The capability extract shows in thirty seconds what runs as standard and what carries a surcharge.
See capabilities →Send the files and an engineer reviews them against your build — free, before you enquire.
To the data check →At 77 GHz a wavelength in the laminate is under two millimetres, the skin depth is a fraction of a micron, and an etch error of ten microns is a measurable impedance error. The board stops being a carrier and becomes the antenna.
Three things dominate radar board performance and none of them is layer count. The loss factor of the laminate determines how much of the transmit budget survives the feed network. The stability of the dielectric constant over temperature determines whether the antenna is still tuned at −40 °C and at +105 °C. And the etch tolerance determines whether the structures you drew are the structures you got.
The economics push firmly towards hybrid build-ups. RF laminate on the two layers carrying the antenna and feed, FR-4 for the rest of the stack, bonded with a matched prepreg. The performance where it matters is essentially unchanged and the material bill falls dramatically — which on a high-volume automotive radar is the difference between a viable product and an interesting prototype.
RF laminate only on the antenna and feed layers, FR-4 elsewhere. The single most effective cost measure available on a radar board.
± 10 µm on RF layers, because on a 200 µm line that is already a 5 % impedance error. Line width control is the real performance parameter.
Radar sits inside a vehicle programme, so it needs the same IATF-oriented processes, PPAP support and traceability as any other automotive board.
Values for the RF portion of a hybrid build. The FR-4 layers follow the normal multilayer figures.
| Parameter | Typical | Note |
|---|---|---|
| Dk (77 GHz) | 3.0 – 3.6 | Held to ± 0.05 across the panel |
| Df (77 GHz) | < 0.004 | Sets insertion loss in the feed network |
| Copper foil | VLP < 1 µm | Skin depth is under 0.5 µm at 24 GHz |
| Etch tolerance | ± 10 µm | Line width error becomes impedance error |
| Build | hybrid | RF layers plus FR-4 core |
| Impedance coupon | per panel | TDR measured, report available |
| Surface finish | immersion silver | Lower loss than ENIG above 10 GHz |
| Registration | ± 75 µm | Antenna geometry to reference plane |
Guide values for pre-selection. Binding values are confirmed per project against your data.
The points below are what most often comes back as a query after data review. Settling them before layout freeze saves a loop.
Put RF laminate only on the layers carrying the antenna and feed. A full RF stack costs several times more and, on almost every design, performs no better where it counts.
At 24 GHz the skin depth is smaller than the roughness of standard foil, so the current follows every peak and valley. Low-profile foil is part of the loss budget, not a refinement.
If a 10 µm width error breaks the design, the design is too tight for volume manufacturing. Widen the critical structures rather than tightening the tolerance.
An antenna tuned on the bench and deployed across an automotive temperature range sees real Dk drift. Get the material curve before the tuning is frozen.
Hybrid, in almost every case. The RF material is only doing work on the layers carrying the antenna and feed structures; everywhere else it is expensive FR-4. The exception is a design where several layers carry critical structures.
Immersion silver. ENIG puts nickel in the current path, and nickel is magnetic and lossy at the skin depth the current occupies — above roughly 10 GHz that becomes measurable.
We publish ± 10 %, proven on a coupon. Anything tighter on the feed structures is subject to a technical review of material and stack-up. At millimetre wave the material Dk tolerance often contributes more error than the geometry does, so the material specification matters as much as the tolerance class.
If it is going into a vehicle, yes — same IATF-oriented processes, PPAP support, IMDS and traceability as any other automotive board. The RF requirements sit on top of that, not instead of it.
Send us your data or a rough sketch. We check feasibility, name the cost drivers and answer within one working day.