It Depends—and That Is Where Engineering Starts

Last week I received an email that made me stop, smile and sit quietly for a moment before replying. It was not a new technical puzzle, although I love those. The message put words to something I have felt for years but do not always say out loud.

If you have ever written technical material for a hands-on and opinionated community, you will recognise the gratitude—and the relief. Someone had found a text that did not merely say what to do. It also admitted why the same action may work in one station and fail in another.

Clean Answers Feel Good

Most of us, me included, like certainty:

• Use this many radials.
• A dipole is always better than a vertical.
• A magnetic loop is efficient on HF.
• Put the choke here.
• Ferrite always fixes common mode.
• The book says so.

Those sentences give us something concrete to do. Antennas, however, do not live only in textbooks. They live in awkward gardens, on uneven roofs, beside gutters, above soil whose properties change with location and moisture, inside apartments full of switching electronics, and under trees that grow exactly where we did not want them.

In that world, always is an expensive word.

“It depends” is an admission that a conclusion belongs to stated conditions. In amateur radio those conditions change constantly—and sometimes without our noticing.

A Model Is a Foundation, Not the Finished House

Textbooks are not the problem. They give us a shared language and isolate mechanisms so we can learn them. A useful model must choose a boundary and make assumptions: free space or specified ground, ideal conductors, a uniform line, a particular source and load, or a noise environment represented by one number.

The trouble starts when a result crosses into another installation without its assumptions. A free-space pattern is not an installed pattern. A component’s small-signal impedance is not its high-power thermal rating. A 50 Ω VNA trace at one reference plane does not establish antenna efficiency, field strength or common-mode current elsewhere.

LLNL’s Numerical Electromagnetic Code is powerful precisely because geometry, segmentation, ground and excitation are declared inputs. A model can test a mechanism under those inputs. It does not silently include the gutter, cable bundle or wet tree that was never described.

How Context Evaporates Into Folklore

I have seen the pattern again and again. A statement appears in a book, is repeated during a club presentation, becomes a rule of thumb and eventually hardens into dogma. Someone tries it under different conditions, gets a different result and concludes either “I did it wrong” or “the book is wrong.”

Sometimes an explanation is wrong. Sometimes the setup is different. Sometimes both are true. More often, the useful conditions were dropped while the conclusion travelled.

“This worked for me” becomes “this works.” Context evaporates; the slogan survives.

The Reality Factor Has a Measurement Boundary

The phrase Benjamin used—the reality factor—does not mean abandoning theory. It means attaching theory and measurement to the system that produced the result.

Sometimes everything is built competently and the result is still modest because space, loss, noise or propagation sets the limit. That is not a moral failure. It is the installation telling us where the boundary is.

Reader Questions Begin Where Diagrams Stop

Benjamin also mentioned that I use reader questions as article starting points. Yes—I receive many, and I am grateful for them. They are usually more valuable than topics invented in isolation.

A textbook can begin with, “Consider a half-wave dipole in free space.” A real email begins with, “The SWR looks fine, the reports are poor, and something at the operating position tingles or heats when I transmit.”

That last observation is not a diagnostic instrument. Stop transmitting and treat unexpected RF contact, heating or arcing as a safety fault until the current path and exposure have been assessed.

The useful questions then become:

• What should I do next under the constraints I actually have?
• Which variable is likely to dominate this result?
• What is the simplest safe change that would distinguish two explanations?
• At which reference plane—and with what uncertainty—was the evidence measured?

That is why my writing returns to context, trade-offs and measurement. Pretending a difficult system is simple wastes people’s time.

Nuance Is Not Randomness

“It depends” does not mean nobody knows anything, every antenna is equivalent or intuition outranks physics. It means a few variables can dominate the outcome, and we should identify them before prescribing hardware.

A useful answer has the shape of a map:

• If condition A applies, this action follows.
• If condition B applies, choose the other path.
• If we do not know which condition applies, make a measurement that separates them.

The NIST uncertainty guide expresses the same discipline in metrology: define the measurand, identify important influence quantities and report uncertainty instead of treating one displayed number as perfect truth.

Why Antenna Myths Are So Durable

Antennas combine invisible fields and currents, difficult-to-control surroundings and selective feedback. Propagation changes. Signal reports use different receivers, bandwidths and operators. Memories favour the remarkable contact. A change made on a good day can receive credit for what the ionosphere did.

Experience is not meaningless. It becomes useful evidence when the comparison states its conditions, controls obvious confounders and repeats. The IEEE 149 antenna-measurement practice exists because a pattern, gain or polarisation claim requires a defined range, geometry, instrumentation and uncertainty—not merely a successful contact.

For noise, Recommendation ITU-R P.372 separates atmospheric, galactic and man-made radio noise by frequency and environment. It is another reminder that “the noise floor” is not one universal property of an antenna label.

The Promise Behind These Articles

When I turn a question into an article, I try to keep a few promises:

• Name the assumptions. What must be true for the recommendation to hold?
• Offer a path, not a slogan. Give the reader a decision or test, not just a conclusion to repeat.
• Separate physics from preference. Cost, effort, risk, appearance and space are legitimate design constraints, but they are not RF laws.
• Respect the station people can build. Not everyone can install a tower, bury radials or route open-wire line.
• Encourage proportionate measurement. A limited measurement with a declared boundary is more useful than borrowed certainty from someone else’s garden.

If I fail at this—and sometimes I will—I hope readers call it out. The goal is not to become “the person with the answers.” The goal is to make the community better at asking the right questions.

Thank You, Benjamin

Thank you for the kind words, but especially for reminding me why it is worth writing the slower, less absolute version of the story. The version with conditions. The version that can end with three valid options because the best one depends on the station.

And yes, there is a practical side: I do commercial work, and I will not pretend otherwise. I have always believed that sharing knowledge openly and honestly helps the right people find RF.Guru for the right reasons.

So keep the questions coming. Keep challenging absolutes. Keep insisting on the reality factor.

Engineering References

• NIST Technical Note 1297: Evaluating and Expressing Measurement Uncertainty
• IEEE 149-2021: Recommended Practice for Antenna Measurements
• Recommendation ITU-R P.372: Radio Noise
• Lawrence Livermore National Laboratory: Numerical Electromagnetic Code v5

Questions, antenna-factor records or height trials to share? Contact RF.Guru.