In this section:
1) As much as we can, FIRST, we make any improvements, clean up any loss and detuning issues in and around the inverted L, to make it as efficient as we can.
When we're done doing all the improvements...
2) We use one of various methods to match the L's feedpoint to 50 Ω so it will look good on the shack SWR meter and support amps with limited matching range.
Since publishing the NCJ article in May 2012 we have engaged in correspondence, a hundred or so threads a year. This email or phone or in-person communication was related to the 5/16 Wave Single Wire Folded Counterpoise more commonly known as the FCP.
We quickly realized most FCP installers were using the inverted L for their 160m aerial wire. That was their choice.
Installers put up an L because they had seen them working before, and because in mechanical and practical terms an inverted L was doable on their property. Experience and struggle-produced inverted L understandings have taught us that an L over FCP can be tamed, "saddle-broken" into an excellent antenna. This still-evolving article lays out the details of the taming.
Most of our early correspondents were asking about getting their new aerial wire + FCP tuned up and looking good. "Looking good" meant a comforting and amplifier-compatible 1:1 50Ω SWR in the shack. In solving these issues, it became clear:
Most problems at Inverted-L-over-FCP sites were not the FCP. Issues were the L itself, its local environment, and a tuning device to match the L's suddenly efficient 20-35Ω feed to 50Ω coax.
Installers, in the beginning very much including K2AV and W0UCE, simply didn't know much about an inverted L. Though we would not want to insult anyone, the suspicion persists that largely hams still know little about an inverted L. Its mechanical simplicity is an effective disguise for its true complexity.
Working with these correspondents and their problems unearthed priceless information probably not discoverable any other way. emerged from that experience, and we hope to pass it all on to the reader without the bafflement, expense and frustrating poking in the dark to gain understanding starting from zero.
Completely cleaned up, a common quarter-wave-ish inverted L over FCP with an isolation transformer (L/IsoT/FCP) is a very strong performing antenna. It is probably the most effective 160 aerial wire possible at many locations, with simple support requirements.
Multibanding: 160+80 160+80+40
Many hams have room and support for good antennas on 80 and 40. Others are cramped for space, have interaction issues and other problems with their lower band antennas, or would have to take down an existing antenna to put up a 160 L/FCP. One solution is to add some complexity to the 160 L/FCP to achieve 160+80 or 160+80+40 in the same antenna space.
80 meters can be added to the L/FCP without changes to the aerial wire. The L is used on 80 as an end-fed half-wave, a best-performing single wire antenna for both local and DX, with advantages over dipoles, inverted V's and simple verticals. The pattern is fairly omnidirectional. This modification adds the complexity of matching and band switching at the feedpoint. See
To add 40 meters to a 160/80 L over FCP, a 35.5 foot wire is attached to the bend, running beneath and gradually dropping to 5.5 feet below the horizontal wire. The augmented inverted L is called a "Trident L" for the three wires, vertical, the 80/160m horizontal and the 40m "Trident".
The modification increases the complexity of the aerial wire itself, and of the band switching and impedance matching. But the reward for the complexity is an excellent 160/80/40 antenna that only occupies the single airspace and footprint of the 160m L/FCP. The FCP short, feedpoint to center of the third fold, already switching the FCP to 80m operation, also serves 40m.
The 40m-favored directions cover 270 contiguous degrees with the same weaker quadrant as 160m. For critical aerial wire details and how they generate the 40m pattern see
You can listen for K2AV on his relocated Inverted L, now a Trident L, which is a very medium size, medium height L. The L is up 66 feet (55 over FCP at 11). It's currently tuned with Method B using a "heavy" 4 inch toroid with 21:26 turns ratio, at some point to be replaced with 22:27.
The original K2AV L over FCP was a 3/8 wave L, up 90 out 105 foot L, out next to the service road. But the Town of Apex, North Carolina ran a 13 kV primary power line down the service road within 25 feet of the old L. For safety and RF noise reasons, that forced relocating the L to a pair of trees up near the house. You could find or erect the mechanical equal to the relocated K2AV L on many small properties with some 50-60-70 foot trees.
Please do not interpret the K2AV dimensions as this web site's "recommended" wire lengths and height for an inverted L over an FCP.†† Study the design principles in The K2AV dimensions were generated by applying the principles in Design an Inv L to K2AV's circumstances. If you use K2AV's dimensions you may be selling yourself short. **KNOW THE REASONS WHY** K2AV used those dimensions, and then design for YOUR situation.
†† The Trident L has a third wire, the "Trident" wire, that has a specified length and separation from the horizontal wire. The Trident wire can't be changed without undoing the 40 meter pattern. The 40 meter pattern does not depend on the vertical and horizontal wires of the 160 L.
Aside from using an FCP as the counterpoise, nearly all the issues and countermeasures listed in "Your Shrinking RF", also known as "The Loss List", were developed with correspondents and tabulated after the old K2AV L went up. Loss List items have all been addressed in the new K2AV L, mostly were never addressed in the old L. Some Loss List items would have been very difficult to remedy at the old site.
The new K2AV L has always clearly outperformed the old K2AV L on 160, most obviously to Europe and west coast USA. K2AV says it's nice not to mourn losing the old site to the power company. During the same twenty minute time span in summertime 03Z CWT contests, K2AV has had 160m EU RBN spots and 160m QSOs with California stations. That never happened on the old L.
While it's not possible to assign comparison dB benefits to each individual change between the two, being unable to rank individual changes does not diminish enjoying the new L's obvious total 160m improvement to Europe and the USA West Coast.
This section treats numerous issues that have flummoxed hams trying to use inverted L's. It includes taking a careful, detailed look at your planned or existing station looking for loss and other issues that have defeated other installers.
You may have read all you can take in today, but if you are going to use the ubiquitous inverted L, DO come back later rested, and take on this section and There is much to be gained.
Pay special attention to items throughout marked **Like This** which are murk and confusion factors from correspondence on inverted L over FCP. These warnings are repeated all around this web site to make sure they are seen, so many times do they recur as confusion issues in correspondence.
Two particular warning items stand out starkly from the rest as singular roots of misinformation, confusion and discouragement. These two warnings are involved in easily 2/3 of our correspondence. They are somehow readily forgotten by correspondents who can need multiple reminders over the course of their project. Initially some correspondents are not even convinced either of these warnings are true. We have no explanation why these particular two persist to afflict novice and old-timer alike to such degree. The hyper-emphasis below, indicating the literary equivalent of jumping up and down, screaming at the top of our voices while waving lit highway flares, is deliberate. For cause. No apologies for screaming over the internet will be forthcoming.
Not even close. Think 20Ω to 35Ω.
Varies with dimensions and environment.
♦♦ Lower SWR does NOT
predict improved performance ♦♦
A dummy load has a broad, perfect SWR.
And at 100% loss, is a worse antenna than a light bulb.
**Again, loss should be remedied before tuning work because successfully completing a number of loss countermeasures almost always changes feedpoint resistance and reactance, sometimes very large changes.**
One common example occurs doing a "160m de-resonance" loss mitigation on nearby 80/40 dipoles, vees, OCF doublets, etc. Excited by a nearby 160m antenna, these 80/40 antennas' radiating wires and their feedlines can behave as a T or L "weed" parasitic element with a lot of its "vertical wire" (shield of feedline to dipole/vee/doublet) laying on the ground. This indirectly hard-couples the 160 antenna to lossy ground. These weed parasitic elements can also "pull" the feed impedance/resonance of the 160 antenna. Such unintended couplings often make the 160 antenna poorly responsive to 160 antenna adjustments that are otherwise best practice and work well.
1) The inverted L unintentially couples nearby secondary conductors, "pulling" the impedance. The degree and specifics of this unintended pull vary unpredictably with the secondary conductors involved. This loss factor often harms RX by coupling noise otherwise not picked up by the intended antenna conductors.
2) The inverted L has a lot of dielectric and/or resistive loss somewhere in the near field, possibly 15Ω to 30Ω effective series loss in that 50Ω, producing 1.5 to 4 dB potential recoverable loss.
160m R, when X=0 at K2AV's current L/FCP feedpoint, varies from a wintertime no leaves, no sap 28Ω, to a summertime full leaf and juicy sap 31Ω, and to one 38Ω in hurricane standing water. It is not a 50 ohm antenna. An adjustable isolation transformer, 21:26 turns ratio, matches the 28-31Ω to 50 ohm coax.
RigExpert, various models, portable, R,X and other graphical presentations, USB to PC for larger, detailed screen, saving and processing data for in-shack view of feedpoint state. At this writing, it appears to be the least expensive self-contained with all required features. After setting center frequency and range, remaining work can be done with "Scan R,X" and "Display All" menu items. Continuously variable Zref can only be set in software display. The 30 model is the cheapest but cannot store readings unless hooked to a PC or tablet while making measurements. The 54 model adds 6 meters and also has 100 memory slots to retain reading data for later transfer to PC. Other models are available with additional features and cost.
AIM family 4170, 4300, etc, previously from W5BIG.com and dealers. Presentation from Windows PC. Portable only with windows tablet or small laptop running software. These have been around for a long while, rock solid to those who are used to and equipped for them. No longer manufactured or sold retail.
Mini Radio Solutions "miniVNA Pro2", two port VNA, requires Windows PC or Android smart phone for presentation. Contains Lithium Ion battery and Bluetooth. Charging via USB. Communication to Windows or Android via USB or Bluetooth. Allows placing VNA at height without connecting wires to interfere with normal operation of VNA or "pulling" the antenna off its true parameters. Supplied PC programming is Java-script. App supplied for Android. Apparently no longer available retail and no way to service since the pandemic. RIP.
DG8SAQ VNA from SDR-Kits. This is the 2 port VNA per se only. Uses supplied software runing on Windows or Linux computer for presentation. Connection and power via USB cable. To be portable requires a Windows or Linux tablet or laptop. But quite inexpensive if you have the computer.
Recent correspondence has identified a sneaky tuning issue, in situations where the shield of the antenna's coax to the shack is heavily induced by the antenna's aerial wires. This is worse on 160 meters because common mode blocking performance of typical devices and feed system configurations decreases as frequency decreases.
The actual Z in the circuit at the feedpoint while the IsoT is connected directly to the shack coax, ready to transmit, is one value. The feedpoint Z when the IsoT is only connected to an RF analyzer is a different value, occasionally wildly different. The installer's adjustment efforts are matching to the wrong R, X and F0, all while following instructions and measuring at the correct places.
It can occur particularly if the coax shield has an electrical quarter-wave-ish length between a shield grounding point and a high value common mode current block at the antenna end, as with an isolation transformer (IsoT). This can put a rather high common mode voltage at the IsoT. Although there is no danger of a breakdown using web site specs, it does feed back into the system to some degree and can shift the tuning measures/settings for presenting 50+j0 to the coax.
This feedback will be present when the system is in normal operating mode. It will have some effect on X, R and F0 as a normal part of the antenna behavior. But if the coax to the shack is simply disconnected and laid aside while measuring/tuning with the RF analyzer, the feedback from the coax shield is removed and the system is tuned up without it.
When the analyzer is removed and the shack coax is reconnected, the feedback is reintroduced. Then K2AV can get a complaint that the tune-up procedure is not working.
However, if the coax shield, only, retains its connection to the antenna system with a simple jumper, then that influence, however large or small, is present and is taken into account along with all else as the tuning is done.
In the image to the left, K2AV has the jumper wired directly to the analyzer's coax jumper shield, reportedly so he won't forget to make the connection. A simple jumper with a large clip on both ends, or just two large clips soldered together, can connect the two PL259 barrels involved, serving the same purpose.
K2AV made his jumper out of RG400 with a pair of UG175U adapters to fit the RG400 to the PL259's. The RG400 has a PTFE jacket and dielectric, which will not melt while soldering. The RG400 conductors are silvered copper strands which will not break with bending or develop oxides that do not conduct. RG400 is designed for use in aircraft cable runs and bends easily without breakage. The wire to the clip is soldered direct to the shield of the RG400. The RG400 is rated to 7 kW at 50 ohms. In his new (2018) choke cookbook, K9YC uses RG400 as the winding conductor.
At this writing the clips can be found on Amazon.com with a search on "B007ZZFUDA" without the quotes. The clips are also found with stretchy PVC insulation, which can be removed to allow opening jaws wide enough to go around a PL259 barrel. Short lengths of RG400 are inexpensive on eBay. At this writing six inch RG400 with a PL259 on both ends are available and inexpensive.
This text does not address the question as to whether an installation's induced common mode effect is excessive, or causes noise or other problems, or that likely antennas other than L/FCP can be so affected. But by attaching the jumper as illustrated above, noting the analyzer reading, and then removing the clip from the barrel, the change in R and X will indicate the presence and degree of induction. The installer can take appropriate action, or none as seems fit. How to gauge and remedy the effect must be left to a later article added to the Loss List. A link will be added here when such text is added to the site.
In "An Unhappy Example" above, a graphing RF analyzer would simply show X crossing zero at 1.837, where R=35Ω. For purposes of illustration let's assume that our center frequency of choice below is always 1.825 MHz. Using the graphing analyzer you then have five basic strategies with possible combinations for matching the antenna feed to 50Ω coax.
Method A: Prune/extend horizontal for R=50Ω then tune out X.
Method B: Prune/extend horizontal for X = 0 then adjust isolation transformer turns ratio to approach R=50Ω.
Method C: Non-resonant dimensions for best RF then matching network.
Method D: Non-resonant dimensions for best RF then use coaxial series matching transformer as the coax to shack.
Method T: Non-resonant dimensions for best RF, possible simple fixed components to moderate the Z from isolation transformer, larger feedline to shack, tuner in the shack.
Note: There may be significant work involved, some methods more work than others. In text below, 1.825 is used as a center point. Substitute your own desired center point.
Note: Do not change FCP dimensions to tune antenna. Clearly reduces performance. Already tried in 2010 with significant RF loss by K2AV and W0UCE, who still had a lot to learn back then.
Important: Using ANY of the methods below, especially if your isolation transformer shack side winding connects immediately to coax to the shack, DO use the between the RF analyst, the IsoT, and the coax shield to the shack. DO have the feedline from the IsoT to amplifier (or transceiver if no amp) in its regular operating configuration, as if you were going to screw the coax onto the IsoT and transmit right now.
**Especially on 160m, for tuning, have all your station's coax shields in their final configuration (length, layout, routing, connections), except the L/FCP feed coax center conductor not connected at the IsoT.**
**Not having final, ready-to-transmit configuration with the special jumper has caused large, wild variations in resonance and apparent feed impedance. This can easily cause you to spend time tuning, even removing turns from an adjustable IsoT, to match to a specific impedance that no longer exists when you hook everything up to operate.**
Method A: Prune/extend horizontal for R=50Ω then tune out X. Note: Try to limit prune/extend of the horizontal (horizontal + drooper) to ±3 ft (±1m) to remain in the performance sweet range. Larger adjustments can be made but gradually start to reduce performance.
Method A1 Matching is done on the antenna side of the isolation transformer, but measured on the coax side of the isolation transformer to absorb non 1:1 aspects of the isolation transformer.
Find the point on the graph where R (not Z) = 50Ω. If that point is not 1.825, then prune/lengthen the L horizontal wire to move the R (not Z) = 50Ω point to 1.825. Then depending on the sign of the X at 1.825, in series use a transmit power level variable capacitor or (rarely) a tapped or variable inductor at the base of the vertical wire to adjust the X to zero at 1.825.
Wire length and capacitor adjustments may interact a bit, repeat the process if needed. A vacuum variable capacitor is commonly used for QRO. Air gap variable capacitors are sometimes used for lower power levels, but must be protected from dampness.
If the uncorrected inductive X value at R=50Ω is too low, method A1 has a blind spot where the required value of a single cap is too large to be practical. Add a series coil to increase the inductive reactance and bring the X correction into range of the cap. Or better, implement A2 and use the cap elsewhere. Either narrows the SWR bandwidth a little.
Method A2 Same as A1, but does not require a vacuum variable capacitor. Place in series with the antenna a fixed high power, high current transmitting cap in series with a coil and move a tap on the coil to zero out X at the 50Ω point. This narrows the SWR bandwidth a little. When using the RF analyzer, use receiving capacitors to determine the value just small enough to go with a tap on a 3 uH coil, this will allow you to determine the appropriate pF value of the capacitor.
To obtain sufficient current rating in fixed transmitting capacitance, multiple identical caps can be used in parallel. Better to have the combined current rating well in excess of calculated max than discovering insufficient rating in the middle of a contest in a blizzard or rainstorm.
Method A advantages: The Z = 50+j0 point can be set precisely at the transformer output, and then the series adjustment at the antenna can fine tune the 1.5:1 or 2:1 SWR range points in the shack to best advantage. Method A is compatible with the 160/80 dual banding extension. Method A2 can switch between different taps on the coil to cover multiple 1.5:1 SWR ranges on 160.
Method A disadvantages: A project box is needed at the antenna to house a vacuum variable or fixed cap, or coil as specified above. This makes a DIY isolation transformer, or ordering commercial transformer without the enclosure a good idea, to wire it into the box housing the cap.
If You're Lucky: If a high power, high current 2000, 2200, or 3300 pf transmitting cap, together with pruning the horizontal/drooper produces results close enough for you, it all can be put in a plastic 6 x 6 enclosure, including transplanted hardware and guts of commercial transformer.
Method B: Prune for X = 0 then adjust isolation transformer turns ratio to approach R = 50Ω.
Note: Limit the horizontal (horizontal + drooper) prune/extend to ±3 ft (±1m) to remain in the performance sweet range. Larger adjustments gradually start to reduce performance.
Prune or lengthen the L horizontal wire to move the X=0 point to 1.825. You likely will have a somewhat different R value, say R = 38Ω, after you do this. 35Ω-40Ω is a fairly common result. Then for R in the high thirties Ω, you can remove the two end turns from the antenna/FCP winding of the standard 3 inch core isolation transformer. This R may be close enough to 50 for you. At this point you can prune/lengthen the horizontal wire to optimally place your 1.5:1 or 2:1 SWR range points in the shack.
This method works because a turns ratio of 18:20 gives 18/20 = 0.9. 0.9 squared = 0.81 impedance ratio. Theoretically 0.81 times 50Ω = 40.5Ω on antenna to produce 50Ω at coax connection to the shack. Given a normal degree of non-1:1 transformation for a core/winding/application of this type, this is not an exact value.
However, given the ease of pulling out one turn through the core at either end of the antenna/FCP winding and reconnecting, it's possible this easy reduction of two turns may make the SWR curve "good enough" on a thirty something ohm L/FCP. Restoring the two turns to original state is easy if the wire has not been yet trimmed.
To wind an IsoT specifically engineered for Method B, with an input range of 23Ω to 46Ω and matching to within 1 or 2 ohms, see
Advantages: Method B is dirt simple out at the antenna feedpoint after everything is done and closed up. Lacks the series capacitor blind spot in method A. Just another antenna with a 4" x 4" or 6" x 6" grey box between the antenna and counterpoise wires, with coax running off it to the shack. With one of the Advanced X:Y turns ratio transformers, one can transform 23 to 40 Ω to 50 in roughly 2.5Ω steps by removing end turns per the adjustment procedure.
Disadvantages: Working with turns ratios can be tedious. Changes in the aerial and FCP may require redoing the winding.
Method C: Non-resonant dimensions for best RF then matching network.
Build an adjustable L matching network between the isolation transformer and coax to shack. Quite a few hams are capable of designing these, especially now with the collection of excellent network analysis and design programs available. This method may complicate dual-banding by requiring additional switched circuitry. This will have to be your own design project.
Advantages: Assuming taps on coils or adjustable capacitors or both, easier to adjust after changes to antenna, as in unavoidable gradual implementation of loss list fixes.
Disadvantages: 160/80 dual banding may require switching the network in and out, designing modifications for the 160/80 dual banding circuit to operate correctly.
Method D: Non-resonant dimensions for best RF then coax series matching transformer as coax to shack.
Build a coaxial series matching transformer, same strategy as Method C, but accomplishes the same in the coax running to the shack.
A nice little free calculator for this purpose may be found here, courtesy Greg, W8WWV. It's based on the formula for the method found in the ARRL Antenna Handbook, 18th Ed.
Advantages: Useable with a stock DIY or commercial isolation transformer, giving the antenna the external simplicity of Method B without having to mess with turns ratio. For the dual band scheme, a simple one range on 160, one range on 80 version can be done easily since the dual banding scheme has an 80-meters-is-active voltage available, both in the station and out at the antenna. This is the voltage that energizes the FCP shorting relay on 80 meters.
Disadvantages: Without some changes, cannot be used with 160/80 dual banding or 160 range-switching because the coax series matching transformer is a single frequency device. The coax lengths are specific to your antenna on a single 160 frequency range only. Switching for more than one range on a band would require switching separate coax runs for frequency switching.
Method T: Non-resonant dimensions for best RF, simple components to moderate the Z at transformer output, larger coax, tuner in shack.
Use any of various methods to convert antenna Z to the rough vicinity of 50 + j0. Use a larger feedline, RG213 minimum or small hardline. Use a stand-alone or built-in tuner in the shack.
Larger coax, better yet small hardline, will have less resistance in the center conductor or shield where current maximums occur, reducing the dB loss at any power level.
Advantages: For those always operating transceivers barefoot, a built-in auto-tuner may have excellent range. Adjust the minimal matching at the antenna to place and maximize the auto-tuner range. At K2AV the K3's built-in auto-tuner matches 1.8 to 1.9 at 1:1 and to 1.920 at 1.5:1.
With a no-tuner outcome near 50+j0 at 1.825, some stand-alone tuners in the shack can match all the way to 2 MHz. At K2AV, at 1.975 MHZ, the ATR-30 sees Z = 2-j25 (SWR=30:1) at the shack end of 82 feet of LDF4-50A half inch hardline from the L/FCP/IsoT, and still can produce 50+j0 (1:1 SWR). K2AV limits himself to 1000 watts up there. For those who operate CW and have regular low power up-band SSB contacts, impedance matching by the numbers in the shack may be all that is needed.
Disadvantages:If an amplifier is used, a tuner in the shack may not stand up to voltages at the extreme frequencies. At K2AV this method did not stand up to more than 300 or 400 watts above 1.95 MHz due to voltages present, usually arcing across SO239 chassis connectors in the very solidly built ATR-30 tuner.
This disadvantage can be overcome: The offending SO239 and PL259 terminated RG213 at the ATR-30 was replaced with 7/16 DIN connections and LDF4-50A 1/2 inch hardline on one antenna connection. Now the ATR-30 tuner, connectors and hardline can handle an efficient 8:1 SWR at 1.5 kW at 1.999 MHz. The arrangement is stable and a table of tuner settings vs. frequency allows reliable tune by the numbers.
This works well for run frequencies above 1.9 in 160 SSB contests, or regular QSO's up in the high end. With an antenna feed point tuned to the low end of 160, random search and pounce below 1.9 can use the auto-tuner of the KPA1500, but S&P above 1.9 needs to reduce power out. For 1500 watts up to 2.0 MHz, using a switched in 1.850 Mhz tuning at the IsoT, above 1.9 MHz can be tuned by the KPA1500 with the ATR-30 in-line tuned to 1.95.