Wearable Respiratory FSR: Fix the Mechanical Load Path Before Calibration

A wearable respiratory FSR should not be calibrated until the mechanical load path is repeatable. Define how belt expansion reaches the holder or actuator, where that part contacts the active area, how much preload remains at rest, and how the tail is restrained. Then compare the same mounted sample across the intended body positions and light-movement states. Otherwise, a baseline shift caused by belt tension or contact geometry can look like a sensor, circuit or algorithm fault.

An original wearable respiratory-sensor paper submitted on August 6, 2026 describes an FSR embedded in an abdominal belt, a mechanical holder that transfers abdominal expansion, and a low-power acquisition system. Its tests also identify posture effects and motion-induced baseline shifts during light movement. The paper does not specify a Baoshengda component. The buyer-level lesson is that the force-sensing element cannot be reviewed separately from the belt, holder, preload and movement states.

A custom FSR pressure sensor can be specified by active-area shape, outline, tail, connector and electrical response range. The wearable-device team still owns the belt, holder, electronics, signal processing, comfort, skin-contact materials, risk management and any medical validation.

Draw the complete belt-to-sensor load path

Start with a section view through the belt. Mark the outer textile, any elastic layer, the holder or actuator, the FSR active area, backing support and the body-facing layer. The drawing should show which part moves, which part stays supported and where the force closes through the assembly.

A flat force applied at the center of a loose FSR is not the same as belt expansion applied through a curved holder. If the actuator touches only the edge of the active area, tilts during breathing or slides as the wearer moves, the effective contact area changes. The output may change even when the electronics remain stable.

Control these dimensions before requesting a sensor sample:

• actuator footprint relative to the active area;
• holder radius and allowable tilt;
• backing stiffness beneath the sensor;
• belt stretch and adjustment range;
• adhesive or pocket geometry that locates the sensor;
• tail exit direction and the first restrained point.

Do not ask the sensor supplier to infer these details from a belt photograph. A simple section view and an actuator contact outline are more useful than a finished-product rendering.

Set preload before setting the electrical threshold

Preload keeps the actuator in contact at rest, but too much preload can consume useful response range or make fit variation dominate the signal. Too little preload can allow intermittent contact, producing steps or flat sections that resemble an electrical dropout.

Specify a mechanical rest condition first. For example, record belt adjustment, holder position and the force or compression state at a defined reference fit. Then record the readout voltage, divider resistor, supply voltage, sample rate and filtering used for that measurement. A threshold without those conditions is not portable between fixtures or electronics.

Use a preload window rather than one ideal number. The lower edge should maintain contact in the loosest accepted fit. The upper edge should leave enough response range for the intended expansion while avoiding excessive local pressure. The device team must determine the acceptable comfort and safety limits; the FSR supplier can help match sensor geometry and sample response to the stated mechanical window.

Separate mechanical symptoms from electrical symptoms

The fastest troubleshooting starts by changing one layer at a time.

This table does not diagnose a medical condition. It separates assembly variables so the engineering team can decide whether to change the holder, belt, FSR geometry, cable restraint, readout circuit or algorithm.

Restrain the tail without stiffening the sensing zone

A long flexible tail or cable can pull on the sensor when the wearer turns, bends or removes the belt. If the restraint point is too far away, cable movement can rotate the holder or peel the sensor. If it is too close to the active area, a hard clamp or thick adhesive can change local stiffness and create another load path.

Show the tail in the worn and removed states. Define the minimum bend radius, first restraint point, connector orientation and service loop. Keep the connector away from repeated body pressure unless the enclosure and comfort design specifically account for it. Run the motion test with the final cable, not a bench lead that will disappear before production.

For a flexible printed tail, specify contact side, pitch, exposed-contact length and stiffener direction. For a wired lead, specify wire length, strain relief, connector family and mating direction. The sensor drawing and belt drawing should use the same tail-exit datum.

Approve the mounted sample across real use states

A loose-sensor press test confirms only a limited part of the design. Build a representative belt fixture with the intended holder, backing, sensor location and cable restraint. Mark the adjustment position so another technician can reproduce it.

Test at least these states when they are relevant to the product:

• the minimum, nominal and maximum accepted belt adjustment;
• standing, sitting and lying positions;
• normal breathing, deliberate deeper breathing and a no-motion reference;
• light turning, reaching or walking if the device is expected to see those movements;
• removal and refastening by different users or technicians;
• the planned cleaning cycle and environmental conditioning;
• cable flexing and connector mating after the belt has been worn.

Keep raw and processed traces. A filtered graph alone can hide intermittent contact or a mechanical step. Link each trace to the sensor lot, belt revision, holder revision, electronics revision and test state. If the holder or belt changes, repeat the mechanical approval instead of carrying forward the old calibration automatically.

Define the supplier boundary in the drawing package

Baoshengda can review the FSR outline, active-area geometry, printed sensing structure, tail or wired lead, connector option and sample workmanship against the requirements provided. The wearable-device developer remains responsible for the mechanical holder, belt fit, enclosure, electronics, calibration method, signal interpretation, user comfort, skin-contact material decisions, cleaning validation, regulatory pathway and final product performance.

Do not label the component drawing with an unverified medical claim or ask the FSR sample to prove the finished system. State the measurable component conditions instead: contact footprint, force or compression window, response limits, repeatability method, tail routing and mounted acceptance test.

Send an RFQ that can be built and tested

Use the Request Quote page to send a package that includes:

• sensor outline, active-area location and preferred sensing-zone count;
• belt section view and the holder or actuator contact footprint;
• expected force or compression range and the rest-preload window;
• backing material, adhesive or pocket method, and location datums;
• tail direction, cable length, connector, mating orientation and restraint point;
• supply voltage, divider or conditioning circuit, sampling method and threshold logic;
• required body positions, belt adjustments and movement states for sample review;
• cleaning agents, temperature, humidity and expected flex exposure;
• prototype quantity, production estimate and acceptance evidence.

The useful starting question is not whether the FSR can detect force in isolation. It is whether the complete belt-to-sensor load path produces a repeatable input under the positions, preload and movement states the finished device will actually see.