When Boat HF RFI Refuses to Behave: Jan’s Long Road to a Real Fix

Marine HF troubleshooting is often presented as if one clever tweak solves everything. This case did the exact opposite. What started as a straightforward question about RF ground, counterpoise strategy, and tuner placement turned into a long real-world process involving repeated on-board tests, different harbor and open-water conditions, changing return paths, repositioning the tuner, adding multiple ferrite chokes, installing Schaffner filters, and eventually questioning whether the solar controller family itself was part of the problem.

Jan, ON2ZTT, worked through this on board SY Goudurix with unusual persistence. That is precisely why this case is worth documenting. It was not a one-evening “I added a strap and the problem was gone” story. It was a months-long elimination process where several ideas looked promising, some fixes only worked partly, and the usual outside advice did not produce a stable answer quickly.

This was a long-duration field case, not a quick fix. That distinction matters, because many marine HF problems only become understandable once every variable is tested separately and over time.

The original problem

The station on board was suffering from transmit-related RFI in the solar control chain. That sounds simple until you realize how many possible paths exist on a sailboat: the antenna wire itself, the tuner return, the copper strap, the coax shield, common-mode current on control and DC wiring, coupling into nearby equipment, and the controller hardware’s own tolerance to strong RF fields.

Jan was not working with a theoretical bench setup. He was dealing with an actual /MM installation with a real transmatch, a real solar system, real deck routing, real marine hardware, and real compromises in where conductors and equipment could physically go. That is what makes marine HF hard: the system is never just “antenna plus tuner.” It is the boat.

This was not a weekend fix

The troubleshooting stretched across multiple phases:

• July 2025: the discussion started around the classic boat question: what should actually serve as the return path, and what is the real meaning of “RF ground” on a fiberglass sailboat?
• September 2025: Schaffner filters were installed and multiple return-path and counterpoise combinations were tested, with some encouraging but not yet fully stable results.
• October 2025: further photo-documented tests compared the SGC-230 under deck versus above deck, with different counterpoise combinations and choke positions. Some layouts still produced RFI on 30 m, and in some cases 20 m and 17 m as well.
• November to December 2025: additional real-world boat tests showed that a configuration could behave well in one condition and then misbehave again later. That alone is enough reason to distrust quick conclusions.
• Early 2026: after controller replacement and keeping the Schaffners in place, Jan reported that all solar-related RFI was gone. At that point the remaining question was no longer controller RFI, but tuning and matching behavior on 15 m and 17 m with a shorter vertical on a carbon pole.

What was tested along the way

This case became useful because Jan did not stop after one or two trials. The installation evolved through repeated, isolated tests:

• ATU mounted above deck versus under deck
• Short and long counterpoise wires, including roughly 2.5 m, 3.5 m, 3.8 m, 7.5 m, and 9 m variants
• Counterpoises present, shortened, removed, or even rolled up to make them electrically insignificant
• A short copper route from the tuner area to the existing underwater bronze plate
• FT240-31 and FT240-43 ferrite chokes in more than one location on the feedline and cable bundle
• Schaffner line filters installed in the solar path
• A 9.2 m sloper versus a shorter 7.2 m vertical / about 7.68 m effective radiator
• Tests in the marina, on land during maintenance, and at sea / at anchor
• Finally, a replacement test using a Victron MPPT in place of the problematic controller family

What first looked promising

Schaffner filters clearly helped. Early testing showed that once the filters were installed, some bands improved immediately. At one point the installation worked well on every band except 30 m. Later, with the tuner back under deck, multiple ferrite chokes in place, and both experimental counterpoises effectively removed from the picture, Jan reported no RFI across all bands, with successful 17 m operation and decent on-air reports.

That was a real improvement, but it still was not the end of the story. The key lesson is that an apparent fix is not the same thing as a repeatable fix. Later testing with similar ideas still produced RFI in some conditions, especially on 20 m, 17 m, and 30 m. That is exactly why long troubleshooting trails matter: they expose whether a result is robust or merely situational.

The part many people miss: random return conductors can become part of the radiator

One of the clearest patterns in the case was that “adding more metal” did not automatically make the system better. In several tests, extra counterpoise wires and ad-hoc return conductors did not calm the system down. In fact, they often made it less predictable.

That fits a pattern seen repeatedly in asymmetrical HF systems on boats: once those extra conductors begin participating as uncontrolled radiators, they stop being a clean engineering solution and become part of the problem. The fact that some configurations improved when those experimental counterpoise wires were removed should be read very carefully. On a boat, “more conductors” is not the same thing as “better return.”

What finally changed the outcome

The most important turning point came when Jan stopped assuming the antenna side had to carry all the blame and tested a different solar controller brand. He temporarily replaced one Mastervolt unit with a Victron MPPT on the side of the boat closest to the antenna and controller location. The result was immediate and striking: no RFI on any HF band.

That mattered because it changed the diagnosis. Up to that point, the investigation naturally focused on tuner position, copper strap length, “RF ground,” counterpoises, and chokes. Those things still mattered. But the controller swap showed that the controller family itself was not a passive bystander. After Jan later replaced the controllers and kept the Schaffners installed, he reported that all RFI in the solar system was gone.

So was the underwater bronze plate the “magic RF ground”?

No. That would be the wrong lesson to draw.

In this case, the short copper connection to the existing underwater bronze plate became part of a workable and calmer return arrangement once the extra uncontrolled conductors were removed and the filtering/choking strategy was improved. But that is not the same as proving that the bronze plate is some universal RF miracle. The broader lesson is that a return path must be controlled, not romanticized.

That distinction matters. A bronze plate may be part of a functioning marine installation. It does not mean every reference to “RF ground” in ham radio folklore is suddenly correct.

What remained after the solar RFI was solved

Once the controller-related interference was gone, a second and more ordinary engineering question remained: matching and behavior on 15 m and 17 m with a short vertical around 7.68 m total radiator length on a carbon support. That is a very different problem from controller RFI.

At that point the useful questions became things like:

• What is the complex feedpoint impedance on those bands?
• How much is the carbon support detuning the system through capacitive coupling?
• Is the radiator ending up too close to a half-wave condition on 15 m and 17 m, making the impedance more awkward?
• Would a small loading coil or a modest geometry change improve the tuning window?

That is also why this case naturally led into a separate discussion about why a VNA is not the same thing as an antenna-efficiency meter, and why a proper antenna analyzer is usually the more practical tool for this phase of the job.

What this case teaches other boat owners

• Do not trust one-word diagnoses. “RF ground,” “counterpoise,” or “move the tuner” are not complete answers on a boat.
• Do not confuse a temporary improvement with a stable solution. Harbor tests, open-water tests, and later re-tests can behave differently.
• Extra wires can become extra radiators. Random conductors do not automatically improve the return path.
• Filters and chokes matter, but hardware immunity matters too. A noisy or RF-sensitive controller can keep sabotaging the system even when the antenna side is improved.
• SWR is not efficiency. A tuner finding a match does not tell you whether the total installation is radiating efficiently or quietly.
• System troubleshooting beats folklore. The reason this case was solved is that the variables were changed one by one until the actual cause chain emerged.

Why this story is worth publishing

Many external inquiries about marine HF interference end in the same place: generic advice, contradictory opinions, or a tidy theory that does not survive contact with a real boat. Jan’s case is valuable precisely because it did not stop at the first plausible narrative. It kept going until the difference between partial fixes and repeatable fixes became obvious.

That makes this more than a customer anecdote. It is a practical reminder that real RF work is often slow, iterative, and slightly annoying. But that is also how you get to answers that hold up outside the workshop.

Interested in more technical content? Subscribe to our updates for deep-dive RF articles and lab notes.

Questions or experiences to share? Feel free to contact RF.Guru about boat HF installations, counterpoise behavior, and marine RFI troubleshooting.