that its not flashing over with a 100W its probably in the few hundred ohms region at most - or it might even be close to the 50 ohms your Tx needs to see - if the SWR is actually low on the Tx side of the ATU. Did you measure that?

Richard

Re: A 'Modern Vintage" 3.5MHz ATU.
 

Good point - I don't remember the resistive component the VNA gave - should have taken a photo of the screen rather than writing-down what I thought I needed to remember using a biro on the back of my hand!

But the suck-it-and-see approach seems to work OK - no flashovers or things-getting-hot, unlike a friend who built an ATU using similar 'design' ethics and ended up with a crazily-high 'Q', so high that one end of the coil unsoldered itself due to the massive circulating RF current.