Font Size: a A A

Performance Improvement Of SnO2 Formaldehyde Gas Sensor

Posted on:2007-10-23Degree:MasterType:Thesis
Country:ChinaCandidate:H C BaoFull Text:PDF
GTID:2178360212457563Subject:Circuits and Systems
Abstract/Summary:PDF Full Text Request
In this paper, a kind of indirect-heating formaldehyde gas sensor based on SnO2 is introduced, and we try to improve the performance of this gas sensor, such as sensitivity, selectivity and so on.The fabricating processes of SnO2 gas sensors without and with different dopants were described in this paper. Their basic gas sensing characteristics were tested. The sensitivity reached its max value when heating voltage was 4V for the undoped SnO2 gas sensor, but it decreased to 3.5V for some doped SnO2 sensors. The sensitivities of some doped gas sensors were much increased. Based on the test, the sequence of influence of different dopants on the sensitivity of the sensors is Pd>Zr>Ti>Sb>Cu>Ag>Mn. Especially the sensitivities of sensors doped with Pd and Zr to formaldehyde are above 50, which is good result.The selectivity of the sensor could not be improved by doping metals in our experiments. A three-level B-P neural network that can alert learning rate and momentum coefficient automatically was taken to do it. There were 6 nods in input layer, 8 in hide layer and 2 in output layer. A sensor array with six different sensors was used to the mixed air of formaldehyde and ethanol, and got hundreds of data which were the input of the network. And the network learned by itself based on the data circularly until the error of output was below that you set before. Then the network was ready to distinguish quantificational formaldehyde and ethanol from mixed air. When the concentration of formaldehyde in mixed air was 50ppm, 100ppm and 200ppm, respectively, the predicting outputs of this network were 57.1ppm, 111.6ppm and 214.2ppm, respectively. The predicting errors of this network were 13.4%, 11.6% and 7.1%, respectively.Simple IC modules such as 51 SCM and ADC0809 were used to design and print wireless transporting circuit, which was based on voltage output signal of formaldehyde sensing sensors. The advantages of this circuit were simple-structure, low-waste, far distance-transmission and easy-to-operation. The receiving program based on VB language could dynamically indicated the voltage signal and drew the curve of voltage signal in a long period. The whole processes have been fulfilled include signal collection, signal processing, data storage, data sending and receiving, and display.
Keywords/Search Tags:SnO2 Gas Sensor, Formaldehyde, Adulterant, B-P Neural Network, Wireless Sending and Receiving Circuit
PDF Full Text Request
Related items