Team:Nottingham/Hardware

Hardware:


Electronic Nose

Optimist: "The glass is half full."

Pessimist: "The glass is half empty."

Engineer: "The glass is twice as big as it needs to be."

Hardware:
Electronic Nose

Engineering Aim

Current detection methods for botulinum toxin are expensive, time-consuming and conducted in specialized labs by trained scientists only. Consequently, tests for the toxin are neither widely nor readily available. We have set ourselves an aim to overcome this barrier by developing a new device, an electronic nose, for the detection of acetone produced by our acetone-producing reporter strain.

Introduction to the Electronic Nose

We have fulfilled our engineering aim with the production of an automated and handheld device, an electronic nose, which determines the acetone concentrations in the sample between 50 and 500 PPM (0.86 and 86 mM). The device measures the acetone in the food sample under investigation based on the concentration found in the atmosphere above it, using an automated syringe system. The atmospheric sample is then analysed by a gas sensor, further referred to as Figaro TGS822, which provides a voltage reading. Finally, this reading is converted into acetone concentration present in the food sample. This concentration is then shown on an LCD. In addition to this, we have included two LEDs to indicate when the concentration of acetone in the sample’s atmosphere is in excess of the sensor’s limit.




The casing of the device has been designed using Fusion 360 software and 3D printed using silver PLA (polylactic acid), which is a cost-effective and easily accessible material. In addition to the syringe system, the casing contains a battery pack (14.54 V / 3500 mAh) providing the necessary power for all electronic components. For more detailed specifications of the components used, see the “List of Components” section below.


The programme for the device has been fully written using the Arduino IDE software and coded onto the Arduino Uno board. .

Design and Creation Process

The creation process of our device consisted of an intertwined development of the hardware, software and the casing.


Our first task was to design an electronic circuit capable of converting the voltage readings of the gas sensor into concentrations and displaying the results. The preliminary circuit, shown in Figure 1, consisted of:

  • an Arduino Uno control board
  • an LED, representing the envisaged display
  • a battery, ensuring that the detector will be transportable
  • some resistors
  • a gas sensor, Figaro TGS822

All provided by Rapid Electronics.


Figure 1. Preliminary circuit.

The next step was to test our setup and to determine its measurement range and accuracy. To carry out the test, we prepared a series of samples containing different acetone concentrations. We then determined the voltage readings representing the atmosphere above our samples (Figures 2 & 3). Subsequently, we converted those readings into concentrations using an equation created by our modelling team.


Figure 2. Acetone concentration test

Figure 3. Acetone concentration test

Results:


Known acetone concentration/mM

0

0.5

1

2

5

10

20

50

100

Average of 3 readings/mV

110

195

135

127

427

646

664

805

896

Average of 3 readings/mM

0.5

2.20

1.01

0.85

7.89

18.8

20.4

43.2

109.5

From the results above we can see that the sensor can detect acetone. As detection limits of Figaro TGS822 lie between 0.82 mM and 82 mM, the values for 0.5 mM and 100 mM have been altered. Similarly, we can see that the reading for concentration of 2 mM does not follow the observed trend. This is due to a suspected pipetting error. The values read by the sensor are close to the known concentrations, however, they are not exact due to the Figaro TGS822 exposure to humidity and temperature.


The equation below was used to change mV to mM in the code:


x=(yb-cb/a-y+c)^1/n


Where:

x - concentration of acetone in mM

y - voltage reading in mV

a = 859

b = 18.5

c = 90.2

n = 1.2


On Figure 4 we can see the calibration curve of voltage/mV against acetone concentration/mM used to generate the equation for the mV to mM conversion.


Figure 4. Calibration curve of voltage/mV against acetone concentration/mM

Furthermore, the experiment showed us that we needed to change the design of the nose. During our measurements we realised that the readings are strongly dependent on the distance between the sensor and the surface of the liquid, in addition to the fact that air movement will alter the atmosphere above the sample too. It was also found that the results change depending on the temperature and humidity of the environment. As a consequence, we needed to place the sensor in an enclosed space to guarantee a controlled environment during the measurement process.


Our initial solution was simple; a sample would be extracted from a headspace of the tested food via a syringe and subsequently transported to the Figaro TGS822 sensor sealed in a casing.


Although our first design was functional and fulfilled all requirements that were needed to get an accurate reading, it lacked practicality since it was neither transportable nor user-friendly. Thus, the device was not applicable in either challenge testing or testing packaged food. As a consequence, we had to improve our design fulfilling the above mentioned condition of a controlled measuring environment, being handheld and battery-powered, and finally easy to use.


Figure 5. The initial design of the casing

Figure 6. The initial schematic of the circuit

The initial design of the casing and circuit (Figure 5 and 6 respectively).


However, this initial design did not fulfill the aforementioned requirement to guarantee a controlled environment for the measurement of the gaseous acetone concentration. Thus, we had to change the design of the casing significantly to include an automated syringe system (see the next section for more details) to make the testing process more accurate and less time-consuming. We have, therefore, extended the handle and made the bottom of the device wider and longer to accommodate the syringe system. Furthermore, we had to extend the battery box to house a more powerful battery pack which was now also supplying power to the syringe system. Additionally, we have included a removable lid at the back of the design, which would allow access to the hardware if anything breaks or needs replacement (see Figures 7,8,9,10).


Figure 7. The back view of the electronic nose

Figure 8. The top view of the device. The battery box and the designated space for the on/off button visible.

Figure 9. The bottom view of the device. The hole through which the syringe goes out visible.

Figure 10.The frontal view of the device.

The complete circuit design with the stepper motor included and the complete casing (Figure 11 and 12 respectively).


Figure 11. The complete device schematic with the stepper motor and 2 hall effect sensors included.

Figure 12. The complete electronic nose casing design with descriptions of how each space in the design was used.

The Automated Syringe

Our automated syringe system has been designed to increase the accuracy of the acetone readings, by providing an enclosed space for the gas sensor to reduce the effects of temperature, humidity and air movement on the measurement process, as well as full self-sufficiency for the device.


The three main parts of the assembly have been 3D printed. The dimensions and the final assembly product can be seen on Figure 13 and 14. Our design has been created to accommodate a 10 ml syringe.


Figure 13. The frontal view of the syringe pump with dimensions included (all dimensions have been provided in mm).

Figure 14. The top view of the syringe pump with dimensions included (all dimensions have been provided in mm).

The system is powered by a stepper motor, which is attached via a motor coupler to the fully threaded rod. As the stepper turns, the middle part of the assembly moves, causing the plunger of the syringe to move up causing the uptake of a sample, and down to expunge the sample when the measurement is finished (see Figure 15).


Figure 15. The back view of the syringe pump.

The pump was designed to give protection at each end to limit the movement of the middle part. This has been achieved using 2 hall effect sensors and 3 neodymium magnets. The magnets have been placed in the left-hand corner of the central (moving) part, while the hall effect sensors have been placed in adjacent holes on the front and back parts of the pump (one sensor in each part; directions described as if looking at Figure 13).


To assemble the device, we put the ball bearings in the distal holes of the central (moving) part. Through those bearings we placed the round rods, to be used as “tracks” for the syringe plunger and central (moving) part to be moved along. The lead screw was then placed in the central (moving) part, allowing the fully threaded rod, attached via a motor coupler, to turn when being twisted by the stepper motor. This twisting of the threaded rod provides the energy for the central (moving) part to be moved either forward and backwards (see Figure 17 for the full pump assembly).


Figure 16. The side view of the syringe pump.

Figure 17. The full assembly of the automated syringe, including a 10 ml syringe, the motor coupler and the stepper motor.

For all components used in the syringe assembly, see the “List of components” section.

The Device

All the dimensions on Figure 18 are in mm.

The depths of the sections:

Head: 75mm

Handle: 47mm

Bottom: 74mm

Figure 17. The full dimensions of the device (all dimensions have been provided in mm).

So, how will it work?

To power on the device, you need to press the start button, located on the top of the device. Once it is pressed, a message will appear on the LCD saying, “To start the measurement, insert the end of the syringe to the sample.”


A message saying, "to be confirmed" will appear. Hold the device steady in place and wait for the concentration, read by the sensor, to appear on the LCD. During the waiting time, the plunger will be automatically pulled up via the stepper motor mechanism, exposing the sensor to the atmospheric sample. This allows the gas to pass through the sensor and the reading to be taken by it.


If the concentration of acetone is above a safe threshold, the red LED will light up and a message will appear on the LCD screen signalling that excess acetone has been detected.


Once the reading is done, the user will be then instructed not to remove the device until told otherwise. This waiting time is required for the plunger to go down and pump out the collected gas. Another LCD message will inform the user when all the gas is removed and the device is ready to be switched off.


Since the device is powered by batteries to allow for handheld portability, it needs to be charged. To see the required charger, see the “List of components” section below.

Future Improvements

Even though the device we have designed is a prototype which can be used in the form it has been produced, there are a few issues which could be addressed to improve the accuracy of the device further.


The characteristics of the Figaro TGS822 are dependent on temperature and humidity. We have taken steps to reduce the effect of these conditions by placing the sensor in an enclosed space. However, to produce more accurate readings a temperature and humidity sensor should be added. The readings from these sensors could then be used to calibrate/adjust the results obtained by the gas sensor, making them more accurate. Alternatively, the Figaro TGS822 could be swapped for a less condition-affected sensor.


Another point for further evaluation lies in the syringe that we have used. Although the type we are using is sufficient for the device, it has a limited life, which means that if used, will require regular replacement. We have therefore contacted a specialist from marbleproductdesign.co.uk, Nigel Harrison, and asked whether he could suggest an alternative to a standard syringe. He has suggested using a pneumatic cylinder, as it is designed to operate on air for many cycles, providing a longer lifespan and a lower leak rate.

Adknowledgements

We would like to thank the following people for support in the creation of our device:

Edward Kujawinski - University of Nottingham’s EEE technician.

For helping us in the assembly of the components and advice on which ones to include.

Christian Klumpner - University of Nottingham’s Associate Professor in power electronics.

For providing us with the knowledge on the basics of electronic engineering as well as advice on the development of our design.

Alex Ottaway - University of Nottingham’s EEE technician.

For helping us in 3D printing the casing for our device.

Nigel Harrison - Marble product design engineer

For providing us with the knowledge on syringes for a greater understanding of the future of our device

Detection System

  • Arduino Uno: Orangepip Kona328 Arduino UNO Compatible Development Board
  • Sensor: Figaro TGS 822 Organic Solvent Vapors Sensor
  • LCD: Winstar WH1602B-YYH-JT 16x2 LCD Display Yellow/green LED Backlight
  • 2x LED
  • Fuse: ESKA 885011 Subminiature Quick response fuse 5.08mm Pitch 8.4x7.6mm 250mA
  • Motor/stepper/servo shield for the Arduino Uno: Adafruit 1438 Arduino motor/stepper/servo shield Kit
  • Fuse Holder: ESKA 866.002 fuse holder for subminiature fuses 5.6mm
  • Batteries: Ansmann 2447-3032-20 Li-Ion battery pack 14.54V / 3500mAh with PTC
  • Switches: SCI R13 BLACK 2 Pole SPST On-off Locking PI Mnt Push Sw / RVFM US-101A BLACK Push to make the switch
  • Charger: Ansmann 2000-0001-06 IPC30 Charger Li-Ion 1.2A 4 cell

Automated Syringe

  • 1x Stepper motor: RVFM MY3489A-2 Size 17 Hybrid Stepping Motor, 40mm Long
  • 3x 5mm neodymium magnets
  • 2x small hall effect sensors
  • 1x fully threaded rod
  • 2x M6 round rod
  • 2x M6 ball bearings
  • 1x M6 lead screw
  • 1x Stainless Steel Flexible Motor Coupler

References

All circuit diagrams were created using “Fritzing” software
The casing has been designed using Fusion360 software
Syringe inspiration taken from: “3D Printed Syringe Pump Rack” https://www.instructables.com/id/3D-Printed-Syringe-Pump-Rack/
The deeper understanding of the sensor taken from: Ketosense – An Arduino based ketosis detector, https://jenslabs.com/tag/tgs822/