# Antenna tuner design

In a related question about a broadband antenna for SDR it was suggested me to build or buy an antenna tuner.

Since maybe it was not clear to everybody, I'm designing an antenna tuner for reception, not transmission.

Searching around I concluded that most designs are variations on two main ideas.

The first is presented, among others, in a nice video by Stan Gibilisco but applies to wire antennas, whereas I would like to have a loop or dipole antenna.

The second is presented here although is designed for transmission and not reception.

As far as my understanding goes, the one from Gibilisco is essentially a tank circuit defined by L and C that can be tuned to a specific resonant frequency and the inductor is used for impedance matching with the radio.

The other is a series LC circuit (T-network) but I don't understand why a second capacitor is needed. There is a sort of explanation on the link but I can't wrap my head around it.

Furthermore I would like feedback on what I'm thinking of building:

The loop (about 1m diameter) is placed outside and is directly connected to a 1:1 current balun (ferrite ring) that feeds to a coax cable going inside. The variable capacitor (50-1000pF, two gang connected together) is in parallel with a rotary switch that can insert different inductors to have different tank circuits for different bands. A variable inductor is in parallel with the tank circuit and gives the impedance matching for the radio (since I don't have one or I will need to buy one this will be inserted later in the circuit).

1. Is it a sensible design?
2. If 1., is the value of the last inductor (the "matching" inductor) relevant?
3. Is the balun going to change its transformer ratio (that right now is 1:1) when different inductors are switched in the circuit?

4. Related to the second design (the VK5AJL design), why there are two capacitors? Wouldn't a L or tank circuit also work?

So, yes, to tune a resonant antenna, you'll need an adjustable capacitor or inductor. You'd usually go the capacitor route, since adjustable caps are smaller, cheaper, and more exact, usually.

Bonus: There's electronically variable capacitors! They're called varactors or varicaps (or just variable capacitance diode), and they're a very mature (read: old & well-tested and used everywhere) and thus, cheap, method of tuning receivers.

You probably have heard of what a diode does in reverse bias: it doesn't conduct. Of course, that means that there's then two conductive pieces isolated from each other. Since the blocking layer is relatively thin in a diode, these work as a capacitor. By changing the voltage bias, you can make that layer thinner or thicker, and thus, you can change the capacity of that "accidental" capacitor with a DC voltage. Awesome!

When you design a diode for exactly that purpose, you can make sure that this effect is very well-controlled. These diodes then are sold as varicaps/varactors.

http://www.radio-electronics.com/info/data/semicond/varactor-varicap-diodes/circuits.php has a few very minimal circuits illustrating the usage of varicaps. The back-to-back configuration is what you usually encounter:

simulate this circuit – Schematic created using CircuitLab

Now, when you look at a typical small loop (that's actually a specific term for an antenna that is small in comparison to the wavelength, not generally for small antennas) antenna, you'll notice it's usually simply a loop, in parallel with a capacitor.

Now, a loop is a coil is an inductor. Inductor in parallel with capacitor makes resonant circuit. The frequency of resonance is

$$f = \frac1{2\pi\sqrt{CL}}$$.

Regarding practical varicaps that you can buy:

well, look in the places where you'd buy electronic components. I've got good experiences with Mouser, shipping was 5–6€, I think. The BB 640 series (datasheet) looks like a good match for you – nearly 20 times the maximum capacity than minimum capacity means that you can change the resonant frequency over a range of $$\sqrt{20}$$ (see above formula for resonant frequency). If the maximum achievable capacity is too low for you, just put multiple in parallel.

Just throwing together a few numbers: I looked for an online one-turn loop inductance calculator. Found it, and put in a radius 0.5m loop made of 0.5mm diameter copper wire. Got a 2 µH inductance.

Then I said, ok, let's do hundred turns of that. We'll end up with 200 µH inductance. I'm neglecting nonidealities a bit, so we'll throw in a security margin of 50% afterwards.

Now, according to

$$f = \frac1{2\pi\sqrt{LC}}\implies C = \frac1{(2\pi)^2 f^2 L}$$

and plugging in our L = 200 µH and f = 150 kHz, we get a maximum needed C of –including some margin– 6 nF.

For f=1 MHz, we'd need 0.13 nF = 130 pF.

That's bad news since 6 nF is 85 times the maximum capacity of the BB 640. That means you'd either need

• 85 parallel pairs of BB 640, so 170 BB 640 in total,
• a lot more turns on our coil (which isn't the worst solution, by any means!), or
• go up in frequency.

We see that f contributes square-inversely to the needed capacity. So, if we double the frequency we want to reach, we need to take one fourth of the capacity. Since our reliable capacity range is 19x, we can only accomodate frequencies of ca 4.2x the minimum frequency.

We can, however, switch capacitors in and out of our system – and that allows for higher frequencies, should we need them! Lower frequencies need linearly more turns of coil or linearly more varicaps.

You bring up the adjustable air capacitor – of course, that's the easiest solution here, but judging from the mini-whip, I was assuming you'd be looking for a non-mechanical, remotely operatable system.

• Marcus, a varactor never would have worked on my HF loops, because of the limited capacitance range. Mine was like 10 to 200 pF (or more). – Mike Waters Jul 28 '17 at 16:18
• Isn't a normal air capacitor easier? I mean, no batteries and the one I have goes from 25 pF to 500 pF or 50-1000pF if the two gangs are connected together. What is the typical range of a varactor? – Luca Jul 28 '17 at 16:23
• Mike, Luca, stand by, explaining/expanding in my answer :) – Marcus Müller Jul 28 '17 at 16:46
• @MikeWaters 200 pF isn't really a problem – the BB 640 does 69 pF easily, so just put three pairs in parallel (a 2€ solution :) ); the 6 nF that a meter-radius loop would need with 100 turns to go down to 150 kHz is challenging, but it's also an order of magnitude larger :) – Marcus Müller Jul 28 '17 at 16:58

For your loop antenna concept - no, it is not a sensible design but some slight tweaks will bring substantial improvement.

A balun is not a universal impedance transformation or choking device. When it is presented with largely reactive loads, its characteristics differ from ideal. This often results in core losses and in worst cases, destruction of the balun in a transmitting configuration.

A small loop antenna exhibits very high inductive reactance and very low radiation resistance. The variable capacitor is normally placed on the loop side of the diagram in order to be most effective in countering the inductive reactance.

The tuning will be quite sharp so you may wish to design a remote tuning feature for the variable cap.

There are several loop modeling tools on the Internet. Once you have modeled the antenna to determine its complex impedance, then the balun and antenna tuner can be configured to match. Doing it prior to modeling is just taking a shot in the dark.

• I'm sorry but I don't see how this addresses any of my questions. Putting the variable capacitor on the loop side doesn't defeat the idea of having multiple inductors to tune to different bands? Can you expand on that? Secondly, no I won't transmit with it, it is a SDR receiver, I will edit it into the question to be more specific. – Luca Jul 28 '17 at 8:50
• I updated my answer to address your above comment and to reflect the changes to your question. – Glenn W9IQ Jul 29 '17 at 12:47

I use an SPC (series parallel capacitor) matchbox, or transmatch as designed by Lew McCoy W1ICP (SK).

You're not tuning squat. You're matching the transmitter to the antenna system. - Feedline, traps, baluns - everything between TX out to antenna input. The antenna is or should be tuned, cut to frequency, resonant. So whats the problem simple change frequency - change the radiating element - make the dipole longer or shorter whether that dipole is a wire strung between trees or the one on your tribander. What to do? Put a transmatch on the feedline at the TX output and make Zs = Zl - or as close as you can. And don't worry theres no such thing as a 1 : 1 SWR, that would mean 100% efficiency. You are going to have to waste power in the form of heat (3412 BTU per kW) GI and '73 KF9F

• Hello and welcome back Rich. Your post makes some great points, but it doesn't really answer the question, which is probably why it's being downvoted. That's a downside of this site being a question-and-answer site rather than a forum-style chat site. By the way I took the liberty of fixing your all-uppercase text, which does remind me of the stories of the days when telegraphers had uppercase-only "mills", but in the online era it tends to come across as shouting. – rclocher3 Jan 22 at 21:02