A Technical Case for a Low-Loss, Broad Airband Filter with Real FM Suppression March 2026

James Wiebe
4 days ago
7 min read

FlatTop White Paper

—

Radiant Engineering

March 2026

Prepared for technical and market-facing view. Measurements cited herein are assembled-unit VNA results from FlatTop development hardware.

FlatTop was built to solve the actual problem: keep the airband flat, keep insertion loss low, and hit FM hard enough to matter in real receivers.

Executive summary

Most compact “airband filters” force a compromise. Some are broad enough to cover the aviation band, but they throw away too much wanted signal and require a separate FM trap. Others are sharp enough to look impressive, but they are too narrow to be practical broad-airband front ends. FlatTop was built to reject that false choice. It is a 118–136 MHz broad airband bandpass filter designed to keep insertion loss low, maintain a useful passband across the civil COM range, and suppress the adjacent FM broadcast region hard enough to matter in real receivers, all in one box.

The real problem we set out to solve

In the real world, many receivers do not fail because the receiver is “bad.” They fail because the front end is forced to swallow strong out-of-band energy, especially broadcast FM, while trying to receive weak airband signals. Both RTL-SDR Blog and GPIO Labs explicitly describe strong FM broadcast energy as a source of receiver overload and desensitization.

There is an important nuance that many products gloss over. A pure FM notch can kill FM very effectively, but it is not the same thing as a well-shaped airband bandpass filter. A narrow tuned airband filter can look excellent on paper, but may not be practical when the use case is the whole 118–136 MHz COM band. A broad compact airband filter is only useful if it does not throw away too much wanted signal.

FlatTop was engineered around that specific problem statement, not around the easiest marketing graph. It rolls both needs into one product: excellent insertion loss across aviation COM frequencies; superior attenuation of interfering FM signals.

The design target

• Broad civil airband coverage: 118–136 MHz

• Low insertion loss across that entire band

• Controlled ripple and useful skirt behavior

• Real suppression in and below the FM region

• Compact, manufacturable construction

The design was not a one-shot simulation exercise. FlatTop’s development was iterative hardware engineering over roughly a year: prediction, layout, build, VNA measurement, retune, relayout, production board, repeat. At VHF, the final result depends on component values, layout parasitics, coil geometry, launch geometry, shielding, and enclosure interaction.

How we did it

1) Start with an unusually strong fundamental design

The core electrical design was sound from the start. FlatTop’s basic topology was capable of excellent passband form, and the values landed in the right neighborhood for a broad 118–136 MHz airband filter. A weak starting design can be “tuned” for weeks and still never become great. FlatTop’s starting point was not weak.

2) Treat layout as part of the filter

At these frequencies, the PCB is not an innocent bystander. Trace lengths, pad geometry, ground currents, coil placement, and proximity effects are all part of the filter. FlatTop’s layout was engineered, not merely routed. The VNA trace kept getting better as the physical realization approached the intended electrical behavior.

3) Use the enclosure as part of the RF solution

This is where FlatTop becomes more interesting than a normal board-level filter. The can is not just a shield. In the final assembly, the can lid is part of the electronics. It contributes capacitance, changes coupling, and materially improves the FM-notch behavior. In measured units, the lid-on assembly consistently produced a deep notch in the FM region around 105 MHz.

Privately, we will admit that part of this outcome was lucky. When we discovered it, we locked it into the design. Publicly, the more important point is that the FM trap effect is real, repeatable, and now engineered into the product.

4) Iterate until the hardware, not the simulation, is right

FlatTop was pushed toward its present form through repeated build-and-measure cycles: prototype filter concepts, production-fabricated boards, coil geometry changes, VNA characterization, relayout and retuning, and repeated confirmation on assembled hardware. This is how good RF hardware gets built — with a loop of theory, measurement, humility, and stubbornness. It takes time and a “never quit” attitude.

Measured FlatTop performance

In four assembled units, FlatTop showed the following measured VNA results.

Sample	Passband IL across 118–136 MHz	Attenuation at 108 MHz	FM-band notch near 105 MHz
1	0.85 to 1.65 dB	34 dB	49–54 dB class
2	0.83 to 1.78 dB	41 dB	49–54 dB class
3	0.76 to 1.68 dB	36.5 dB	49–54 dB class
4	0.97 to 2.00 dB	43 dB	49–54 dB class

These are measured assembled-unit results. The design takeaway is simple: FlatTop is now a repeatable product, not a heroic one-off sample.

Figure 2: Proof Of Life: Flattop production

Why that matters more than a pretty graph

Earlier in development, we evaluated a low-cost imported filter that was impressively ripple-free. On first glance, it looked polished. Its insertion loss was about 5 dB. That is the cautionary tale.

A smooth trace is not the goal. A useful trace is the goal. In receiver-front-end work, 5 dB of loss is not an academic blemish; it represents a loss of more than ¾ of the original signal strength. It is a direct tax on wanted signal. That is exactly the kind of tradeoff FlatTop was designed to avoid. If a compact filter looks gorgeous but eats 4–5 dB of signal to get there, it is still a poor answer for weak-signal airband work.

Competitive analysis

Against compact broad airband filters (using their published data)

GPIO Labs’ enclosed 118–138 MHz Airband Bandpass Filter is one of the clearest reference products in the small, affordable category. GPIO publishes 4 dB attenuation across the passband and 10 dB attenuation at 108 MHz. Against that published claim, FlatTop’s measured result is not subtle: about 0.8–2.0 dB across the passband, 34–43 dB attenuation at 108 MHz, and a repeatable 49–54 dB notch near 105 MHz.

That means FlatTop is not merely “better tuned.” It is solving a different level of problem. Relative to the common compact broad-airband filter class, FlatTop preserves more wanted airband energy and rejects far more unwanted FM-band energy.

Against dedicated FM notch filters

A dedicated FM stop filter can still win on raw FM murder. RTL-SDR Blog advertises over 50 dB attenuation across 88–108 MHz with less than 0.5 dB insertion loss from 0 to 1 GHz. GPIO’s 9th-order FM notch advertises 40 dB attenuation at 108 MHz and only 0.4 dB attenuation at 150 MHz. FlatTop is not a pure notch. It is a broad airband bandpass with built-in FM suppression.

The honest comparison is this: if the mission is pure FM elimination, a dedicated notch remains the specialist’s tool. If the mission is whole-airband cleanup in one box, FlatTop offers something more integrated and, for many users, more useful.

Against compact tuned airband filters

ZCG’s ZLBF Series compact airband filter publishes ≤1.5 dB typical insertion loss, but only over a ±1 MHz pass bandwidth, with factory tuning and >30 dB rejection at f0 ± 5 MHz. That is a good product for a different job. ZCG’s compact filter is optimized as a narrow tuned channel-region filter. FlatTop is optimized as a broad all-airband front-end filter.

Against cavity filters

Amphenol Procom’s 118–136 MHz cavity products explicitly emphasize large ø200 mm cavities, high Q, and very narrow passbands, with adjustable insertion loss figures such as 0.6–2.0 dB in one published variant. Those are serious products for serious fixed-site roles. FlatTop is not trying to replace a cavity in a rack. It is trying to win in the compact front-end space where cavity hardware is too large, too specific, or too expensive to be the right answer.

What makes FlatTop technically credible

• It is broad without being lazy. The passband stays broad enough for the whole civil COM range while preserving low loss and useful shape.

• It is low-loss without being toothless. FlatTop still delivers meaningful rejection at 108 MHz and a deep notch near 105 MHz.

• The enclosure is not cosmetic. The can lid materially changes the response for the better and is part of the tuned RF structure.

• It was engineered iteratively, not decorated after the fact. Strong base design, careful layout, repeated measurement, informed retuning, and production validation all matter here.

• It performs like a product, not a cherry-picked board. Four similar assembled units create credibility.

Where FlatTop is better, plainly stated

FlatTop is better than many compact broad-airband filters because it delivers a stronger combination of lower insertion loss across the full airband, more meaningful suppression near the FM edge, repeatable assembled-unit behavior, and a real integrated front-end solution instead of a partial fix. That is the heart of the case — not “best at every metric in the universe,” but better where real compact airband users actually live.

Where we are deliberately not over-claiming

• FlatTop is not claiming to outperform a dedicated FM notch at pure FM stop depth.

• FlatTop is not claiming to replace a high-power cavity filter in infrastructure applications.

• FlatTop is not claiming that luck played no role in discovering how beneficial the lid geometry would be.

Instead, FlatTop claims something more useful: through design discipline, careful layout, enclosure-aware engineering, and a year of iteration, Radiant built a compact broad-airband filter that appears to beat the usual compact alternatives where it matters most.

Conclusion

FlatTop did not become good by accident. Its final form came from an astonishingly good fundamental design, layout treated as part of the filter, enclosure behavior exploited rather than ignored, and iterative measurement-driven development over roughly a year.

The result is a compact 118–136 MHz airband bandpass filter with roughly 0.8–2.0 dB measured insertion loss across the passband, 34–43 dB measured attenuation at 108 MHz, and a repeatable 49–54 dB notch near 105 MHz in assembled units. Some competitor products are broader but too lossy. Some are sharper but too narrow. Some are excellent notches, but not real broad airband front ends. Some are true professional filters, but in a different class entirely. FlatTop’s achievement is that it occupies the middle ground unusually well: broad enough, low-loss enough, and aggressive enough below the band to be a better compact airband answer than the usual compromises.