You've selected the right sensor. You've read the datasheets, compared sensitivity values, and reviewed cross-sensitivity tables. Now you need to turn that sensor into a working, field-deployable, certified product.

Sensor selection and gas detection device development are not the same problem. Sensor selection is one decision. Device development is twenty decisions that constrain each other. Your power architecture shapes your alarm algorithm. Your MCU choice limits your communication options. Your enclosure design changes sensor response time.

We've covered sensor selection criteria in our gas sensor selection guide and electrochemical sensor guide. This article moves one level up: the design decisions that follow sensor selection, and why getting them in the right order matters.

Why Gas Detection Device Development Is Harder Than It Looks

The description makes it sound simple: place the sensor, read the ADC, trigger an alarm when the threshold is crossed.

The actual picture is different. Electrochemical sensors produce current signals in the nanoampere range. A typical CO sensor has a sensitivity of 50–90 nA/ppm. At 10 ppm, the output current is in the 500–900 nA range. Converting that to a low-noise voltage requires a potentiostat circuit. Potentiostat design is an analog engineering problem in its own right.

PCB layout compounds the challenge. Leakage currents cause measurement errors when your signal is at the nanoampere scale. Temperature variation and humidity shift sensor output. Without compensation, the device produces false alarms in the field, or worse, fails to alarm when it should.

The most dangerous scenario in gas detection is not a false alarm. It's a device that fails to alarm when it needs to. Recognizing this at the design stage is what separates products from prototypes.

Add certification requirements, environmental conditions, and product lifetime, and development requires managing several parameters simultaneously:

Analog circuit design
Embedded firmware
Mechanical design
Performance testing and verification

Power Architecture: Battery, Mains, or Hybrid?

Define power architecture in the beginning of the project. Changing it later is the most expensive redesign you can make in terms of time and development cost. Mains/adapter powered, battery-powered, or hybrid, the choice determines every decision that follows.

Battery-powered design: Everything is constrained by the power budget. The difference in battery life between a continuously sampling device and a periodically sampling device can range from months to years. That delta is rarely driven by the sensor. It's driven by MCU selection, firmware design, and circuit efficiency.

Mains/adapter powered design: Power constraint disappears, but a new problem appears. When external power fails, the backup battery transition must be seamless. Alarms cannot reset. Communications cannot drop. Devices that skip testing this transition in the prototype phase encounter this in a field stage. When backup power activates, the device may also need to reduce sampling frequency to extend backup runtime.

Shared concern across both architectures: Mains power introduces 50 Hz harmonic noise. Switching adapters and on-board DC/DC converters introduce high-frequency switching noise. In both cases, sensitive analog circuits require adequate isolation from these noise sources. Your power architecture choice determines which noise sources you're managing.

Microcontroller Selection

MCU selection in gas detection device development is frequently done with the wrong criteria. The right question is not "which MCU is most capable?" It's: "is this MCU at the right trade-off point for this product?"

The answer depends on what you're building.

Residential CO detector (EN 50291): Cost pressure is high. An 8-bit or low-cost 32-bit MCU with sufficient ADC resolution and minimum flash/RAM is appropriate. Certification is required regardless of architecture, so cost-efficiency balance is the determining factor.

Industrial gas detection device: Requirements shift to communication interfaces (4–20 mA, RS485 Modbus RTU), memory capacity, and hardware safety features. Reliability and standard compliance take priority over unit cost.

Battery-powered design (either category): MCU sleep-mode current consumption becomes the binding constraint. The difference between 1 µA and 10 µA in sleep mode has a direct, calculable impact on annual battery life.

ADC performance matters in every case. While datasheet resolution specs reflect ideal conditions, in real scenarios, particularly after waking from low-power mode, ADC settling time is non-negligible. Firmware that doesn't account for settling time generates a read error at the start of every measurement cycle. This failure mode is difficult to find in the lab without deliberate test scenarios. In the field, identifying the root cause is significantly harder.

Additional criteria to include in MCU selection:

EEPROM or flash emulation for storing zero and span calibration data
Watchdog and brown-out detection for hardware safety requirements
Compliance with applicable certification standards

MCU selection is a design constraint, not a starting point. The wrong MCU creates redesign cost in later phases, particularly during low-power optimization, calibration data management, or certification.

Communication Interface

Gas detection devices communicate in different ways depending on the application: 4–20 mA analog output, RS485 Modbus RTU, digital communications or a combination. Whatever the requirement, it must be defined before MCU selection — not after. An MCU with insufficient UART peripherals or memory cannot support a Modbus RTU stack without a hardware revision. Defining communication requirements late is one of the most common causes of hardware revisions due to changing the MCU.

The choice also has firmware implications. Modbus RTU requires a protocol stack running alongside measurement and alarm logic. On resource-constrained MCUs, this creates scheduling and memory pressure that must be accounted for in the firmware architecture from the start.

For devices with 4–20 mA output, the signal conditioning circuit must be designed alongside the analog front-end.

Alarm Strategy: TWA or Instantaneous Threshold?

Alarm algorithm design is both a technical and a product philosophy decision. Model what each approach means in the field before deploying.

Instantaneous threshold: Simple and fast. Alarm when concentration exceeds a set point. But short-term, non-hazardous concentration spikes trigger alarms too. A garage door opening, a kitchen burner igniting, these can briefly exceed thresholds. Repeated false alarms erode user trust and create alarm fatigue: users stop responding to alarms. For safety devices, that's an unacceptable outcome.

Time-weighted average (TWA): Solves the false alarm problem. The human body's response to toxic gases like CO depends on cumulative dose, not instantaneous concentration. TWA algorithms model this relationship: low concentration over long duration and high concentration over short duration are evaluated against different threshold-time curves. False alarms decrease; sensitivity to real hazards is preserved.

TWA creates its own engineering requirements. Window size, update frequency, and concentration-time curve calibration each affect algorithm accuracy and computational load. On a battery-powered device, that load must be balanced against the power budget.

Which approach is correct depends on the application. Residential CO alarms use time-based alarm algorithms because user experience is critical. Industrial environments may prioritize fast response to high instantaneous concentrations. Many products combine both: TWA at lower concentrations, instantaneous threshold above a defined ceiling, for example, 300 ppm.

One direct connection: EN 50291 tests alarm behavior at specific concentration-time combinations, 50 ppm over 60–90 minutes, and 300 ppm within 3 minutes. Designing the algorithm without referencing these test conditions creates certification surprises.

Gas Detection Device Development: Certification

Certification is where the decisions made during design are systematically evaluated. For residential CO detectors, EN 50291 applies. For parking garage applications, BS EN 50545-1 is the reference standard. Both define device performance, alarm behavior, and environmental robustness requirements.

Core requirements these standards impose:

Alarm response at defined concentration-time combinations (EN 50291: alarm within 60–90 minutes at 50 ppm CO; within 3 minutes at 300 ppm CO etc)
Performance retention across temperature and humidity range (typical: −10°C to +40°C)
Performance retention under electromagnetic disturbances and electrical safety compliance under Low Voltage Directive (LVD) requirements
Fault indication for sensor failure or circuit malfunction
End-of-life indication

Certification should not be treated as a final checkpoint. Doing so typically leads to unnecessary redesign cycles, increased cost, and schedule delays. Integrating certification requirements into design decisions from the start reduces these costs. Alarm algorithm design and analog front-end choices are the highest-leverage points.

Production and Field Realities

The prototype works. Lab measurements are accurate. Alarms are triggered at expected conditions. Gas detection device development enters a new phase — field conditions, which differs from lab conditions.

Production variation: Two electrochemical sensors from the same batch produce different output values at identical concentrations. Closing that gap requires individual device calibration. How calibration integrates into production, reference gas handling, test duration, writing calibration data to the device, directly affects production cost and line throughput. Designing the calibration procedure at the prototype phase prevents a costly problem in production.

Field drift: Electrochemical sensors drift over time. This is predictable and manageable, but the management plan must exist before the product ships. How does the device notify users that calibration is needed? How is periodic recalibration performed? Where is calibration data stored? These questions have hardware and firmware implications that must be resolved in the design phase.

Silent failure: Sensors degrade before they fail completely. During degradation, the device appears functional. It displays readings. Alarms don't trigger. But measurement accuracy is declining. A device without a self-test and sensor state of health monitoring mechanism enters the worst field scenario: the user trusts it, the device doesn't measure.

Every Decision Is Connected

The core difficulty in gas detection device development is that none of these decisions are independent.

Power architecture determines alarm algorithm behavior and sampling frequency.
Certification requirements shape alarm algorithm design.
MCU selection constrains communication options.
Calibration strategy determines production and maintenance cost.
Enclosure design affects sensor response time and certification test results.

Making these decisions in the right sequence, with full visibility into their downstream effects, minimizes redesign cost. A decision that looks cheap in the prototype phase can cost significantly more when it surfaces during certification, or after field deployment.

A good gas detection device is the result of dozens of consistently aligned design decisions. End-to-end sensor-based device development requires experience managing these interdependencies. The decisions that matter most are the ones made before the first prototype.

Design Checklist

✔ Power architecture (battery/mains/hybrid) defined at project start ✔ Backup battery switchover scenario tested in prototype ✔ MCU sleep-mode current compatible with power budget ✔ Calibration data storage (EEPROM/flash) added to the embedded firmware ✔ Alarm algorithm (TWA/instantaneous) mapped to target standard requirements ✔ Production calibration procedure designed before prototype completion ✔ Self-test and sensor state of health monitoring included in firmware development ✔ Certification requirements (EN 50291 / BS EN 50545-1) reflected in design decisions from project start

If you need technical support for gas detection device development, contact us. HEFA Teknoloji provides end-to-end R&D services including sensor selection, electronics design, and embedded software development.

References

SGX Sensortech. Electrochemical Sensors Application Note 2: Design of Electronics for Electrochemical Gas Sensors.
BS EN 50291-1, Electrical apparatus for the detection of carbon monoxide in domestic premises.
BS EN 50545-1, Electrical apparatus for the detection and measurement of toxic and combustible gases in car parks and tunnels.