Medical telemetry is the only radio category where the antenna is often inside a human body, the transmitter may be powered by a battery that cannot be replaced, and the environment is a dense soup of Wi-Fi, nurse-call systems and imaging equipment. That combination makes it a poor fit for generic IoT design habits. A pacemaker link that reconnects slowly is an inconvenience; a telemetry link that cannot be interrogated is a surgical procedure. This article explains how implant and body-worn links are architected, what the newest standards and clinical workflows require from the radio, and where domestic silicon can realistically be screened, including the harsh truth that one whole segment has almost no domestic supply. For the broader device-class picture, see our medical and industrial RF overview.
Application logic: three radios, three constraint sets
- Implant telemetry (MICS, 402 to 405 MHz) — a tiny transceiver inside the device talks to an external programmer or home monitor a couple of metres away. The band is globally allocated for this purpose, the emission limits are severe, and the link must work through tissue, clothing and a metre of air with a battery measured in microamp-years.
- Body-worn patch (BLE, 2.4 GHz) — ECG, SpO2 and continuous glucose patches stream to a phone or gateway. The radio is commodity, the constraint is average current: a patch must run for days on a 30 to 60 mAh cell while sampling continuously.
- Clinical gateway (Wi-Fi or cellular) — bedside monitors and telemetry servers aggregate patch data onto the hospital network. Here the radio is standard, but the security and availability requirements are not: patient data cannot leave the premises unencrypted, and a gateway that drops a session mid-shift is a clinical event.
- Coexistence — all three live in the same ward. MRI suites add pulsed fields, electrosurgery adds broadband noise, and a crowded 2.4 GHz band is normal rather than exceptional.
- Energy and thermal limits — an implant cannot dissipate heat into tissue beyond strict limits, and a patch cannot exceed a skin-contact temperature. Both constraints push toward low duty cycle and short transmissions.
- Regulatory surface — EMC per IEC 60601-1-2, radio approvals per region, and for implants additional absorption and MRI-compatibility evidence. The radio choice is therefore a regulatory choice as much as an electrical one.
The parameters that decide a link that works through a body
- Link budget through tissue — tissue loss at 400 MHz is several decibels higher than free space and varies with posture. Receiver sensitivity and antenna efficiency at the implant end decide the usable distance far more than transmit power, which is capped tightly.
- Duty cycle and latency — implant links are duty-cycled by regulation and by energy. What matters is how quickly the external unit can wake the implant, exchange a burst and shut it down again; a slow wake cycle drains the battery that cannot be changed.
- Average current — for patches, the honest metric is microamps averaged over a day, including the sensor front end, the radio and the DC-DC converter quiescent draw. A radio datasheet quoting microamps in sleep says nothing about a design that keeps an ADC running.
- Coexistence robustness — the receiver must survive a neighbouring Wi-Fi access point at close range without losing a sample stream; blocking and adjacent-channel rejection are the figures to ask for.
- Security — pairing, key storage and authenticated firmware updates are now expected; regulators and hospital IT both ask for evidence, and a radio without secure key storage forces an external element into an already tiny board.
- Reliability and longevity — medical volumes are small compared with consumer volumes, so the supplier’s process stability and long-term availability matter more than the unit price. An end-of-life notice on a medical BOM is a redesign, not a re-order.
What is changing: home monitoring, dual-band implants and BLE everywhere
Three shifts are visible. First, care moved home: remote interrogation of implants and continuous patch monitoring turned telemetry from a clinic-only function into a 24/7 service, which means the link must survive domestic Wi-Fi and a patient who walks around the house. Second, dual-band implants: MICS remains the long-range workhorse, but some designs add a 2.4 GHz BLE channel for fast bulk uploads when a programmer is in contact, splitting the energy budget between a low-rate always-available link and a high-rate occasional one. Third, BLE became the default for wearables, which pushed silicon vendors to integrate the analog front end, the radio and the security block into one low-leakage die, and pushed designers to treat average current, not peak performance, as the headline specification. A parallel trend is wireless charging for patches and implants, which removes the connector but adds a coil and a charging controller into an already constrained mechanical envelope.
Domestic alternatives: suppliers and indicative cost
Be direct about the segmentation. For MICS-band implant transceivers, the volume supply has historically come from a small number of imported specialists; domestic supply at chip level is thin and should not be assumed. For body-worn BLE, the picture is completely different: Tailing Micro’s TLSR series and Unisoc-class BLE SoCs are widely used in domestic wearables, and Artery’s low-power MCUs handle the sensor and power side. For the analog front end such as ECG, bio-impedance and pulse oximetry, domestic suppliers including Chipsea have published medical-oriented AFE families, though the qualification evidence for a specific clinical claim must be obtained per project. The practical domestic strategy for a patch product is therefore a domestic BLE SoC plus a qualified AFE, with the imported part retained only where a regulator has already approved it in your predicate device.
A realistic patch BOM: a Nordic nRF52840 as the reference BLE platform (stocked for comparison builds), an Artery AT32F421C8T7 for sensor scheduling and power sequencing where a separate host is needed, a COS29152WU for the always-on analog rail, and a TS2502PN on the charging contacts that the patient touches every day. Our low-power Bluetooth SoC article compares the BLE options in more depth.
Price figures below are indicative public-market reference ranges for moderate volumes, compiled from publicly observable listings. They fluctuate with the spot market, package option, grade (industrial, automotive, screened) and order quantity, and must not be read as a quotation. Ask us for a firm price on your exact part numbers and quantities.
| Function (imported anchor) | Example import part / class | Indicative import price (USD) | Domestic candidate | Indicative domestic price (USD) |
|---|---|---|---|---|
| MICS-band implant transceiver | Imported MICS transceiver class | 6.0 – 14.0 | No volume domestic equivalent identified (family-level gap) | — |
| Body-worn BLE SoC | Nordic nRF52 class | 2.0 – 4.0 | Tailing Micro TLSR series (family level) | 0.80 – 1.80 |
| BLE SoC with integrated PA | Imported BLE SoC with PA | 2.5 – 4.5 | Unisoc-class BLE SoC (family level) | 1.00 – 2.20 |
| ECG / bio-impedance AFE | Imported medical AFE | 3.0 – 8.0 | Chipsea medical AFE families (qualification-dependent) | 1.20 – 3.50 |
| Low-power host MCU | STM32L-class Cortex-M | 1.8 – 3.8 | AT32F421C8T7 (Artery) | 0.40 – 1.00 |
| Low quiescent LDO | Imported 1 uA-class LDO | 0.15 – 0.45 | COS29152WU (Cosine) | 0.04 – 0.14 |
| Charging-contact ESD | Imported low-capacitance TVS | 0.15 – 0.40 | TS2502PN (Titan Micro) | 0.03 – 0.12 |
Where the cost difference actually comes from
On a patch, the radio is 20 to 30 percent of the electronics cost, so a domestic BLE SoC genuinely moves unit economics, which is why domestic designs dominate consumer-grade wearables. On an implant, the radio is a small fraction of a device whose cost is dominated by packaging, batteries, sterilisation and clinical evidence. Substituting the transceiver there saves almost nothing and risks a regulatory submission. The rational split is therefore: domestic silicon where the product is disposable or consumer-adjacent; imported silicon where the product is implanted or has an approved predicate. A second cost line people forget is firmware and stack licensing: an imported part often arrives with a qualified BLE stack and security library, and rebuilding that around a domestic SoC has an engineering cost that dwarfs the per-unit saving at low volumes.
Validation flow before you commit
Measure average current over a full clinical use cycle, not in a sleep-mode demo: include the AFE, the DC-DC converter and the radio wake overhead. Then run an EMC pre-scan per IEC 60601-1-2 with the device in its real enclosure, because a patch antenna against skin behaves nothing like a bare board. Test coexistence by placing the device next to a hospital access point and an electrosurgery generator and checking for sample loss over hours rather than minutes. Finally, confirm the supplier’s longevity and change-notification policy in writing; in medical products the second-source question is answered by documentation, not by a sample kit.
Frequently asked questions
Can a BLE SoC be used inside an implant?
In principle the band and the power budget are the obstacles rather than the protocol. MICS exists precisely because 2.4 GHz through tissue and at implant power levels is a poor trade; most implants stay in the MICS band and add BLE only for external accessories.
Is domestic silicon acceptable for a regulated medical device?
It can be, but the evidence burden is on you. What regulators ask for is your verification data, not the silicon’s country of origin; a domestic part with a documented change-notification policy is easier to defend than an imported part with none.
Why is average current more important than sleep current?
Because the design rarely sleeps. Sensor sampling, DC-DC quiescent draw and radio wake overhead accumulate; a datasheet sleep figure can be two orders of magnitude below the real daily average.
Do patches need their own security element?
If the SoC provides secure key storage and authenticated updates, an external element is usually unnecessary. If it does not, adding one costs board area and current, so check early.
About us & trademarks: JTDZ Tech (Shenzhen Jingteng Electronics Technology Co., Ltd.) is an independent distributor of electronic components. We are not an authorised distributor or representative of the imported manufacturers named in this article. Manufacturer, brand and part names are used for identification purposes only; all trademarks remain the property of their respective owners. Cross-reference information in this article is based on publicly available sources, is subject to change, and is not a substitute for your own engineering validation. Always verify specifications against the manufacturer's datasheet before design-in or purchase.
