Satellite Terminals, Hotspots and the New Accessory Category

Walk the accessory aisle of any electronics retailer and you will find a wall of phone cases, a rack of tablet folios, and nothing at all for the satellite internet terminal that started shipping in consumer volumes. Protective case sourcing for satellite internet terminals and mobile hotspots puts a buyer in an unusual position: no incumbent brand owns the shelf, no factory has a public mold sitting on a rack, and there is no reference design to copy from a competitor's teardown. That is the opening and the problem in the same sentence.

Be careful with the first half of that sentence. Nobody can hand you a defensible unit-demand forecast for this category, and any supplier who offers you one is selling, not informing. What can be described honestly is the engineering: these devices are genuinely harder to case than a phone or a payment terminal, and that difficulty is what decides whether your programme lands on time or dies at second sample. The rest of this article is about the difficulty.

The category arrives before its accessory ecosystem does

Accessory demand for a new hardware category tends to arrive roughly one product generation after the hardware itself. The first generation goes to early adopters who accept the device as-is. The second generation goes to people who mount it on a truck, drop it in a Pelican case, carry it to a job site, and lend it to a colleague — and those people go looking for a case that does not exist.

By the time that demand is visible in search data, tooling has to already be cut. A silicone case for a device you have never handled is not a two-week project; it is CAD interpretation, a mold, a T1 sample, a fit review on real hardware, steel modifications, and a T2. The suppliers who make money in a new category are the ones who started before the demand was legible.

The same lag has already played out one category over: handheld thermal imagers reached mainstream trade pricing well before anyone tooled decent cases for them, which is the subject of a companion piece on thermal cameras going mainstream while the cases did not.

Connectivity hardware — fixed and portable satellite terminals, 5G mobile hotspots, LTE routers used as field kit — is the current example. It also happens to be one of the harder families to case correctly, for three reasons that have nothing to do with marketing.

Why connectivity hardware is hard to case

The antenna keep-out zone is a steel condition, not a note on a drawing

Every one of these devices radiates. A hotspot has cellular and Wi-Fi antennas, usually printed on flex or stamped metal, often tucked behind a plastic window on one face. A satellite terminal has a much larger aperture and frequently a GNSS patch antenna on the top surface for positioning. A case that covers those surfaces changes their behaviour.

The mechanism is dielectric loading. A silicone wall pressed flat against a patch antenna raises the effective permittivity the antenna sees and shifts its resonant frequency down. On a narrowband patch, a couple of millimetres of rubber in direct contact is enough to move the resonance off the band and cost you real signal. The fix is not a material choice — it is geometry. You either hold an air gap over the radiating area using standoff ribs, or you thin the wall in that zone, or you open a window. All three are decisions about the shape of the steel, which means they have to be settled before the mold is cut, not after the first sample tests badly.

Pigment matters too, and this is the detail most accessory briefs miss. Black silicone is normally coloured with carbon black, and carbon black is not electrically inert — at high loading it raises dielectric loss. If your product is a black case and the antenna window sits under a black wall, ask the molder which black pigment system the compound uses, and whether a non-carbon-black black (iron-oxide or organic pigment based) is available for that region of the part. The same warning applies harder to any metallic-effect finish: metal-flake pigments, vacuum metallising and chrome-look trim over an antenna zone are a straightforward way to ruin a device. If you want a metallic accent, put it somewhere the RF engineer has blessed.

These devices run hot continuously, and silicone insulates

A phone is thermally bursty. A satellite terminal or a hotspot serving a household is a continuous-duty radio: the power amplifier and modem run for hours, and the enclosure was designed by the device OEM as part of the heat path. Wrapping it in rubber changes that calculation.

Silicone rubber is a thermal insulator — its conductivity sits around two orders of magnitude below aluminium. A full-wrap case with no ventilation raises skin temperature, and on a device with thermal management in firmware the visible result is not a melted case, it is throttled throughput on a hot afternoon. Your customer will not blame physics; they will blame your case.

The design answer is a case that stands the device off its own hot faces and gives air somewhere to go: vent channels moulded through the wall, standoff ribs that break the contact patch into strips rather than a continuous blanket, and open regions over the hottest zone rather than a decorative grille that is closed on the inside. It is worth asking a prospective molder directly whether it has done thermal-relief geometry before or whether your project is where it learns. A supplier that already runs cases for thermal cameras, 3D scanners and AI edge boxes has met this problem; one whose portfolio is phone cases has not. A companion piece on designing heat out of a case for AI edge and thermal-camera hardware goes through the geometry in detail.

Outdoors is a different specification from in-pocket

Phone cases live in pockets. Satellite terminals live on roofs, poles, vehicle racks and job sites; hotspots live on dashboards and in truck cabs. That moves the durability question from abrasion to UV, ozone and thermal cycling.

Silicone is the better base polymer for that duty than TPU — thermoplastic polyurethane is tough and clear but yellows under UV and is vulnerable to hydrolysis in hot, humid service. But the base polymer is only half the answer. Colour retention outdoors is governed mostly by pigment lightfastness, not by the rubber, and a poorly chosen organic pigment in a bright colourway will fade long before the part fails mechanically. If the product is going outside, ask for the pigment system, not just the Shore hardness, and ask what accelerated exposure evidence exists for that specific colour.

What to compare across device families

If you already source cases for one device type, the useful exercise is to see where connectivity hardware differs from what you know.

The pattern is that the further right you go, the more of the design is dictated by things a case designer cannot see — and the less any public mold library helps you. This is the opposite of the POS terminal business, where a factory holding ready-made molds for common terminal models can sample in days. Nobody holds a rack of satellite terminal molds. You are paying for first tooling, and you should plan the money and the calendar accordingly. If you have not costed that before, the mechanics of mold cost in China is the background reading.

Proving the case does not break the device

This is the step most accessory brands skip, and it is the one that turns a category bet into a defensible product. You do not need a certified anechoic chamber to catch the obvious failures. You need a repeatable comparison.

A workable pre-tooling protocol:

1. Fix the environment. Same location, same mount height, same orientation, same time window. Log it. Changing any of these invalidates the comparison.
2. Baseline the bare device. Record whatever the device exposes — signal metric, SNR, throughput on a fixed test — over a run long enough to see variance, not a single reading.
3. Test with a 3D-printed mock of the case geometry before steel is cut. Print in a material with roughly comparable permittivity to your intended silicone, or better, cast a soft prototype. The point is to catch a keep-out mistake while it costs a print, not a mold modification.
4. Repeat in at least three orientations. Antenna problems are directional; a single azimuth will hide them.
5. Re-run on the T1 sample in the production compound and the production colour. Pigment and cure differ from your prototype. If you specified a metallic or high-loading black, this is the run that tells you.
6. Run the thermal case in parallel. Loaded device, cased and uncased, ambient logged, for long enough to reach steady state — several hours, not twenty minutes. Watch for throughput decay, which is the fingerprint of firmware throttling.
7. Record everything in a one-page report and make it an attachment to the purchase order. A supplier who has signed against a test report behaves differently from one who has not.

Ask any candidate supplier which of those steps it can run in-house and which you will have to arrange. There is no shame in a molder subcontracting RF work — there is a real problem in one that claims capability it cannot show you. For the wider question of how to interrogate a supplier on radio-adjacent hardware, the CMH guide to sourcing RF and sensor components in China covers the vocabulary.

Where a molder fits into a category with no reference design

Because there is no public mold, the whole programme rests on custom tooling capability: 3D data interpretation, structural review, material selection, and the willingness to modify steel after a fit review rather than argue about whose CAD was wrong.

WJM Silicone is a reasonable example of the profile to look for. It is a 12,000 m² plant in Longgang, Shenzhen, running silicone compression molding and plastic injection under one roof with a 12-person R&D team handling 3D modelling, structural analysis, material selection and thermal simulation, and it lists satellite internet terminals, mobile hotspots and AI edge devices among the categories it cases, with vent channels and heat-dissipation structures as stated design features. Its 72-hour sampling figure applies to its existing public molds for payment terminals — a new connectivity device is a custom tooling project, and you should ask for that lead time separately rather than assuming the headline number carries over. The same applies to RF validation: treat it as an open question to put to the supplier, not an assumption. Whether your programme runs to plan depends far more on how the sample-to-production timeline is managed than on any single spec on a quotation.

Common questions

Can a silicone case really affect a satellite terminal's performance?

Yes, and the mechanism is well understood. Material in direct contact with a radiating element loads it electrically and shifts its tuning; conductive or lossy fillers make it worse. The size of the effect depends on the antenna's bandwidth and how close the material sits. A wide-band antenna with a plastic radome already designed in may barely notice a thin case; a narrowband patch under a thick black wall can lose meaningful signal. This is why the test protocol matters more than any supplier's assurance.

Should the case be silicone, TPU, or a two-material build?

For outdoor, continuous-duty devices, silicone handles UV, ozone and sustained heat better than TPU. TPU offers higher tear strength and better optical clarity where you want a transparent zone. Where a case needs both a rigid mounting interface and a compliant bumper, an overmolded construction — soft material moulded onto a rigid PC or ABS carrier — is the usual answer, and it needs a factory that runs both processes rather than subcontracting one of them.

How do I get dimensional data if I am not the device OEM?

Two routes: a 3D scan of a physical unit, or measured CAD reconstruction. Scans are fast but carry error on sharp features and deep recesses, so a good molder will want the physical device as well as the scan file. Send the actual unit for the fit review, and expect the molder to ask for both — the files a case molder needs before it can quote covers what a usable data package looks like — a case approved against a scan and never tried on hardware is the classic route to a T2 modification you pay for.

What is the real risk in a new-category accessory programme?

Device revision. Connectivity hardware gets mid-cycle hardware changes, and a few tenths of a millimetre moved on a port or a boss can make your mold obsolete. Write the CAD revision or the device hardware revision into the tooling agreement, agree in advance who pays for steel changes if the device changes, and keep the first production run small enough that a revision does not strand a container of inventory.

What to ask the supplier next

Before you commit tooling money on a device in this category, put these in writing:

• Which of the device's antenna zones do you propose to keep clear, and how — air gap, thinned wall, or open window? Show it on the 3D model.
• What pigment system does the production compound use in black, and is a non-carbon-black option available for the antenna region?
• What thermal relief is designed into the case, and can you show a previous part where you did the same thing?
• What is the lead time from approved 3D data to T1 sample on custom tooling, and what does a steel modification cycle add?
• What RF or environmental testing happens in-house, what is subcontracted, and which lab?
• Who owns the mold, where is it stored, and what happens to it if the device is revised?
• Send a soft prototype or 3D print of the proposed geometry before steel is cut.

Send the physical device, not just the file, and ask for the fit review to be done on it in front of you on a video call. In a category with no reference design, the supplier's willingness to iterate before cutting steel is the single best predictor of whether the programme ships.