Characterization of a Bio/Gas Sensor
By Charles Guthy and Angel Madera
16.541 Introduction to Biosensors
5/9/2007
- Overview
This is the new Bio/Gas Sensor Characterizer. Used for the characterization and qualification of resistive gas sensors, it measures the changing electrical resistance of a gas sensor by passing different gases and concentrations and mixtures thereof and measuring the resultant voltage changes across the sensor.
The design of the Characterizer is quite straightforward, consisting of an airtight chamber for gas, necessary pipes and gauges, and test circuitry. The user first puts the gas sensor, along with its wiring, into the glass tube. Being sure that the wiring has been brought outside the tube, the user then puts the rubber stopper on the tube’s open end. The user attaches two gas canisters to the system, turns on a DC power supply (not included), sets it to the desired voltage (generally +5VDC, although other voltages can be used), and turns each canister’s spigot.
The test circuitry is also very simple in design. It is made of a Wheatstone Bridge (for proper measurement of changing voltage) and an instrumentation amplifier op-amp topology for data collection and analysis. When gas flows over a test sensor, the Wheatstone Bridge detects the voltage change and sends it to the instrumentation amplifier. The amplifier topology, in turn, magnifies the signal from the Bridge and outputs a signal at least 3 times stronger than the input signal voltage, limited only by the supply voltage. The factor that the instrumentation amplifier multiplies in the input signal by can be changed to the user’s preference by adjusting R8.
- Product Pictures
- AutoCAD Drawings
- Operating Circuit
- Specification Sheet
Abs. Max. Supply Voltage: ±22VDC (±5VDC or ±12VDC recommended)
Operating Temperature Range: 0ºC to +70ºC
Gain Range: 2 to 10 V/V
Measurable Sensor Resistance Range: 0 to 400K ohms
Slew Rate: 400000 V/s
Noise: 0.55 uVp-p
- Graphs
This graph shows the results of the first experiment with the test circuit to prove its potential worth. As you can see, the output voltage (Gain x v2-v2) has a curve that eventually falls to ground. This experiment was done with a 220 kilo-ohm resistor within the Wheatstone Bridge connected in parallel with a decade box (used to simulate a resistive gas sensor)
We later did more testing, removing the 220 kilo-ohm resistor and placing a 5 kilo-ohm resistor in series with a 20 kilo-ohm potentiometer in between the outputs of the first two op-amps (together known as a gain resistor). We performed two tests, one with the pot set all the way counter clockwise (Rg=26.1 kilo-ohms) and the other set all the way clockwise (Rg=5.001 kilo-ohms). The curves generated were quite interesting. Both were much sharper than the curve above, with the clockwise curve beginning a steep fall when the decade box was set at 100 kilo-ohms and stopping at ground at around 400 kilo-ohms, and the counter-clockwise curve beginning an even steeper fall at 200 kilo-ohms and stopping ar about 400 kilo-ohms. It can be inferred that, had the op-amps’ power supply been +/- 15VDC instead of simply +VDC, we would have seen a much more complete curve that dropped sharply from a little less that +15VDC to a little more than -15VDC.
- Possibilities for Expansion
The Bio/Gas Sensor Characterizer provides the user a limited ability to expand or otherwise change the setup to suit test requirements, localized entirely within the test circuit. Signal gain may be adjusted (3.6x to 10x) by setting the 20 kilo-ohm pot at R8. In the future, when hardy (projected lifetime of over 10 million cycles), adequately sized potentiometers can be located, the user will be able to adjust the voltage across both nodes of the Wheatstone Bridge.
In addition, op-amp buffers and passive, low-pass filters (with a cutoff frequency of at least 30Hz) may be installed to improve signal quality. A CERDIP op-amp, such as Analog Devices’ AD620, or Texas Instruments’ INA126, specifically designed as an instrumentation amplifier, may also be obtained and installed. This would improve performance, reduce the amount of necessary wiring, and effectively “ruggedize” the circuit. It is our group’s determination to bring this already fruitful project to a successful conclusion.
- Appendix
Original Schematic
Raw Data from the First Experiment with the Test Circuit (220 kilo-ohms in parallel with the decade box)
Resistance / Vin (VDC) / Vo (VDC)10000 / 0.123 / 4.350
20000 / 0.232 / 4.350
30000 / 0.332 / 4.350
40000 / 0.424 / 4.340
50000 / 0.510 / 4.230
55000 / 0.549 / 4.150
60000 / 0.587 / 4.060
65000 / 0.624 / 3.980
70000 / 0.657 / 3.900
75000 / 0.692 / 3.840
80000 / 0.726 / 3.780
85000 / 0.758 / 3.710
90000 / 0.789 / 3.650
95000 / 0.818 / 3.590
100000 / 0.849 / 3.530
105000 / 0.876 / 3.470
110000 / 0.904 / 3.410
115000 / 0.930 / 3.360
120000 / 0.954 / 3.310
125000 / 0.979 / 3.250
130000 / 1.003 / 3.210
135000 / 1.026 / 3.160
140000 / 1.048 / 3.110
145000 / 1.068 / 3.070
150000 / 1.088 / 3.020
155000 / 1.112 / 2.980
160000 / 1.131 / 2.940
165000 / 1.152 / 2.900
170000 / 1.168 / 2.860
175000 / 1.188 / 2.830
180000 / 1.207 / 2.800
185000 / 1.224 / 2.750
190000 / 1.238 / 2.720
195000 / 1.254 / 2.680
200000 / 1.267 / 2.650
210000 / 1.296 / 2.590
220000 / 1.329 / 2.530
230000 / 1.357 / 2.480
240000 / 1.383 / 2.410
250000 / 1.408 / 2.360
260000 / 1.432 / 2.310
270000 / 1.454 / 2.260
280000 / 1.482 / 2.210
290000 / 1.501 / 2.170
300000 / 1.522 / 2.120
310000 / 1.537 / 2.080
320000 / 1.557 / 2.040
330000 / 1.572 / 1.990
340000 / 1.587 / 1.960
350000 / 1.608 / 1.920
360000 / 1.625 / 1.890
370000 / 1.641 / 1.860
380000 / 1.654 / 1.820
390000 / 1.663 / 1.792
400000 / 1.683 / 1.766
410000 / 1.698 / 1.738
420000 / 1.712 / 1.711
430000 / 1.724 / 1.682
440000 / 1.735 / 1.654
450000 / 1.747 / 1.629
460000 / 1.759 / 1.605
470000 / 1.772 / 1.583
480000 / 1.785 / 1.560
490000 / 1.794 / 1.539
500000 / 1.810 / 1.514
List of Materials
1