My Homebrew Antenna Coupler

Discussion in 'Amateur Radio' started by W5HRO, May 10, 2015.

  1. W5HRO

    W5HRO Administrator

    I pulled my unfinished HB antenna coupler out of the garage and am getting ready to finish it. It's been covered up out there for the past few years. It's in a cabinet that matches my HB transmitter and it will sit on top of it.


    I finally found another Johnson KW matchbox for the extra parts and knobs.

    Then here is the circuit drawing...
  2. W5HRO

    W5HRO Administrator

    After thinking about it I may just modify everything per the below. Since I now have another big matching KW matchbox parallel tuning cap I may connect the current one in there to the input link winding then add 1 or 2 doorknobs on a switch to work all the way down to 160-meters. Then I can use the other one in parallel with the output winding. I also have two other 300pF to 400pF (I think) HV tuning caps I can throw in series on the output as well. Those are much smaller in size and they came out of an old Nye Viking or Ameritron T-network tuner. I don't remember which one. I also have the knobs from a Johnson 275 matchbox for those so everything would match.

  3. W5HRO

    W5HRO Administrator

    I did some more work on the coupler the night before last and I think the drawing below is almost there. It's really close now.

    One thing I discovered is the coil out of the KW matchbox is too large in size to fit between the spacing of the two Vcaps. I never planned to use the coil from one anyway so I think I will just use the B&W stock or wind my own with the link winding and make the whole thing enough uH's to cover 160-meters with the big Vcap. The Johnson coil is too small for 160 and the link winding gap to the secondary winding is a bit too wide anyway. What I can do though is use the two 5 position continuous shorting switches for SW2 in the drawing below. I can then maybe add one extra switch for 160-meters when on 80-meters or just make that a manual switch inside of the cabinet somehow. Just lifting the lid on the top to do that would be easy.

    The two big Johnson split 220pF Vcaps are working great too. It's nice having two of those now because one also works perfect on the input link winding side. Those caps are C1 in the KW matchbox (152-517-3) and are 110pF per section. What I also discovered is having the other caps in series with the output is a plus. The bigger ones I had are like 240pF which is too big so I used a couple more I found which are only 100pF each. With those set halfway to around 50pF it actually helps balance my antenna just by tweaking each one a little. It might be nice to add a couple of small lamps on the front panel to monitor the balance. I would just need to find the right size lamps that wouldn't mess anything up or maybe switch them in and out of the circuit.

    I'll post the final circuit and pics of the inside and outside of the coupler when everything is finished.

    I've shorted the 110pF stator sections of each Vcap together to make them 220pF total when in series, but I'm not grounding the metal frame (rotor) of C2. It is floating and mounted on ceramic insulators. The Vcaps are actually sort of reverse connected in my coupler. In the actual KW matchbox the frame (rotor) is grounded and that's part of the problem. You really cant fully remove the common mode current on the output side in the KW or 275 Johnson matchboxes. Removing the center strap from C2 to ground wont do it because there's no way to remove it from C1. Removing the center ground strap from C2 in the Johnson's will only make things worse as others in the past have already noticed. The only thing you could do with the Johnson matchboxes is install ceramic standoffs under C1 and float it like I am doing, but you would have to drill a hole in the front to raise the shaft and the knob up higher. The knob's skirt would probably cover up the old hole, but it might look a little funny with the left knob up higher than the one on the right.

    Anyway, it may be a little confusing in my drawing above, but when SW1 is in the 40-80M position the total "C" is 220pF max and not 55pF. There was really no other way to draw it to make it still look like a split-stator cap other than the way I drew it. I've updated the drawing the best I could.

  4. W5HRO

    W5HRO Administrator

    Tonight I came home and stripped down the coupler and disconnected all of the components.

    I think what I might do is just use one series cap for the 50-ohm output via the switch below.



    The above switch will switch this series cap below into the circuit when using a coax fed antenna. The switch will just short across the cap when not in use.


    I did find the piece of B&W stock below out in the garage and it's exactly 70uH, but I may run down to the hardware store tomorrow and just get some heavy bare copper wire or tubing and wind one with a bit wider spacing, but the one below may work fine.


    Anyway, below is the inside of the stripped down coupler with all of the wires/straps removed etc. I went ahead and remounted the Johnson input winding cap with metal standoffs. I will just keep the insulated ceramic standoffs on the output link side and just have the parallel cap for OWL. I can always switch in the one series cap if an imbalance occurs. I don't think I have room for two series caps.

  5. W5HRO

    W5HRO Administrator

    I'm thinking about changing my design to just a T-network or similar network even though it’s a high pass configuration. The new little MFJ mobile tuner I bought works so well and matches my coax fed 75-meter dipole on all of the HF ham bands except on 160-meters due to the short length that I think it will work better if I do. I doubt I'll will even mess around with OWL again, at least not unless I move out in the country somewhere someday and have land to where I can easily use it.

    Below is the MFJ schematic and I’m wondering why they used a balun with a few turns at the end of the coil to chassis ground. The selector switch shorts across that to chassis ground at each position. I’m wondering if maybe they did that to pad things in case someone switches the switch while transmitting or if in some way it helps prevent common mode current from getting in there and looping around somehow. It might be a good idea to add it. Does anyone have a clue as to why it’s there though? One thing I noticed is the ground symbol on the switch is different. I wonder if maybe inside of the tuner that means it’s just shorting the windings together and it’s not really chassis ground, but it sure looks like a ground symbol to me.I guess the balun could be adding "L" to the coil though if its not.

  6. WD5JKO

    WD5JKO Member


    I'm not too sure about the grounding symbols, but my guess is that coil at the bottom gets switched in for 160m coverage. If so, for the other bands that coil would be bypassed.

  7. W5HRO

    W5HRO Administrator

    That would make sense too if it were not for the way the schematic is drawn. It may be just a mistake with the drawing or it's just not very well explained. It shows it switched to "A" which would be the maximum inductance. It shows it going to ground at that point which would be shorted across the balun. The main coil does look awful small though to cover 160-meters, but who knows.

    Anyway, here is the P.O.S. below inside. I found the pic on the net. The P.O.S. works great though. Without taking the cover off of mine and tracing position "A" down I won’t know for sure.

  8. W5HRO

    W5HRO Administrator

    Ok I removed the 2 screws and slid the top cover off and the inductor switch common is tied to chassis ground. After looking at how the coil is wired to the balun to ground the balun is in fact shorted across to chassis ground when the switch is in position "A" so the balun is NOT being used to add more inductance for 160-meters. It's strange, but that’s how it’s all connected so the schematic is in fact correct except for one little detail. The main coil is cut close at the bottom and the balun is inserted in series with it, but there is only like one complete turn, maybe a half a turn of the main coil to chassis ground, but none of the "A" through "L" taps on the switch are connected below the balun. The lowest switch tap is still position "A" which is above the balun.

    The only thing I can think of is that its either to pad things if someone rotates the switch while transmitting or it is somehow absorbing common mode current on the chassis or for looping. My best guess is they added it for common mode, but I'm still trying to visualize how it works if so.
  9. W6MQI

    W6MQI Member

    Not quite sure what OWL is I thought this was a type of matching system used on yagi's? T's are great tuners also had great luck with balanced tuners using two rotary inductors gang together with a timing belt. One tuner I always wanted to build was a link coupled never got around to this maybe someday from what I've read they're very low loss. Your MFJ tuner pictured above what's the stated impedance range? Have you measured internal losses with the MFJ yet? One other thing I've never been a big fan of toriods in tuners guess there fine at low power levels below 100w.

    Guess you answered my question sort of I would have thought for sure that toriod was being used for 160?
  10. W5HRO

    W5HRO Administrator

    Open Wire Line
    It's basically connected like the below. Doesn't make a whole lot of sense does it? Unless they are trying to force a small common mode current loop and keep it circulating thru the balun to ground around and around, but would it even do that?

  11. W5HRO

    W5HRO Administrator

    I did some reading and it looks like if you use smaller "L" with larger "C" in a T-network tuner the efficiency can be as high as 95% even with a 20:1 antenna SWR.

    I do in fact now think I was right and MFJ added that choke for a reason and it probably helps improve the "Q" of the inductor given the fact the main coil looks a bit smaller than what you would think for 160-meter operation. I think they may have added the choke as a workaround and it improves performance on all bands even though the switch shorts across it to ground. I think that's how it actually works and it prevents bad things from happening because they used less coil and more "C". The choke does not actually add tuning "L" to the main coil though.

    Anyway, I think something like a 16uH to 20uH roller inductor with its wiper disconnected from its end and tied down to ground instead with the choke’s ground just like they have done in the MFJ circuit and then with two large 500pF tuning caps would probably be ideal. Anyone know if that 16uH roller inductor MFJ sells can have it's wiper disconnected or isolated from it's one end, or know of another high power roller someone else makes where it can be done?
  12. W5HRO

    W5HRO Administrator

    Here is my updated redesign plan. If I use a smaller inductor and bigger 500pF caps the efficiency can be as high as 95% with a 20:1 SWR antenna load. The bigger inductor with 200pF caps will still work, but with slightly less efficiency. I already have two big 200pF caps.

    T-Network_Tuner.png .
  13. W5HRO

    W5HRO Administrator

    I found my old pic back from I first started going through the stuff I dragged home from Tulsa in 2009. I still have these big caps out in the garage. They are HV 500pF Chinese surplus caps. The only problem is they are so big that they will not fit in my HB tuner cabinet and the shafts are a large metric size. I had the end of the shaft milled down to 1/4" on the same 350pF cap version I used in the HB RF deck, but that was a long time ago when I knew someone nearby who had a machine shop.

  14. KD6VXI

    KD6VXI Member

    I think that toroid inductor on the cold end of the coil is there only to increase LMin of the main inductor.

    I don't believe you'll find the mix gives you much choking, just additional L.

  15. W5HRO

    W5HRO Administrator

    Actually the toroid choke in the MFJ tuner is just under 100uH, but the whole thing gets shorted across via the switch so it's not part of the tuning network. It adds some L at the very end of the coil to help improve the overall Q of the main coil and absorbs some of the common mode current that will start looping (circulating) through it. My plan is to use a very big toroid or 2 or 3 put together to handle more power and current and have like around 200uH there.

    If you remove that toroid in the MFJ tuner it will still tune all of the bands fine, but when you remove your hand from the knobs or case the settings will move. The meters will rise up and down as you ground and unground the tuner with you hand. That choke does settle it down. If you use bigger tuning caps say like 500pF and less coil it will be even worse so it really needs that choke.

    I think because the MFJ is a small mobile tuner that it was designed to work in all environments like when it's outdoors with a poor or no ground, but I think it's a good idea for any T-network tuner design.
  16. David

    David Member

    Hi Forum!

    Hm...seems to me HRO's design (top) is a link coupled balanced tuner. No baluns. Floating air inductor true balanced tuner. So...I am wondering whether anyone has ever built a unit whose floating secondary consists of a roller in series with maybe 10 turns of the BW HRO shows, EACH SIDE of the link (which is typically mabye 4 turns at the center of the BW stock). IOW: NO TAPS ...just connect the ladder feeders to each roller and crank for SWR. I need something that will allow for dynamic ON AIR tuning...not this stop TX, set the tap, and do TX and chk SWR on and on. Yes...once the taps are found they can be recorded for THAT TX. Change the rig and possibly the taps need to change. Doesn't matter: I am seeking the DYNAMIC TUNER idea. NO BALUNS!

    WA6PBJ (Peanut Butter and Jelly)
  17. W5HRO

    W5HRO Administrator

    You would still need to have the main fixed secondary coil, but you could possibly use small roller inductors, one at each end to give it a little more fine tune, but you still need the split capacitor and probably a couple of the fixed taps to cover all of the HF bands.

    I've changed mine to an improved T-network design using a coax feeder again because the overall performance and efficiency you will get on all of the bands is much better. Using open-wire line is overrated despite what you hear. It’s pretty hard to have everything balanced on all of the bands all of the time using the same length of wire for the antenna, and the feed-line will be prone to radiating on transmit and picking up noise on receive when the balance isn't dead on. The trick is to use bigger 500pF capacitors with vernier reduction drives and a smaller coil with a toroid at the end to ground using the T-network design instead. You want a "Q" factor of only around 4 to 5 down low with a coil inductance of around 10uH to 20uH max for 160-meters. Then inductance will then be very tiny up on 10-meters but you will tune for a bit higher "Q" factor up there.
  18. W5HRO

    W5HRO Administrator

    By the way if anyone is interested I went ahead and updated the drawing to what I'm currently using. I have a fixed coil with an alligator clip tap setup so I still need to find the correct type of roller inductor if possible.

    One of the key things I also discovered is the RF ammeter needs to be in series with the coil to be able to use it to properly tune and protect the meter and it works perfectly. You don't even need an SWR meter connected. If you adjust everything for a max reading on the meter the SWR is dead smack 1:1 after. The other thing too is it gave me the ability to use the big 0A to 10A Westinghouse meter I found still new a few years ago because the current is double through the coil. A 500W AM carrier coming from my HB rig into 50-ohms is 3.16A, but in series with the coil it's 6.32A so the bigger meter works. I just divide the reading by 2 to get the correct power level into a 50-ohm dummy-load. It would actually be best to also have a lower current meter in series with the cap on the output side to the antenna, but to protect the meter if you were to transmit with the antenna disconnected by mistake it's best to place it in series with the coil instead.

    I can guarantee you this tuner or matchbox will out preform any tuner for made for open-wire line and it uses coax instead. It will tune a single 160-meter halfwave dipole from 160-meters up to 10-meters or a single 80-meter dipole from 80-meters up to 10-meters and you don't need traps or anything and you will get over 90% efficiency on all of the bands.

  19. W5HRO

    W5HRO Administrator

    One additional last thing I was thinking of adding to the matchbox is the two smaller 0A to 5A meters I have. It might be nice to be able to see the input power vs output power during tuning. They would never be the same though because if its 50-ohms on the input and say like 200-ohms on the output the readings will be different. Adding the meters though would probably aid in tuning so I could peak each capacitor quicker. The current through the coil is actually the input current plus the output current combined and is why having the meter there works so well. So basically, the meter reading on the right when added to the one on the left should equal the reading on the one in the middle in series with the coil. I just need to make sure I use a good choke on the output to help protect the meter if I do it.

    I have not pulled the white plastic off the front of the panel yet so I could still drill the two holes for the other meters.


    I have two of these unused meters below.

    Then here's the old pic of the big one in the middle.