Most antenna material choices come down to a balance between loss, cost, and processability.

But there is a class of designs where the balance point is not loss. It is PIM.

Passive intermodulation is one of the hardest problems to deal with in base station antennas and satellite communication systems. Unlike insertion loss, it cannot be calculated out of a link budget. Unlike return loss, it cannot be seen during tuning. It is a new interfering frequency generated by nonlinearities in materials or contact points when high-power signals pass through passive components. If those frequencies land inside a receive band, base station sensitivity drops—and you often will not realize it at the design stage.

We recently built a 2-layer F4BTME350 board with a finished thickness of just 0.3mm. It serves exactly this kind of application: PIM-sensitive, and height-constrained. But this article is not just about how well the board was built. It is about when F4BTME350 is necessary, and when it is overkill.

The Board Itself

The base material is F4BTME350, a 2-layer rigid PCB. Dimensions are 89.77mm × 45.16mm, held to ±0.15mm. Minimum trace and space is 6/8 mils, and minimum hole size is 0.25mm. No blind vias.

Finished thickness is 0.3mm, with 1oz (35 μm) copper on the outer layers and 20 μm via plating. Surface finish is immersion gold. Top silkscreen is white, with no bottom silkscreen. Top solder mask is black, with no bottom solder mask. Every board is 100% electrically tested before shipment, built to IPC-Class-2, with artwork supplied in Gerber RS-274-X.

The statistics: 12 components, 38 total pads, 9 vias, 2 nets. Not many pads—but each one has a clear job. This is a small, focused design, not a general-purpose board.

The stackup is 35 μm copper, a 0.254mm (10mil) F4BTME350 core, and 35 μm copper. At 0.3mm finished, the board is roughly one-fifth the thickness of a standard 1.6mm FR-4 board. The core is only 10mil, so the board is noticeably flexible, and warpage and handling damage need careful control during processing.

The Real Value of F4BTME350: PIM

F4BTME350 has a dielectric constant of 3.5 and a dissipation factor of 0.0025 at 10GHz. Those numbers are not extreme on their own—the Dk is slightly lower than FR-4, and the Df is better than standard FR-4 but not as low as pure PTFE.

What makes it hard to replace in base station antenna and satellite communication applications is its PIM performance: ≤ -160 dBc.

That number needs context. At 1900 MHz, a well-designed antenna typically targets PIM between -150 and -155 dBc. At -160 dBc, the material's own PIM contribution is pushed down to a level that is effectively negligible.

The key to achieving this is the copper foil. F4BTME350 uses reverse-treated copper foil (RTF) rather than standard electrodeposited copper. RTF foil has lower surface roughness, and rough copper surfaces are a significant source of PIM. The F4BTM series uses ED copper and suits applications without PIM requirements; the F4BTME series switches to RTF foil specifically to minimize nonlinearity in PIM-sensitive designs.

In other words, the premium you pay for F4BTME350 buys PIM, not Dk or Df.

The Processing Reality of a 0.254mm Core

A 10mil core is on the thin side for the F4BTME family. At this thickness, the board's mechanical rigidity comes mainly from the copper layers and the subsequent solder mask and surface finish. The core itself provides limited structural support.

The drilling challenge is the 0.25mm minimum hole size. On a 0.3mm finished thickness, this is roughly a 1.2:1 aspect ratio—not extreme, but on a thin board, support and chip evacuation during drilling need attention. All 9 vias are through-hole, and the 20 μm plating must be uniform across the hole walls under thin-board conditions.

The surface finish is immersion gold. For RF boards, this is a safe choice—no nickel-layer loss problem like ENIG, good flatness, reliable soldering. Combined with F4BTME350's low-roughness RTF copper, immersion gold helps keep conductor loss low.

The board has black solder mask on the top only, with none on the bottom. This is common in antenna boards: solder mask has dielectric properties that can affect RF trace impedance, and removing it on the bottom eliminates that variable while exposing copper for heat dissipation or grounding contact. Black top mask with white silkscreen provides enough contrast for assembly and inspection. No bottom silkscreen, because the bottom is typically a ground plane or mounting surface where markings serve no purpose.

When F4BTME350 Is Necessary

When PIM is the system-level limiting factor. Base station receive sensitivity is directly affected by PIM. If a -150 dBc material would raise the noise floor enough to degrade receive performance, then -160 dBc is not "better"—it is required.

When height is hard-constrained. A 0.3mm finished thickness suits highly integrated antenna arrays or compact RF front ends. A thin board reduces overall stack height while maintaining electrical performance.

When frequency and power are both high. Satellite communications, phased array antennas, high-power amplifiers—these applications are sensitive to both thermal stability and nonlinearity, and F4BTME350's CTE (10 ppm/°C in X, 12 ppm/°C in Y) and low thermal coefficient of Dk have real value under those conditions.

When You Should Not Use It

This is the part most material write-ups skip, and it is the most useful part for readers.

If your design operates below 1 GHz. At low frequencies, PIM is usually not a system-level bottleneck. F4BTME350's -160 dBc specification is overkill. The F4BTM series with ED copper may be entirely sufficient at lower cost.

If PIM is not a system-level concern. Not every RF board runs high power. Low noise amplifiers, small-signal receive chains, lab test boards—in these, PIM will not be the limiting factor, and paying a premium for it makes no sense.

If you do not need 0.3mm. Thin boards are harder to process, with more complex yield control. If module height allows, a standard-thickness F4BTME board is easier to process, more cost-effective, and mechanically more reliable. Thin is not an advantage in itself—it only becomes one when height is constrained.

If cost is the first constraint. F4BTME350's RTF copper and low-PIM formulation make it a premium material. In cost-sensitive applications where PIM requirements are modest, standard FR-4 or the F4BTM series is the more reasonable engineering choice.

The Bottom Line

F4BTME350 is not a general-purpose antenna material. It is a targeted solution: it earns its premium when PIM is the limiting factor in your link budget, and when height must be compressed to the 0.3mm level.

If your design does not meet both conditions, you are likely paying for performance you will never use.