Inspiration
The CheapStat is a budget friendly device developed at UC Santa Barbara several years ago. We aspired to build this circuit and modify the software for our purposes. We downloaded the zip file from the website of CheapStat. After unzipping the file, we got three folders: schematic, mcu and gui. The folder schematic contains EagleCAD files for schematic and board layout, and Gerber files for PCB manufacture. The folder mcu contains the firmware for the MCU and source files. The folder gui contains the java program for receiving results from the CheapStat. We decided to build one CheapStat and tested whether or not it works. So we sent the gerber files contained in the folder schematic to a PCB manufacturer. After waiting for one week, we got a CheapStat board from the PCB manufacturer.
Fixing bugs and improvement
Bug 1
After receiving the CheapStat board, we programmed the firmware into the mcu on the board, connected the CheapStat to a laptop through a pair of USB connectors and opened the java program contained in the folder gui. The CheapStat just had one input device, a 4-directional switch with center push(model. SKQUAAA010). There was a stick on the switch, shown in Figure 1.Pushing the stick up or down can move the current cursor displayed on the LCD screen up or down. You can choose current item by pushing the tick to the right and return to the previous menu list by pushing the stick to the left.
Figure 1. Photo of the 4-directional switch with center push(model. SKQUAAA010).
When we connected the CheapStat to the laptap, the LCD screen showed ”CheapStat” on the first line. Then three profiles were showed on the displayer. We wanted to move the cursor down to check more profiles, so we pushed the stick down, but the cursor didn’t move at all. We were afraid that the firmware for the mcu had some bugs. However, after careful checking, we found that it’s the PCB manufacturer’s fault since it made the switch in a upside down position when mounting it on the PCB board. We reinstalled the switch in normal position and the switch worked perfectly.
| parameters | value |
|---|---|
| start | -200mV |
| stop | 700mV |
| sampling step | 5mV |
| number of scans | 4 |
Table 1. Profile used in CV test.
Bug 2
We took a Cyclic voltammetry test using new CheapStat. The profile used in the test is shown in the table 1. We prepared a sample for test. We did a CV test for the sample on a commercial potentstat and did a CV test for the same sample on the CheapStat. We compared the results got from the commercial device and the CheapStat, which are shown in Fig2. As in Fig2, we found that the voltage at the maximum current shifts to the left as the increase of the cyclic voltammograms and the shift between adjacent cyclic voltammograms is about 20mV, but the result gotten from the the commercial device didn’t have shifts at all. We repeated the test several times, and the shifts still existed. We guessed that it maybe a software bug, because the sampling rate was 4mV per sample in our test and the shift is about 5 times the sampling interval. As expected, we checked the voltage datas generated by the client software and found it was not consistent with the voltages generated by the mcu during test, which caused the shift of the cyclic voltammogram. We tried to fix the client software, which was written in Java, but because of unknown reasons, we cannot rebuild the java program, so we chose to generate voltage datas in our data processing software and not use the voltage datas generated by the client software.
Expansion for Optical Detector
We also wanted a device capable of detecting the strength of biology luminescence.
Preparation and Design
We bought a Photomultiplier tube(Model. CR314) and a high-voltage power supplier(Model. CC238 ) from Hamamatsu Photonics (China) Co.,Ltd. The photomultiplier tube is useful for light detection of very week light, in which the absorption of a photon results in the emission of a electron. The photomultiplier has a good property that the output current has a linear relationship with the incident light flux. So if we measure the value of the output current, then we will get the value of the incident flux through some calculations. The mcu on the CheapStat had several empty ADC input channels and the ADC on the mcu can measure the voltage between 0 to 2V. So we chose to design a I-V converting circuit to convert the output current signal to a voltage signal in the range of 0-2V and use modified CheapStat to measure the voltage. The schematic was shown in Fig3. The schematic also include the power supply circuit which supplied +15 volt- age to the high-voltage power supplier and the circuit for adjusting the voltage supplied to the Photomultiplier tube.
Figure 2. The result got from a CV test by the CheapStat. You can see that the voltage at the maximum current shifts to the left as the increase of the cyclic voltammograms.
Figure 3. Schematic of I-V Converter.
The photocathode of the photomultiplier can not be exposed to daylight or another strong source of light. In order to make our device more convenient to use, we designed a container for the photomultiplier tube. The 3D perspective drawing of the container is shown in Fig4. The container consists of three part: photomultiplier room, sample room and aperture room. The photomultiplier tube is put into the photomultiplier room, the samples to be measured are put into the sample room and a circular aperture is put into the aperture room which has a role controlling how much light enters into the photomultiplier tube.
Figure 4. 3D perspective drawing of the container.
Testing
Initially, we doesn’t put anything into the device and measure the dark current, which is 8.6nA. Then we put the sample into the sample room. We can see that the current is increasing slowly as the passage of the time as in fig5, although there are some fluctuations, which were caused by the external noise. We can see that the total current when putting in the sample is approximately 60nA and the pure current caused by putting the sample is subtraction of the dark current from the total current, which is approximately 51.4nA. The value got from a commercial enzyme-labeled instrument is approximately 1600. At this point, we can see that the sensitivity of our device is enough for testing.
Due to the lack of time, we didn’t do more experiments to test our device. It’s a pity! If we have more time, we will do more testings.
Figure 5.This figure shows the increase of the current with the passage of the time.
Attachment
The files for manufacturing devices, instructions, communcation program and data processing program are contained in the following zip file. There is readme file under the root folder, you can get the contents of each subfolder from it.
Device.zip