Electronic Nose (E-Nose) Gas Sensor Array Demo

An electronic nose works like the biological sense of smell: instead of one selective sensor, an array of broadly-tuned sensors produces a response pattern, and the pattern identifies the odor. Students usually meet this idea only in papers. This project makes it a bench demo. Four MQ-series gas sensors (MQ-2, MQ-3, MQ-7, MQ-135) sit on a board inside a sealed acrylic chamber; an Arduino reads each sensor continuously and an LCD shows all four responses as live bar graphs. Injecting a sample with the syringe — fruit scent, alcohol vapor, clean air — makes the bars jump in visibly different combinations, and the pattern-recording procedure lets the student build a small reference library of odor fingerprints.

1. The sensor array warms up inside the sealed chamber until the LCD baselines stabilize on clean air.
2. The student records the four baseline bar levels as the reference "clean air" pattern.
3. A sample (e.g. fruit scent from the vial) is injected through the port with the syringe.
4. Each MQ sensor responds according to its sensitivity profile, and the LCD bar graphs jump in a sample-specific combination.
5. The student records the four readings in the pattern table; repeating with different samples builds the odor fingerprint library.
6. The procedure explains how a nearest-pattern comparison on the normalized readings identifies an unknown sample, demonstrated by the student.

Tech stack:

• MQ-2, MQ-3, MQ-7, MQ-135 gas sensors
• Sealed acrylic test chamber
• Arduino controller board
• Character LCD with bar-graph firmware
• Syringe sample injector
• Sample vials (fruit scent and blanks)
• Perforated sensor mounting board
• Pattern recording tables

• [Four-sensor MQ array] MQ-2, MQ-3, MQ-7 and MQ-135 sensors give four overlapping sensitivities, the minimum array for visibly distinct odor patterns.
• [Sealed acrylic chamber] A transparent screw-sealed chamber holds the sensor board and the odor sample, with an injection port for the syringe.
• [Live response bar graphs] The LCD shows each sensor's response as a real-time bar graph, so pattern differences are visible the instant a sample is injected.
• [Syringe sampling] A syringe injects controlled sample volumes through the chamber port, making the demonstration repeatable.
• [Pattern-recording procedure] A documented procedure for recording baseline, injecting samples, and building a reference table of odor fingerprints.
• [Labeled sensor board] Every sensor is labeled on masking tape (MQ-2/MQ-3/MQ-7/MQ-135), traceable for the viva.
• [Complete viva kit] Report covering gas-sensor physics, e-nose principles and pattern recognition basics, plus PPT and viva Q&A.

• Acrylic e-nose chamber with sensor board
• Four MQ gas sensors, labeled
• Arduino controller with readout firmware
• LCD display module
• Syringe and sample vials
• Complete wiring
• Pattern-recording procedure with blank tables
• Project report PDF (gas sensors, e-nose theory, pattern recognition)
• PPT presentation for final review
• Viva Q&A preparation document

• MQ sensors are indicative, not calibrated analytical instruments; readings are relative response levels, not ppm concentrations.
• Sensors need a warm-up/stabilization period before readings are meaningful; the procedure enforces this.
• Sensor responses drift with temperature, humidity and age; the report treats this as a documented error source.
• Odor "identification" is a nearest-pattern demo on the student's own recorded library, not a validated classifier — no accuracy figure is claimed.
• Only safe household odor samples are used; no toxic or flammable gas testing is included or permitted.