Short answer: conformal-coat the board and brush silicone or urethane onto the solder joints; keep full potting off the antenna, the lithium cell and the USB port. A coating film is about 0.04% of the 12.5 cm wavelength of a 2.4 GHz signal, so it barely moves the antenna. Full potting buries the antenna in dielectric, and the number to budget is 10 dB of link loss, or roughly 3x shorter range to a phone. The cell is the second constraint: buried in compound it cannot shed charge heat, cannot vent in a fault and cannot be inspected for swelling. Below is the side-by-side, the range arithmetic for a quarter-sized board, and the protection stack to specify by environment.
What potting does to a 2.4 GHz antenna
An antenna tuned in air resonates at 2.4 GHz, and its geometry assumes air around it. Cured silicone has a dielectric constant of about εr 2.7-3.5 and urethane about 3-4, typical values from compound datasheets. Surround the antenna with εr 3 instead of air and the resonance pulls down-band while the compound adds dielectric loss. For a small chip or PCB trace antenna fully immersed, the loss budget to plan around is about 10 dB.
10 dB of lost margin is a 10x cut in received power. In free space, range falls by the square root of that, about 3.2x; the worked example below runs the arithmetic. Two potted devices lose 10 dB each, and their range falls by roughly 10x.
You have three ways out, in the order most boards should try them:
- Mask a keep-out so the antenna stays in air. Module datasheets call for a clear zone of about 10 mm around a chip antenna; a dammed area that size costs little volume.
- Move the RF outside the pot: a module with an antenna connector, and a sealed feedthrough carrying the antenna out of the compound.
- Re-tune for the embedded state. Resonant length scales with 1/√εr, so in εr 3 the arm shortens to about 0.58 of its free-air length. That board under-performs in air, so commit to one state before you build.
If potting is mandatory, pot in stages: release agent on the board, half the compound, cure, test the link on a bench, then finish the pour. It limits the cure heat and leaves you an escape hatch.
The housing shell matters separately: plastic enclosure effects on antenna performance at 2.4 GHz cover what a molded wall does to the same link, and PCB antenna placement and clearance cover the layout side.
Potting versus coating, side by side
| Property | Conformal coating | Full potting (silicone or urethane) |
|---|---|---|
| Film thickness | 25-75 µm dry film (brush and dip practice) | Whatever fills the cavity; 2-5 mm walls on a quarter-sized board |
| Antenna effect at 2.4 GHz | About 0.04% of the 12.5 cm wavelength at 50 µm | Budget about 10 dB; resonance pulls down-band in εr 3 |
| Water protection | Splash, condensation, humidity; adds no rating | Seals the board, but the product still leaks at ports and button |
| Heat | No measurable added thermal resistance | Traps charger and cell heat in the cavity |
| Rework | Film strips and joints re-solder | Cutting back in usually destroys the board |
| Lithium cell | Safe near the cell, keep it off the can | Full encasement blocks heat, vent and inspection |
| Mechanical capture | None | Holds heavy parts, strain-relieves wires |
| Best use | Humid, condensing, chemically dirty air | Vibration, immersion-duty internals, tamper resistance |
The rows that decide this comparison are the antenna row and the cell row. Everything else is preference.
Where full potting wins
Potting earns its place on four jobs. It captures heavy parts — a cell, a holder, a big connector — so vibration cannot fret their joints. It strain-relieves the battery wires, where solder alone fails first. It faces chemical splash and immersion duty that a film cannot. A soft silicone also adds tamper resistance on a consumer build.
For the solder joints specifically, spot-potting is the middle path: dam the area around the battery tabs and connector backshells and fill just that. If you expect to open a joint again, re-enterable potting compound exists for exactly that reason. Skip rigid epoxy on this build; its cure shrinkage and high modulus load large joints and the cell. Silicone or a soft urethane is the fit.
Potting protects the board, never the product. Water still walks in at the USB socket, the pairing button and any cable entry, and sealing those is a separate job: keeping USB and ethernet ports waterproof on an enclosure.
Where conformal coating wins
Coating wins on every axis the antenna and the battery care about. At 50 µm the film is 0.04% of the 12.5 cm wavelength, so the antenna keeps its match. The board sheds heat as before, joints stay inspectable and reworkable, and the cell stays visible. Brushed as a spot coat over the battery solder joints and any reworked pad, it does the exact job you wanted the encapsulate for.
Keep the film off connector contacts, the button and the antenna zone; masking so coating stays out of the connectors is its own small discipline. Chemistry: acrylic reworks easiest, silicone takes the widest temperature range, urethane resists abrasion best. For the outdoor call between film, potting and a sealed inner box, the field practices for conformal coating on outdoor IoT boards walk the same comparison with test data.
The battery: keep compound off the cell
A single lithium cell charged over USB runs warm in two places: the can and the charger. For the common 3.7 V cell charged to 4.2 V at 250 mA, a linear charger dissipates (5 − 4.2) V × 0.25 A ≈ 0.2 W, and the cell adds its own resistive heating on top. Cell datasheets set the charge window at 0-45 °C and discharge to about 60 °C.
A full encasement works against all three numbers. The 0.2 W and the cell heat now soak out through compound instead of air, the safety vent can be blocked at the moment it matters, and swelling stays hidden until it is a hazard. So hold the cell with a clip or a soft pad, loop the wires for strain relief, and brush silicone over the joints only. If a potted assembly ever swells or runs hot on charge, retire it whole; cutting compound to reach a stressed cell is the wrong move.
What potting costs a quarter-sized BLE board at 2.4 GHz
The case: a Bluetooth board and its single cell, each about the size of two stacked US quarters (24.26 mm across, so call it 25 mm), potted as one block in soft silicone, talking to a phone.
Inputs
- Frequency 2.4 GHz, so the wavelength is c/f = 300/2.4 = 12.5 cm (physics).
- Conformal film 50 µm, mid-range of the 25-75 µm brush-and-dip practice.
- Potting loss budget 10 dB for the fully immersed antenna (the planning figure above).
- Baseline range 10 m line-of-sight to a phone in air, typical for a 0 dBm-class BLE link.
Rule: free-space path loss rises 6 dB per doubling of distance, so a margin loss of L dB divides range by 10^(L/20).
Arithmetic
- Coating: 50 µm against 125,000 µm is a ratio of 2500:1 to the wavelength, or 0.04%. No measurable detuning at that scale.
- One potted end: 10^(10/20) = 3.16, so 10 m ÷ 3.16 = 3.2 m. The link to a phone drops from 10 m to about 3 m.
- Both ends potted, board to board: 10 dB + 10 dB = 20 dB total, and 10^(20/20) = 10, so 10 m ÷ 10 = 1 m.
Result: full potting costs about 3x range to a phone and 10x board to board. Decision: if the spec needs more than about 3 m indoors to a phone, the antenna cannot stay in standard compound — mask it, feed it out, or re-tune for εr 3. If 3 m covers the product's use, potting through is viable, with the cell rules of the section above.
The decision by environment
| Where it lives | Protection stack |
|---|---|
| Dry indoor, handled | Acrylic conformal coat, silicone spot-pot on the battery joints, antenna masked |
| Splash, rain, wet hands | Add a sealed USB path: grommet or overmolded pigtail; urethane spot-pots where it gets wear |
| Washdown, hose-down | The stack above, inside a stainless housing manufacturer-rated NEMA 4X / IP66; not UL listed |
| Prolonged submersion | Full pot of the internals, cable out through a potted gland, housing built to a Type 6P class |
Washdown and outdoor are where a housing earns its keep, because potting alone never earns a rating. If the assembly goes into one of our stainless boxes, you set the material, the size and every cutout per piece, and it is delivered in 10 business days from your approval of the drawing; start at custom stainless enclosures or the stainless size finder.
What to put on the spec
- Coating: acrylic or silicone conformal, 25-75 µm dry film; spot-coat the battery joints; keep it off contacts, button and antenna zone.
- Potting where used: a slow-cure, low-exotherm two-part silicone or urethane; read the cured dielectric constant off the datasheet (silicones typically εr 2.7-3.5) before anything RF stays buried.
- Antenna: a dammed keep-out of about 10 mm, or an antenna re-tuned for the embedded εr; confirm it with the staged pour.
- Cell: clip or soft pad, strain-relieved wires, the 0-45 °C charge window honored, and the 0.2 W charger heat given air or a thermal path out of the cavity.
- Ports: USB behind a grommet or an overmolded pigtail; button under a sealed boot.
- Housing for wet duty: stainless, manufacturer-rated Type 4X / IP66, cut for the connectors you actually have.
That stack — film on the board, compound on the joints, air around the antenna and the cell, seals at every port — is the answer for this class of build. Full potting comes in only when the range budget above closes and vibration or immersion demands it.
Catalog dimensions to compare
Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 10 × 8 × 6 in | 304 stainless | 10 x 8 x 6 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 12 × 10 × 6 in | 304 stainless | 12 x 10 x 6 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 16 × 12 × 8 in | 304 stainless | 16 x 12 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 16 × 14 × 8 in | 304 stainless | 16 x 14 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
Will silicone potting block the Bluetooth link?
Only through the antenna. Compound over the RF chip's ground plane costs little at 2.4 GHz, where the wavelength is 12.5 cm; burying the antenna is what detunes it. Mask a keep-out of about 10 mm around a chip antenna, or re-tune for the cured dielectric — silicone εr 2.7-3.5, urethane 3-4 — and confirm the link after the first half of a staged pour.
Can the rechargeable lithium battery stay inside the potting?
No. A cell needs its 0-45 °C charge window, a path for charge heat, and a vent that can open in a fault; encasement removes all three and hides swelling. Hold it in a clip or a soft pad, spot-pot only the wire joints with a strain-relief loop, and retire the whole assembly if a potted cell ever swells or heats on charge.
Silicone or urethane for potting a board like this?
Silicone when you want low stress and a way back in: it cures soft, shrinks little and cuts away from a joint without lifting pads. Urethane when the assembly is handled or abraded, because it grips harder and wears better. Either way pick a slow-cure, low-exotherm two-part; a bulk pour that races its cure cooks the parts it was meant to protect.
How thick can a coating get before the antenna notices?
Keep anything over RF below about 0.5 mm and the effect stays small: 0.5 mm is 0.4% of the 12.5 cm wavelength. Conformal films run 25-75 µm dry, an order of magnitude under that cap. Full potting is a different regime entirely, because the antenna sits inside a dielectric of εr about 3 rather than behind a thin film.
Does potting replace an enclosure rating?
Only the board. Potting adds no rating of its own, and the product still leaks at the USB port, the button and the cable entry. For hose-down duty, put the assembly in a stainless housing manufacturer-rated NEMA 4X / IP66 and not UL listed, seal the USB behind a grommet or an overmolded pigtail, and step up to a Type 6P housing for prolonged submersion.
