What is the best mosfet for box mod



The Balanced Universal ATU

 (PA0FRI's "S-MATCH ATU" published in RSGB's RadCom March 2003 and VERON's ELECTRON November 2003)


How the idea arose.

Micro Metals / Amidon toroids suitable for this ATU.

A 1500 W S ​​match.

A model of the third variation

Home brewed with cleaned and painted components from the junk box.

With this system it was possible to match for SWR = 1 on all 9 HF bands with respective a 2 × 17 dipole and a W3DZZ. Both antennas were feed with approximately 10 m ladder line.

Undoubtedly a G5RV can be matched for SWR = 1.



Several possibilities of the S Match design.


After an extensive research and experimenting I achieved the design shown in the above drawings of the Balanced Universal ATU System. To make it distinctive from other types of ATU's, I call my design S-Match ©. The system is intended to accommodate balanced antenna systems transmitting on a very wide frequency spectrum, ranging from 160 m band up to 10 m band. Likewise, it is able to take proper care of impedances and associated reactances ranging from less than 20? and up to 3000?. Built with the proper components, it is able to transfer loads of more than 1000 watts. This balanced antenna-tuning unit has only three components: a capacitor, a roller inductor and a balun or RF transformer. The input circuits are isolated from the output circuits, resulting in a system, which is also suitable for single-wire antennas and coax fed antennas.



The efficiency of mechanical and electrical a well-built S-Match is equivalent to other ATU's in this ("fig) table. See my Test Symmetrical ATU's.




In the design of one of my ATU’s, the balun using T200A-2 toroid is mounted between two PCB disks and located in between the 50 µH roller-inductor and the 12 - 1200 pF vacuum capacitor.

An antenna like my present antenna system is matched using only 200 pF on all nine HF bands, still my S-Match is built with a 1200 pF vacuum capacitor. My ATU was able to easily match the antennas of a couple of friends. In one case it was barely adequate to match another friend's 4 × 40m loop antenna on 160 m. I had to add an extra switchable 1200 pF capacitor in parallel with the roller inductor to handle any impedance. (see further below)




With some antenna systems a better efficiency may be achieved if the antenna is connected to the capacitor as shown in the drawing by the dotted line. It is also possible to use two separate toroids. I didn't notice any difference between an ATU with two cores and an ATU built with a single toroid.





 Transformers made with Amidon toroids. The T200A-2 toroid is a nearly exact substitute of two T200-2 units sandwiched together.

ATTENTION: Do not twist the black and green wire!

The inductance of the turns is a part of the resonant circuit with the roller-inductor and the capacitor.

Wires are twisted for more thickness.

The rule-of-thumb for the windings is:

ATU for 10 - 160 m: (a-b) = (c-e) in series with (f-d)3 µH

ATU for 10 - 30 m: (a-b) = (c-e) in series with (f-d) ≤ 3 µH

ATU for 30 - 160 m: (a-b) = (c-e) in series with (f-d) ≥ 3 µH

3 µH Is a compromis for 10 m - 160 m!

The transformer contributes a lot to the final result, which is my conclusion after I tested many types of windings. A couple of ham friends successfully tested a number of my own ideas for a suitable HF transformer / balun. The above chosen type of transformation ensures that input and output connections are galvanic isolated. The experiments show that an inductance of about 3 µH is a good compromise for all HF bands. Sometimes the minimum inductance (coil + balun) is too large for some 10-30 m antenna systems. Subsequently, extending or shortening the feeding line was necessary to change the antenna system impedance to cope with the tuner, or build a separate ATU with fewer turns e.g. less than 3 µH on the toroid.

The transformer / balun is part of the tuning system. Most ferrite toroids are not suitable for high power or high RF current while the indicated Amidon iron powder core is.


The sequence of the windings. Left side: blue = input, black & red = capacitor. Right side: black & red = inductor.



The windings should provide high-voltage insulation by using Teflon covered wire. Another option would be to use RG58 coax cable with the outer isolation and screen stripped off. Do not twist the wires of the primary windings with the wires of the secondary windings. In the picture (see fig ”) the light blue with the blue wires.


RZ3AE's S-Match ATU

RZ3AE has sent me an email with attachment: "Your idea for a balanced tuner allows for a very interesting design. A translated part of his Russian article, he wrote:" With the unconventional design and the use of special components a small symmetrical tuner may be created.

RZ3AE uses two ferrite tubes as balun.

The balun in the original design was replaced with two stacked ferrite tubes of high permeability and the roller inductor by a T200-2 toroidal coil with 15 taps. Taps 1 through 3 one turn, 4 to 6 two turns, 7 and 8 three turns, and 9 to 15 four turns evenly distributed over the rest of the toroid.

The balun is suitable for an output less than or equal to 100 W and the wire is insulated with Teflon.

The various amateur bands may be adjusted to 600 Ohm with VSWR <1.1 and 1000 ohms with SWR to 1.3. For impedance less than 50 ohms, the antenna must be connected to the variable capacitor;

See his internet publication: RZ3AE S-tuner



DL7RU has personally experienced that it is not easy to wind Teflon covered wire on a toroid and keep them firmly in place. DL7RU: I have now found a method to do the job much easier. First of all, I start with a couple of tight pulled cable ties on each side of the toroid. Under these ties I put smaller and thinner ones. This fixes the beginning and end of the windings so that it is easier to settle the wires.



DL7RU found out for himself and as he saw it on this website: wrapping a toroidal core with Teflon braid to build an S-Match is not that easy. He writes:

I have now found a method with which larger toroidal cores can be wound much more easily. First I attach larger cable ties for the beginning and the end of the winding to the toroidal core and tighten them well. Under these cable ties I then insert thinner cable ties with which I fix the winding wires (see pictures). So I no longer need to clamp the beginning of the winding in the vice in order to be able to wind the core tightly and the wire lies nicely in the winding direction.

I put a few more turns on the HF transformer because I only want to use the tuner for the 40, 80, 160 m band. The transformer in the picture has ~ 8 µH for the coupling-in winding and ~ 2 × 4 µH for the coupling-out side. I think that should be enough for the 160 m tape. I have already built an S-Match for the higher bands (10 - 30 m band) and I am very satisfied with this concept. I operate (fig ») a ground plane antenna which I feed via a Wireman chicken ladder (CQ 553). It works really well. The antenna is 6.23 m long (5/8 for 28.5 MHz) and so are the four radials.



With some antenna systems a better efficiency can be obtained when the antenna is connected as shown by the dotted line. It is also possible to use two separate toroids and I have not noticed any difference between an ATU with two cores or with a single core.





If any unbalance occurs, try switching to the other connection or interchange the connections to the roller inductor.

My experience shows that some unbalance in an ATU occurs due to the construction and the location of the components. For example, a variable capacitor has frame and movable plates, which are not in any balance with respect to the chassis or system ground. The basic mechanical construction of roller coils is also not symmetrical.

Always begin with matching the system for SWR = 1, to obtain maximum balance. Unbalance is usually the result of inaccurate tuning. With the usual baluns it may look as if the feeding line is symmetrical, but the center taps of the bifilar and trifilar wound toroids force the balance with respect to "ground" or chassis (fig.a).





Some hams on Internet sites state that in comparison tests the S-Match failed regarding balance. The authors brought no evidence whether the balance of the "reference" ATU was tested properly ....




You may conduct proper tests if your feeder line is in balance with real ground. Use the S-match or any other unbalanced tuner for the test. Connect one of the feeder lines to the ATU (fig. C), match for lowest SWR and measure the RF current. Repeat the test with the other end of the feeder line (fig. D). Check and mark whether both currents are equal or not. As a "ground" you can use a single wire in the backyard or a wire to your central heating system.

The ("fig) wiring is also important and it is best to install as showed in the image.






The use of a roller inductor enables tuning only by turning knobs. Some hams don’t like roller coils, not realizing that a switched inductance with many taps has a lot more self-capacitance and self-inductance. Most of the time you can’t achieve a proper adjustment because the needed point is somewhere in between the taps. Still, I provide in this present application a switched inductance of 25 µH, which will be adequate for most of the antenna systems. An extra variable capacitor of 470 pF from a vintage AM receiver can be added in series with the input to help reducing the VSWR.
Another method is with a quasi roller inductor .. This is achieved by switching a large variable capacitor in series with the switchable coil. This ensures a fine control of the self-inductance.



The earlier made notes had been lost and it was decided to repeat the testing of the different designs to determine which one was the best solution. At the same time it was investigated to see if there was even a better solution than those known to-date. A temporary set-up, shown below, was used.

At present my favorite systems are fig 2 and fig 5.



There was hardly any difference in the efficiency of any of the four designs. Afterwards the symmetry became the focus. The settings of the S-Match need to be adjusted for a SWR = 1 between the transmitter and the ATU in order to achieve the best possible balance. When this was done, design figure 2 showed to be the better one for use on the frequencies between 1.8 MHz and 10.150 MHz. Design figure 3 showed to have the best symmetry for the frequencies between 14 MHz and 30 MHz. No sound explanation has been discovered for this difference. Design fig.2 requires slightly more inductance than design fig.3 (which has the antenna parallel to the capacitor). Between 14 MHz and 30 MHz the higher inductance requirement of the roller inductor is advantageous. Please refer to the below mentioned comments of a German HAM on using the S-Match on the 10 meter band.


The capacitor can be a split stator with a grounded stator. If the parts are built in a metal or screened cabinet, they should be assembled on at least 5 cm distance of all metal parts to minimize the asymmetry of the used components.


These components were used for the repeated test of the design.

Test of the figure 5 design. The capacitor is suited for a 400 W ATU but survives a test with 700 W on 80 and 160 m!

These small caps are suited for at least a 100 W ATU.

Final test


All pictures are just illustrative examples of this S-Match design.

This simple ATU with variable capacitors (from a vintage AM receiver) and a T200-2 toroid is suitable for 400 watts (fig »). The matching range can be increased with an extra-switched 100–220 pF / 5 kV fixed capacitor and used when the antenna impedance is below 50 Ω. With my inverted dipole it was necessary to switch-in the fixed capacitor in parallel with the roller inductor only on the 80 m band.



A T-match unit can easily be converted to an S-match by removing the second capacitor or bypassing the first capacitor. The free hole in the front panel of a converted commercial unit can be used for a switch for paralleling fixed capacitors to the tuning capacitor to provide additional capacitance.




S-Match conversion of a MFJ-989D Versa Tuner V. The modification was successful. See topic.


DJ4ZC has made a number of modifications that I have not (yet) tested, but his motivation is plausible.



The RF transformer got an extra compensation winding, making it a trifilar system.«blue) winding is identical to the primary winding. However, the "hot" side is not connected to the 50 Ohm input, but connected to earth / ground and is therefore in the opposite phase. The other end is installed floating. Furthermore, the secondary windings are arranged between these two primary windings. With this method, the capacitive coupling of the secondary winding produces equal currents in the primary, but since they are opposite in phase, they cancel each other out. Thus, all "common-mode" noise picked up by the feeder line is largely suppressed for the receiver. On some bands, this measure At his location the noise level was reduced by around 20dB.

The transformer was made of ferrite material, because ferrite transformers can be smaller, have fewer losses and the string induction and capacitance are smaller than with powder-iron cores. Using the design information in the Philips Application Note ECO6907 and with measurements taken by W0RUN on 4C65 ferrite, it turned out that a transformer with two stacked T36 / 23/15 cores of 4C65 mix is ​​suitable for a maximum of 1 kW.


To remove any electrical charges that sometimes occur on the antenna and cause noise problems, 3 × 330 k resistors were added in series. That number was due to the expected high voltage with its too short antenna for the 160 m band.



I was regularly asked whether my system can be applied as an anode circuit for an RF amplifier. It is also possible, but due to lack of time I was not able to test the circuit with high power to fully support the idea. That is a good opportunity for you to try it out.


As an encouragement for home brewing see in alphabetical and numerical order examples of S-Match tuners. Not everyone is just as skilful but it is most important that it works well.



Kees, PAØLL who is famous for his LL-tuners, sent me photographs of his "LL S-Match". There was much demand for a small portable antenna system including a small ATU. For this reason his intention was to build and sell a handy ATU including the antenna in a carry box. The final product is built with components, which are made especially for the tuner. C = 11 - 470 pF (knob with 1: _6 reduction), L = 34 µH. The ferrite toroid for the balun was specially made by Amidon for this ATU. I expect that all his handcrafts made ATU's ends as "collector items"!


PE1KQP modified his CL tuner to an S-Match. During test on field days it turns out that his ATU outperforms many commercial types.

PE1ADY's experimental S-Match for his 2 × 20 m dipole and 20 m feeder line. He obtained a SWR = 1 at 1.8 - 29.5 MHz.

 4 × PE2B's S-MATCH



The version of PAØMJM's S-Match. The graphics were made with a miniVNA (www.miniradiosolutions.com) program.


Since I published this design, may hams successfully copy this simple ATU.

PA3EGH's S match

Some of the selected emails are presented:


Once again I recreated something from your ufb homepage, an S-Match ATU. I have tested 3 different ATUs here and all 3 of them.
1. Drake MN2700 with external 1: 4 balun
2. Parallel circuit coupler similar to Johnson or Annecke
3. S-Match according to PA0FRI

The antenna I use here is a 2 × 20 m dipole symmetrically fed with a 240 ohm hose cable (antenna cable for UHF television antennas from the 1970s). The antenna hangs between 15 m and 9 m above the ground.

With all 3 antennas I reach 1: 1 SWR from 160-10 m ... the results of many tests are in the following order:

1. S-match on all belts number 1
2. Parallel circle according to Annecke number 2
3. Drake tuner number 3

You see your S-Match is the best ATU, plus the convenient operation. I built this one with a huge commercial 14 µH roll coil and a 20-200 pF transmitter coil. At the moment I still use 1 Amidon T200-2 as a toroidal core, but I will replace it with a 2 toroidal core solution with 2 x T200-2, because one core gets a bit warm at 1 kW output.

Now to my observations with the S-Match. At 160 m I only need about 7 µH for the roller coil and about 400 pF for the variable capacitor (I had actually expected that the roller coil would have too little inductance at 160 m). With the higher bands, the L is of course even smaller (at 10 m I still have 1 turn of the roller coil). Despite these very small values, the S-Match is still the best on all (!) SW bands.





First of all, I would like to thank you very much for your publication on the S-Match on the Internet. The S-Match coupler has turned out to be the most efficient handicraft project for me in a long time. Why? Now that the various couplers are stacked up with me, similar to yours, one should think it is finally enough. Nothing like that!



During a test the other day with Peter, DC0DX, on the 160m tape, I connected my S-Match test setup (fig »), which I was about to tear apart for another project, just for fun. Thunderstorm, the noise had dropped from almost S7 to around S2, but Peter's field strength remained around the S9. I could hardly believe my eyes. In addition, I received a clear increase in volume from Peter. Unfortunately, this could not have been specified in more detail ("That wouldn't have happened with glasses ...") So quickly put the antenna switch and then compared. No coupler brought such great values.

As much as I was happy about the success, the less I could make a verse about the extremely positive behavior. Unfortunately not until today. Now, however, I asked myself whether the S-Mach coupler only performed so well on my V-antenna (2 × 15 m, approx. 40 ° opening angle).Nowhere had I heard or read of such unbelievably great values. So off to Dieter's garden (DB3DK) and set up the new thick HP comparison coupler and the S-Match. 2 × 20 m antenna. What a surprise: both were absolutely the same in reception!




Apart from the effect on my antenna, the coupler is characterized by its absolute simplicity. The small coupler (fig ») that I have now built contains a toroidal core (439 pF) with a plate spacing of only 1.5 mm and a relatively small roller coil (28 µH) a la Annecke. When tuned, 100 W HF is absolutely no problem. The additional switchable capacitors are not needed on a 2 × 20 m long antenna, but at home they are. In the 80 m band I need 1100 pF, little L and on 160 m the rotary capacitor is enough next to 21 µH of the roll spool. Incidentally, all tapes from 160 to 10 m work perfectly.


It remains to be mentioned that sometimes the antenna has to be placed on the rotary coils and sometimes on the reel spool. Usually the output on the reel spool is to be used.

The tuning behavior is as sharp as a needle on the rotary head and as soft as butter on the reel spool. So I first move the turntable in the right direction and then dip the roller spool. Finding the first response is slightly more difficult than with the HP coupler, but it is still easy to master.




Your design is very interesting, because it is very easy to build (e.g. for HF beginners), saving expensive radio parts and it is small in dimensions (e.g. for portable ham activities). In my case I have used a 34 µH / 3 A coil, a 220 pF / 2 kV capacitor and a toroid T200-2, enough for more than 100 W power. I am using your S-March concept (fig ») now together with my 2 × 12.5m portable dipole and a 13 m feeder line in the field. It's coupling the antenna in good balance from 80 m to 10 m band in a perfect SWR. It fits perfect the little size of my TS 50.











Here is a picture of my SMatch a Super Sym.Koppler, works wonderfully on my dipole with chicken ladder. Compared to Annecke or MFJ 974HB easier to use RX / TX side no difference can be seen.







I have just provisionally loped your S-Match together on a board with old parts from the handicraft box and tested it. Construction time: Just half an hour (including winding the 2 transformers (T184: 2 × 8 turns 1.5 jumper wire for 3µH).

Conclusion: Everything problem-free !!! - I am completely thrilled and will get a few regular devices from it
to build.



DC4JG's S-Match built with a T225A-2 toroid and a variometer.





I read the description of the S-Match coupler on your website and immediately recreated the coupler. The rotary head is a butterfly rotary head 20 to 200 pf and the L is a spherical variometer with 8 to 45 µH. The transformer is a T225A-2. The S-Match works great. I tried many couplers see www.dc4jg.de but none was as unproblematic as the S-Match. With the little rotary cap and the spherical variometer I pass over all the bands 160 to 20 m, the rest has not been tried. But all in all, a good coupler. The coupler is also very well symmetrical on the feeder line. 3 to 4 OMs have already successfully built this coupler.






This would also be a variant that should not be ignored. Have been running this balanced coupler for more than a month. In the meantime I have installed the test setup properly in a plexiglass case. I added a SWR / Power measuring head in the input, as well as a measuring head on the output side to be able to measure the voltage at the transition to the H-conductor. It adjusts from 160-10 m. Since I wanted to know whether there is a big difference between a symmetrical coupler, the Palstar circuit, or according to PA0FRI, I built it with two Russian coils and a rotary coil, result; my other stations could not detect any difference. (Test: as always over several days, 160, 80, 15 and 10 m, 2 × 15.60 m (former G5RV) and 26 m H-ladder as I would have to come to the shack).

I can recommend the easy and inexpensive to manufacture symmetrical coupler according to PA0FRI, I could not measure any asymmetry or phase shift. (Measuring device, coupler, each different termination R's of the assumed impedance on the H-conductor and two identical probes on an HP Oszi) Both couplers behaved almost the same.


The advantage of the S-Match is the size and the low cost factor of the components, it can be set up in the smallest of spaces, a reel is not absolutely necessary, a coil with taps and a minimum of 12-stage switch is also sufficient.

As above a simple double rotary cap with 2 × 500 pF is sufficient. At least 100-330 pF should be switched on in parallel for 160 m. A structure with two T200 cores («fig) is also not necessary, I could not measure any difference in the variant with one core, again one component less.




I also built an S-Match with a split rotating coil 2 × 220 / 1.5 kV, a roll coil 15.5µH (1.5 silver wire), and 2 × T200-2 ring cores. The S-Match does its job excellently from 10 m to 160 m on a symmetrical dipole! But I will add a cross-pointer standing wave measuring device to the S-Mach in the housing. I am super excited and can only recommend it.




ZL2ML emailed the following:

Initially I had a LDG automatic tuner but at times it just did not match properly. I next build a balanced tuner with two roller inductors according to the information provided by AG6K in his publication 'A balanced antenna tuner'. This worked a lot better. But manually turning two mechanically coupled roller inductors is not very light and the design is only a L-C which means that any impedance's that fall between approx. 37.5 and 75 Ohm need large values ​​of either the capacitor or the inductor and hence normally do not get matched properly. (This is also the reason why many automatic ATU's like the LDG do not tune below a SWR of 1: 1.5)



It was a simple matter to test the S-match out (fig ») and then I proceeded to build it properly. Am using a Johnson 229-203 roller inductor 20 µH / 5 A, a Jackson Brother 2 × 300 pF / 1 kV variable capacitor and on 80 meters I add two 180 pF / 2 kV Vitramon capacitors in parallel to obtain a match.



The toroid is a T184-2 ("fig) and it was chosen because it needs less winding to obtain the 3 uH inductance than the other toroids. That means less parasitic capacitance. The windings are # 16 gauge Thermaleze (2 kV isolation) with Teflon tubing. Primary is 11 windings (exactly 3 µH) and the secondary is 2 × 5 windings (fits evenly between the primary windings: if the primary was 10 windings then 1 secondary winding would not be 'nested' between two windings of the primary). As 'base' I used 10 mm plexiglass. But do not use black Plexiglas (like I did): it shows every speck of dust in a photograph! The cover is made of solid walnut, which just rests on the plexiglass and is held in place by the groove that sits over the edge of the front panel. In this way it is easily removed for once a year cleaning and re-lubrication (very sparingly with the proper conductive lubricant) of the roller inductor wheel.


Many thanks to PAØFRI for coming up with this design and I have now disposed of all the other antenna tuners, this one is a keeper.



An OM asked me if he could use a balun without any core. During my early experiments, I did a brief test on this subject but I lost the notes in my rough-copy book. I still remember the two formers -1 "in diameter which were close wound with bifilar turns of well-insulated wire. Each coil was of 1.5 µH inductance. Than, to save space and avoid any external field's influence, I continued the experiments with toroid baluns.