The phrase f4b tp-2 pcb appears in a lot of listings, and it hides a category error: TP-2 is not one of the F4B grades. The F4B ladder is a fixed set of Dk points from 2.17 to 3.00 on woven-glass PTFE; the TP series is a separate line whose Dk is quoted as an adjustable range. Worse, most pages describing “F4B TP-2” quote specs that belong to a different laminate entirely. This page untangles the naming, shows which published numbers came from where, and gives the five checks that settle what a TP-2 sheet actually covers.
F4B grade codes are built to be read: F4BM 220 means a woven-glass PTFE laminate at Dk 2.20, and the nine grades run 2.17 to 3.00 in fixed steps. That structure is the whole point — a designer picks the rung that fits and the number is guaranteed.
TP-2 does not work that way. It belongs to a separate series in the same maker’s extended table, and its Dk is quoted as an adjustable range rather than a fixed point — commonly 2.2 to 3.5 depending on the variant. That makes it useful for tuning a design without changing laminate family, and it makes the “F4B + TP-2” framing wrong: one is a fixed ladder, the other is a formulation range. The grade list that separates the branches is in the complete F4B model list.
| Family | Construction | Dk behaviour | Code example |
|---|---|---|---|
| F4BM | PTFE + woven glass, ED copper | Fixed ladder 2.17–3.00 | F4BM220 = 2.20 |
| F4BME | Same, RTF copper | Same ladder, lower PIM | F4BME220 |
| F4BTM | PTFE + glass + ceramic fill | 2.98–3.5 | F4BTM300 |
| TP series | Separate composite line | Adjustable, quoted 2.2–3.5 | TP-2 |
So the first correction is a naming one: a TP-2 datasheet is not a grade selection from the F4B ladder, and treating it as one is how a design ends up with a Dk nobody can reproduce.
The page this replaces described TP-2 with a parameter set that reads exactly like an imported PTFE datasheet. Compare the figures with the laminates that actually publish them:
| Claimed for TP-2 | Where that figure actually belongs | What it implies |
|---|---|---|
| Dk 2.20 ± 0.02 | RT/duroid 5880 tolerance class (2.20 ± 0.02) | Copied, not measured: F4BM220 itself is ±0.04 |
| Df 0.001 @10GHz | 5880-class figure (0.0009–0.0010) | Plausible for a PTFE, unverified for TP-2 |
| CTE x/y 17, z 24 ppm/°C | z 24 matches RO3003’s 25; PTFE Z-CTE is 200–240 | A Z-CTE of 24 on a soft composite is implausible; this is a ceramic-filled figure |
| Thermal conductivity 0.25 W/m·K | PTFE-class order of magnitude | Low; rules out high-power use regardless of grade |
| Water absorption < 0.02 % | 5880-class figure; F4BM220 documents ≤0.08 % | Optimistic by 4× against the documented value |
None of this means TP-2 is a bad material. It means the page describing it was assembled from other datasheets, so none of its numbers can be used in a tolerance stack. That is the practical harm: a designer who trusts ±0.02 builds an impedance budget that no delivered lot will meet.
Because the maker publishes TP-2 as a range rather than a grade table, the sheet has to answer five questions before it can be specified. Ask in writing:
| Check | Question to ask | Red flag if… |
|---|---|---|
| Material class | PTFE, ceramic-filled PTFE, or a hydrocarbon composite? | The answer changes with the salesperson |
| Dk + tolerance + variant | Which Dk in the 2.2–3.5 range, at what tolerance? | A single Dk quoted for the whole series |
| Df at test frequency | Measured value on your variant, at 10 GHz | A range spanning 0.0009–0.0030 |
| Z-CTE and Tg /Td | Z-CTE in ppm/°C and the decomposition temperature | Z-CTE below 50 ppm/°C on a soft composite |
| Process requirements | Desmear route, press window, aspect-ratio limit | No answer — means it has not been built |
This is the same gate discipline used for imported-laminate substitution: if a parameter cannot be checked, it has not been specified. The four-gate version for Rogers grades is in the substitution review.
Assuming a verified sheet lands closer to the composite description than to the copied one, the design consequences are predictable:
| Property | Effect on the design |
|---|---|
| Adjustable Dk | One laminate family covers several electrical lengths — useful for tuning, but every variant needs its own width table and coupon |
| Dk tolerance | Wider than a fixed grade unless specified; budget impedance accordingly |
| Loss | Between F4BM220 and FR-4 in most cases — verify before assuming the 0.001 figure |
| Z-CTE | Determines via qualification; do not inherit 240 ppm/°C rules from PTFE without checking |
| Process | Composite systems usually press nearer FR-4 temperatures than PTFE, which changes the hybrid strategy entirely |
The last row is the one that surprises people: if TP-2 is a thermoset composite rather than a PTFE, it can share a press cycle with FR-4 and skip sodium-naphthalene desmear — which is a cost advantage, and also the reason the material should never be marketed as an F4B variant. Comparing the two process routes is in high-frequency vs standard PCB processing.
| Requirement | Choose | Reason |
|---|---|---|
| Dk reproducibility across lots | F4B fixed grade | A published ladder with ±0.04–0.06 beats a range |
| Tuning electrical length without changing family | TP series | Adjustable Dk is the point of the series |
| Lowest dielectric loss at 10 GHz | F4BM217 / 220 | 0.0010 documented, not a copied figure |
| Antenna area reduction | F4BM255 or F4BTM | Fixed 2.55 or 2.98–3.5 with known tolerances |
| Process compatible with FR-4 line | TP-2 (if verified) | Composite systems usually press at FR-4 temperatures |
The pattern is simple: choose F4B when the design needs a guaranteed number, choose TP-2 when it needs a tunable one — and never choose either on the basis of a page that mixes their datasheets.
The original page listed base stations, radar, satellite, ADAS, medical imaging and test equipment without a frequency or a length for any of them. With the arithmetic from the frequency comparison, each line resolves into a real requirement — or drops out:
| Application | Band | What the laminate must do | Verdict |
|---|---|---|---|
| Base station antenna | 1.8–3.5 GHz | Low PIM, moderate Dk, panel yield | F4BM/F4BME; TP-2 only if Dk tuning is needed |
| 24 GHz sensor | 24 GHz | ≈ 0.37 dB/in total, Dk tolerance drives yield | Any low-loss grade; verify tolerance |
| 77 GHz radar | 76–81 GHz | 0.36 mm lines, ±6 % from etch alone | Needs a tolerance grade + coupon verification |
| Satellite payload | Ku/Ka | Phase stability over thermal cycling | TCDk matters more than Df; check −142 class |
| MRI / CT | 64–298 MHz | FR-4 is adequate | Drop — no benefit, added cost |
Sorting an application list this way is the fastest way to find out whether a material claim is about the material or about the marketing. The reach arithmetic behind the 24 and 77 GHz rows is in the F4B vs FR-4 frequency comparison.
Whatever the sheet says, the first lot decides. Four checks cover most of the risk on a new material:
| Check | Method | Accept if |
|---|---|---|
| Thickness | Micrometer, 9 points per panel | Inside the quoted band and the ±10 % class |
| Dk / Df | Coupon resonator at the design frequency | Within the grade tolerance — not the “typical” value |
| Peel strength | Pull test on a witness strip | ≥ 18 N/cm for PTFE-class, or the quoted value |
| Solder endurance | 260 °C float, 20 s | No blister, no delamination, no warpage beyond ≤0.03 mm/mm |
A material that passes all four on the first lot is a material you can write a drawing around. One that fails Dk on the first lot is a material whose datasheet was never a datasheet.
| Claim | Correction |
|---|---|
| “F4B TP-2 is a PTFE-based CCL” with Dk 2.20 | TP-2 is a separate composite series with an adjustable Dk range, not an F4B grade at a fixed 2.20 |
| Dk 2.20 ± 0.02 | That tolerance belongs to 5880; F4BM220 is ±0.04, and TP-2 has no published grade table at all |
| Z-CTE 24 ppm/°C, “matched to copper” | A 24 figure matches ceramic-filled PTFE, not a soft composite; and 24 is still 1.4× copper, not a match |
| “Suitable for MRI and CT imaging” | MRI runs at 64–298 MHz where FR-4 is adequate; the material adds cost without benefit |
| Applications listed without band or length | Without a frequency and a run length, an application list cannot be checked — and cannot be quoted |
A workable RFQ for a TP-2 build states the variant and its Dk, the impedance target with the tolerance stack, the copper type and weight, the process route expected (PTFE or FR-4-like), and the acceptance coupons. If the maker cannot name the process route, the material has not been through a fab.
XCE builds both routes in-house — PTFE with plasma or sodium-naphthalene desmear and a 260–300 °C press, thermoset composites on the FR-4 line — so the first question a stackup gets is which chemistry it belongs to. 2–72 layers on 600×1200 mm panels, impedance ±10 % standard and ±3 % coupon-verified, samples in as fast as 24 hours, inspected to IPC-A-600G and IPC-6015, 80+ engineers on 10,000+ builds since 2009. The electrical baseline for the F4B grades themselves is in the F4BM220 introduction and the family overview.
Send us your Gerber files and requirements for a manufacturability review and a tailored quotation.