VK5AJL Home page.Project index.Information index.

BUILD AN HF ANTENNA TUNER
MAKE A HOMEBREW HF-VHF ANTENNA TUNER
(It really isn't all that difficult.)

This page details several T-match tuners. By comparing them, you should be able to construct a tuner to your own particular requirements. Now that a number of people have built them and provided feedback, several things to look out for have been identified.

THIS PAGE

Design considerations

recommendations to avoid problems

Suggested layout

more recommendations to avoid problems

Tuners Wavelength
capability
Power
handling
tuner type
(comments)

Tuner mark 1

80m - 2m

100W PEP - 30w carrier

just a simple tuner

Tuner mark 2

500m - 70cm

400W PEP - 130W carrier

includes inbuilt SWR meter

How they work

Very brief description giving only the basic idea.

Possible traps

things to look out for - possible problems identified in design variation.

Downloads

Various helpful files.

Using a T-Match

How to get a quick tune using a T-Match.

DESIGN CONSIDERATIONS

WHAT WILL YOU USE IT FOR

Normal use will be HF and perhaps some circumstances in VHF bands. I have tuned my HF antenna to 70cm but only as a test.The losses are too great using a tuner with coax so it is recommended to include a simple balun (details below). Using coax with a tuner will gain you nothing except a happier radio. This point can not be stressed enough.

DON'T GO OVERBOARD

Many build tuners and consider a thousand places they can use them and design for all of them. In the end, the thing stays connected to one amplifier or radio and to one antenna or type of antenna. All of these complications introduce possible losses and possible problems. IT IS STRONGLY RECOMMENDED TO KEEP IT SIMPLE STUPID. A single balanced output is recommended. If you really must use a long wire referenced to ground, a simple way of using the same output terminals is decribed below.

TUNER TYPE

Whether you use a T match, Z match or any other kind of matching network doesn't really matter much. Neither coils nor capacitors consume power so, provided resistive components are kept small, there should be very little loss. The major power loss in the units described here is by resistance in the wire that the coil is made out of although there will be others.

Since this tuner does all the impedance matching needed, nothing more than a simple 1:1 current balun is required and strongly recommended. Overall, insertion loss in tuner MK2 is between 0.1 and 0.15db on 10 metres matching roughly 500 ohms and less on the lower bands using such a setup. (Into a dummmy load.) This is less than if a 4:1 voltage balun is used.

POWER LEVELS

The power the tuner will handle is only relevant in terms of the voltage separation and coil resistance (current carrying capability). The voltage separation may be 300kV but if the coil is made of 0.5mm wire, it will handle only a few watts. On the other hand, making the coil out of 6mm copper tube is a good idea at any power level but that doesn't mean the voltage separation will allow more than a few watts.

The maximum voltage will also depend on what impedance you are going to match. This is not just some absolute value but includes an unknown quantity of reactance but, in general, the voltage across the input capacitor will be somewhere near the output voltage.

Table 1 - Recommended minimum voltage separation for given power levels
Please note - this table is only a rough guide and includes capacitors and all other components.

Power Impedance to match Power Impedance to match
pX or pY50Ω100Ω200Ω500Ω1000Ω

pX or pY50Ω100Ω200Ω500Ω1000Ω

10W32v45v63v100v141v500W224v316v447v707v1000v
15W39v55v77v122v173v700W265v374v529v837v1183v
20W45v63v89v141v200v1000W316v447v632v1000v1414v
30W55v77v110v173v245v1500W387v548v775v1225v1732v
50W71v100v141v224v316v2000W447v632v894v1414v2000v
70W84v118v167v265v374v3000W548v775v1095v1732v2449v
100W100v141v200v316v447v5000W797v1000v1414v2236v3162v
150W122v173v245v387v548v7000W592v837v1183v1871v2646v
200W141v200v283v447v632v10000W1000v1414v2000v3162v4472v
300W172v245v346v548v775v15000W1225v1732v2449v3873v5477v

The voltages in the above table are PEAK voltages not those necessary to produce the power listed. Voltage separation for capacitors, and all other components, will not depend on the mode of operation ie. FM, SSB, AM etc. 300W PEP = 300W of power for FM or AM. 300W PEP = approximately 100W of power on SSB however 300W PEP on sideband will have the same peak voltage as 300W of carrier. The amount of compression used on SSB is also irrelevant.

Table 2 - Air gaps required for voltage separation of various components.
"Flat" means a highly polished flat surface while "point" is a sharp point.

GapkV spacingGapkV spacingGapkV spacingGapkV spacing
(mm)flatpoint(mm)flatpoint(mm)flatpoint(mm)flatpoint
0.251.400.951.506.092.202.7510.413.454.0014.604.70
0.502.421.201.756.972.453.0011.253.704.2515.434.95
0.753.381.452.007.842.703.2512.093.954.5016.265.20
1.004.301.702.258.702.953.5012.934.204.7517.085.45
1.255.201.952.509.563.203.7513.774.455.0017.915.70

Flash over will vary on several properties other than just shape such as pressure and humidity. These gaps should also be used only as a guide where the only gap is air. Tracks on PCB should be a lot further apart because of surface flash over.

Table 3 - VERY APPROXIMATE conductor thickness v power level.
Power Impedance to match Power Impedance to match
pX or pY50Ω100Ω200Ω500Ω1000Ω

pX or pY50Ω100Ω200Ω500Ω1000Ω

10W32v45v63v100v141v500W224v316v447v707v1000v
15W39v55v77v122v173v700W265v374v529v837v1183v
20W45v63v89v141v200v1000W316v447v632v1000v1414v
30W55v77v110v173v245v1500W387v548v775v1225v1732v
50W71v100v141v224v316v2000W447v632v894v1414v2000v
70W84v118v167v265v374v3000W548v775v1095v1732v2449v
100W100v141v200v316v447v5000W797v1000v1414v2236v3162v
150W122v173v245v387v548v7000W592v837v1183v1871v2646v
200W141v200v283v447v632v10000W1000v1414v2000v3162v4472v
300W172v245v346v548v775v15000W1225v1732v2449v3873v5477v

The actual power in watts compared to Peak Envelope Power in SSB mode is extremely variable. A value of 1/3 is reasonable but could be anything from 10% for a weak voice to 80% or more if a decent compressor is used.

There are so many other factors such as skin effect, proximity effect, coil diameter, turn separation, conductor type (tube or solid) conductor coating and so forth that will affect the values in the above table. Conductor cross section type will not change the resistance much but will affect the ability to conduct heat away from the used parts to the unused parts.

Voltage is not so much of a consideration because the voltage across the coil can be divided by the number of turns used. Remember though, the voltage across the coil will always be more than across the capacitors. A separation of 1 to 1.5mm between one conducting surface and the next is recommended. Proximity effect will increase flash over values so this 1 to 1.5mm should be plenty.

LAYOUT

IT IS STRONGLY RECOMMENDED A 50 OHM IMPEDANCE IS MAINTAINED RIGHT UP TO THE INPUT CAPACITOR CONNECTION. Both tuner Mark 2 and Mark 3 do this by using a 50 ohm SWR pickup as the input connection but if you decide not to include a SWR meter, the use of 50 ohm coax is recommended. See also GENERAL LAYOUT BELOW

INPUTS

IT IS STRONGLY RECOMMENDED ONLY A SINGLE INPUT SOCKET IS USED AND THAT IT IS A BNC CONNECTOR. To use more involves switches and wires that may cause problems. The tuner will match impedances after the input capacitor but not before.

A PLC-259 or BNC connector will both handle 500 volts. Since the input side will be 50 ohms, powers of up to 5002 / 50 = 5kW are possible with these connectors. Current at this power is 10 amps if transmitting carrier but only 2.5 to 3 amps using SSB (based on the average human voice without compression). A BNC will handle this fine.

OUTPUTS

IF YOU USE A TUNER, THE ONLY COAX YOU SHOULD USE IS BETWEEN IT AND THE RADIO OR AMPLIFIER. A tuner is only suitable to two types of antenna viz. 1) a single long wire using ground as the counterpoise and 2) a balanced antenna of some kind. Even though this is an unbalanced tuner, there is little point using an unbalanced output (coax). The SWR on the input side of the tuner may be 1:1 but on the output it is the same as it was without the tuner and losses the same. Since coax over 1.5:1 introduces too many losses, IT IS STRONGLY RECOMMENDED A 1:1 CURRENT BALUN IS INCLUDED in these units especially if the long wire option is chosen because it is more easily bypassed than a voltage balun. See balun options below.

BALUN

Note first, the terms "voltage balun" and "current balun" are as defined on this site. So long as there are delusions that current and voltage can be transformed (or transferred) as separate entities, this controversy will exist. Voltage baluns have transfer losses while current baluns don't other than the resistance of the wire. A 1:1 CURRENT BALUN IS RECOMMENDED.

CAPACITORS

The usual tendency for amateurs is to get something too big. It is totally pointless getting 3kV spacing capacitors if the rest of the tuner is limited to 100 volts. In fact, if the rest of the tuner will only handle 100 volts, there are advantages in getting capacitors of the same rating.

The largest voltage, other than across the coil, is likely to be across the input capacitor which is usually smaller to slightly bigger than at the output terminals. If you have a PL-259 on the output (not recommended at all) then there is no point getting big capacitors. 500 volt spacing is fine because that's all a PL-259 will handle. IT IS HIGHLY RECOMMENDED THE SMALLEST (IN PHYSICAL SIZE) CAPACITORS ARE USED FOR THE POWER YOU NEED. There will be less losses because of stray resistance and/or inductance. Maximum capacitance can be anywhere from 350 to 500 pF. Minimum doesn't matter so much because there always must be some.

COIL

The coil is where all the power is consumed so the thicker the better. Silver plating this component is also a very good idea. For high power it is advisable to wind the coil with 6mm copper tube. For 160m use, the inductance required will be about 30uH. For 80m and above 15-20uH while 40m and above 8-10uH. IT IS STRONGLY RECOMMENDED TO MAKE TO COIL NO BIGGER IN INDUCTANCE THAN REQUIRED. An approximation calculator in the form of an XLS speadsheet has been provided in downloads below.

GETTING THE BITS

The general population seems to be dumb these days and people don't do things for themselves or make things anymore so getting the tuning capacitors and other parts was like finding a needle in a haystack but I eventually got a couple. Be prepared for a search in your own city. In Adelaide Australia, they are available from Cookson Controls.

GENERAL LAYOUT

Recommended layout.

1) The two connections to the capacitors should be made at the same end of the device. This will reduce series inductance and therefore losses especially at the higher frequencies.
2) The shorter used parts of the coil (least inductance) should be kept to the front of the enclosure to keep the connections to the switch as short as possible.
3) Because of 1) and 2) above, the capacitors should be connected towards the front of the box.
4) A BNC or N-connector are recommended, the BNC for lower power tuners and the N-connector for larger power levels.
5) If a SWR meter is included, the micro-stripline needs to be as long as possible so having the capacitor input connection at the front is an advantage. This should be a 50 ohm micro-strip (6mm wide). If the SWR meter is not included, 50 ohm coaxial cable should be used especially for the higher frequensies.
6) Grounding the of the coil away from the switch grounds both ends of the unused part of the coil. This will increase losses. This is better than having self oscillations because of capacitive coupling between the turns. It will also prevent having massive voltages on the unused free end on the higher frequencies. The magnetic influence of any winding will not extend more than a few turns keeping losses down.

TUNER MARK 1


The circuit is fairly simple but construction a lot more difficult so this page will concentrate on that aspect. The box is made from PCB material off-cuts.

MAKING THE COIL

The coil was the most difficult to get right. I first searched the backyard and found a piece of water pipe of about the correct diameter (50mm). I got some 1.25 mm winding wire from an old transformer and straightened it out. To do this simply hold one end in a vice and the other with some pliers in your hand and give it a SMALL stretch. Sometimes this works on wire you want to re-use but it can crack the insulation and degrade the voltage specs so be careful. For this application it doesn't matter.

I then wrapped the pipe up with a couple of layers of paper and cling wrap. I placed a DOUBLE winding along the pipe for 20 odd turns. Only 14 or so are needed for 80 metres but I gave myself headroom. The ends of the wire were stuck down with tape. The turns were then carefully pulled together. I then used some polystyrene body filler and ran a couple of lines along the coil spaced at 1/3 intervals along the circumference. These must not be too wide because they are used to hold it together only. Air makes a much better dielectric. Once dry, I pulled one winding off breaking the polystyrene as I went but leaving enough to keep the other winding in tact and evenly spaced 1.2mm apart.

Once the middle winding is stripped out, I ran another line of polystyrene along the same places as I did before and CAREFULLY slid the coil off the pipe. Once I removed the inner paper, the coil slid easily back over the pipe. I could then use a power wire brush to strip the insulation from along one gap between the polystyrene so the band connections could be made for the switch.

PUTTING IT ALL TOGETHER

Construction is exactly as shown in the picture above. The coil is held down with hot melt glue using a small standoff made of plastic. Since the tuner will do all the impedance matching needed, only a simple 1:1 current balun is needed to convert to balanced line.

The only thing now is the front panel but, before that, it is a good idea to first to polish all the external exposed copper. A coat of clear lacquer will keep the unit looking nice. The front panel was drawn up in a few minutes with PhotoShop. The panel design was then printed out and laminated with an extra piece of paper behind it. When it is cut out, the back lamination comes away and the paper side can be simply glued to the box with suitable glue.


Completed unit

Costs:-

Capacitors $80
PCB box about $15 depending on where you get it.
Coil - nothing
Switches, chassis sockets - $15
Knobs etc. - $5
Powdered iron toroid for balun $10
TOTAL $115. How much did you pay for your tuner?


TUNER MARK 2

THE NEED TO CHANGE

The problem I had with tuner mark 1 was only on 160 metres. It worked very well at everything from 80 metres to 6m and I used it for some time on all bands between. On 160 metres space limitations mean my antenna simply isn't high enough to have a reasonable pattern to work DX but that's not of concern to a tuner.

A new tuner housing had to be made to accommodate a larger coil. This housing was still less than enough to make micro-strip SWR pick-ups suitable for 40m or less so an opamp was used to amplify the signals from the pick-ups. This can be switched in or out as required. On about 20m (10 watts) and up to 2m (200mW), the op-amp doesn't need to be used. On 40m to 160m, the op-amp can still be avoided but calibration of the meter for correct SWR reading can't be achieved. On the other hand, by switching to REFlected, an antenna can be tuned by simply finding the minimum reading.

The inbuilt SWR meter follows that on my HF radio quite closely. On some bands it is a bit out but if it reads less than 1.1:1, the one on the radio is less than 1.3:1 or vice versa. {uThe linear amplifier is more happy when the inbuilt SWR meter on the tuner is used rather than the one in the radio.}

MAJOR CHANGE

The major change was the size of the coil. This was increased to 32 turns of the same construction. The first take off point is ½ turn from the start of the coil and moves in progressively larger steps from one to 8 turns for the last. New taps are at 1, 2, 3, 5, 7, 9 12, 15, 18, 26 and the rest of the coil for a 12 position switch. These figures are only rough and a tune can usually be obtained on any band on several switch positions.

With feeder length adjustments, I now have a better than 1.2:1 tune on all bands from 160m to 2m including WARC. In fact, using the full coil, I can get a reasonable tune by ear to listen to the local AM radio stations in the high kilohertz.

Layout showing SWR pickup on input.

NOTE: The electronics (op-amp) must be shielded from the rest of the tuner. The PCB panel between them achieves this. The output from the SWR pickups must also be RF bypassed on both sides of this shield.

SWR METER AND SWITCHING

Wiring diagram for inbuilt SWR version
SWR Amplifier circuit

At left are the wiring and circuit diagrams of tuner MK2. A 3 pole 4 position switch selects either direct input from FWD and REF pickups or amplified signals from FWD and REF pickups. In the direct mode of operation (1st and 2nd positions) the battery is not used and switched out. In the 3rd and 4th positions the amplifier and battery are switched on.

The opamp can be any dual general purpose opamp. Some of the resistor values may have to be changed depending on the input impedance of the opamp used.

One half of the opamp (1a) is used to flash a LED every few seconds to indicate the unit is on. Ra and Rb can be used to select the duration and flash time of the LED.

The second section of the opamp (1b) is used to amplify the signals from the SWR pickups. It is simply connected as a non-inverting amplifier. Rc needs to be selected depending of the efficiency of the pickups and the power used.

Connection points between the switch wiring, battery and amplifier circuit are (A) switch to battery, (B) battery negative, (C) opamp input and (D) opamp output and are shown on both diagrams.

The inbuilt SWR meter is very handy and works better than the one in my radio. Both these meters read slightly differently no matter how long the patch cord or whether the linear amplifier is in circuit or not (between radio and tuner). If I tune up using the radio's SWR meter, the solid state linear trips the protection at about 1/2 power. If I tune up using the meter inbuilt into this unit, I can get full power out of the linear without drama.

COMPLETED UNIT

Completed front panel

The front panel was made by printing the desired text and graphics on a printer. It was then laminated with a backing piece of paper so that the lamination would on be on one side of the front panel. This facilitates gluing the front panel onto the front of the unit using PVA woodworking glue. Because of the laminating material this takes quite some time to dry but will dry sufficiently over a few days.

The knobs were made by cutting out some sheet fibreglass with a hole saw then making and gluing the scale in the same manner as the front panel. This was then glued to some commercial knobs using epoxy. See downloads below for some graphics you can use for making such knobs all in one file.

The scale on the SWR meter was made in the same manner.


Graphics for the knobs can be downloaded here as .tif images (670kB), as .jpg images (412kB) or as both .tif and .jpg (1Mb)

HOW IT WORKS

This description is not truly accurate and very brief. Perfect components have been assumed and both source and load impedance largely ignored. A full description involving all of these other things is too complicated for this page and the values outside of the tuner are not known anyway. IT IS ACCURATE only on basically how it actually works.

I have heard people describe a tuner as just a low pass filter. Since the impedance of a capacitor decreases with frequency and the impedance of an inductor increases, this just doesn't make sense. This T match should work better as a high pass filter. In fact, it isn't used as any sort of filter at all. It is used as an adjustable impedance matching auto-transformer.

For this example the frequency will be constant and can be anything. For a different frequency, component values change but the effects are the same.

Example components

Suppose the coil has an inductive reactance of 500 ohms. Suppose also capacitor C1 is chosen so it has a capacitive reactance of 400 ohms. The current through the capacitor is 180 degrees out of phase with the inductor so the radio side of the tuner will see an inductive reactance of 100 ohms. If the inductive reactance between D and G is also 100 ohms, points A and D will always be at the same potential.

That is to say, since A and D are always at the same potential, they are virtually connected.

Similarly for the antenna side of the tuner. Suppose C2 is chosen so that it has a capacitive reactance of 100 ohms. The same thing applies between E (chosen to be 400 ohms away from G) and C so the antenna side will see an inductive reactance of 400 ohms. E and C are virtually connected.

Example components

At some frequency where component values achieve the reactances as described above, we now end up the equivalent circuit shown at left (bright traces).

If it looks like and duck and quacks like a duck, its a duck. If it looks like an auto-transformer and acts like and auto-transformer, its an auto-transformer. In this case 1:4.

Extra notes:-

1) In the diagrams above, only the impedance has been considered, not the number of turns. The number of turns between G and D will be the same as between D and E. (Double the turns quadruple the reactance).
2) In this example, the voltage at point B (with respect to G) will be 2.2 times the input voltage (D to G). The voltage at point C will be double the voltage at point D.
3) This works just as well if C2 is removed and a direct connection made between E and C and with C1 left in place. Similarly for C1 and connection A to D with C2 left in place. It will also work just as well with both C1 and C2 removed and connections A to D and E to C made.
4) Because the load between C and G may be capacitive or inductive, the value of C2 could vary from this value to achieve the same 1:4 match. Likewise C1.

USING IT

Many tuners have a 4:1 balun included in them. This is a complete waste of time and energy because this tuner does all the impedance matching necessary. All that is required is a simple 1:1 current balun with sufficient turns. This will have a lower insertion loss than anything else. From there on this combination can be used to drive a dipole either centre or offset fed or a sky loop.

To tune a band:-

1) Set the capacitors half way then rotate the band switch until the maximum volume of noise or incoming signal is heard.
2) Adjust the two capacitors in turn until the same max volume of incoming noise is found.
3) Find a free frequency (by asking) and select a mode such as FM or AM (RTTY on some radios) where a carrier will be emitted, and turn the power right down.
4) Select REV first on the SWR meter and, while emitting a carrier, fine tune the two capacitors in turn for minimum reading.
5) Calibrate the SWR meter on FWD then set to REV and take the final SWR reading with further fine tuning if needed.
6) If you really can't get a good tune (better than 1.2:1) then try moving the band switch up and down one position and repeating steps 2) - 5)

There is no antenna or band I have not been able to tune in a few seconds with a bit of practice. If you really can't get a good tune within 20 seconds, you have something wrong.

Top of page

All text and images on this site are Copyright to John Langsford (vk5ajl).
You may provide links on other sites or use the information and pictures for your own personal use.
You may use the text or images for redisplay or quotation provided you acknowledge the source ie. vk5ajl.com.
I think that's pretty fair, don't you?