March 31, 2002
We finally have all of the necessary parts to start building and testing our project. Some initial problems were getting the correct amperage for the air compressor and manifold. The air compressor needed 17.2 Amps to run, and that is a lot of current! Two of our teammates wanted to build a power supply, which I thought would be an unnecessary burden on time. So I went out and bought a power supply that could support 25 Amps of current. After testing it, we concluded that the power supply worked great with our air compressor and are finally able to begin construction on our hand.
Also, we needed 150mA to power one of the solenoids at 12 V. This is actually a pretty significant problem. For each solenoid (20), we need to have an op-amp connected to a power transistor that would supply the voltage and the current. The power transistors cost about 2 bucks a piece, so about 40 dollars for the transistors. The op-amps cost about a buck each, so that’s 20 dollars more. Plus, we need 2 W resistors to take the current, which is another twenty dollars. Overall, our amplification is going to cost us about eighty dollars. A small problem has a profound cost.
We are prepared to start testing out the product. I have to write a program that will control two solenoids, and our goal is to successfully move a cylinder up and down.
April 1, 2002
Our goal of successfully moving a cylinder is fastly approaching. Today I was playing in lab and got am amplification of 12 Volts out with 150mA of current. The current was needed to successfully move a cylinder.
Today I found an analog demultiplexer that would solve some of our problems with the amplification. The chip is a CD4067B, which takes one input and selects which channel to output the input signal to. I found this very convenient because our microcontroller can send outputs directly to that instead of wasting the onboard ports. Also, the chip allows for inhibiting, so all the outputs can be shut off immediately. The only problem with the amplification is the requirement of the transistors. The plus side is there is now no need for the op-amps, which would take up considerable space.
May 10th, 2002
A lot more problems. A lot more solutions!
The analog demultiplexer wouldn’t work because it didn’t output the amount of voltage we needed to operate the solenoids in the manifold. This caused me to search the web for transistor circuits that would output the 12 volts and 150mA. However, the feasibility of 20 transistor circuits, much less the appearance, stalled the search. I was randomly searching the web for solenoid drivers when I came across a senior project report in PDF format that used a few valves for their project. They used the sn754410 chips, which houses 4 solenoid drivers that outputs 12 volts at up to 1 Amp per solenoid driver. My goodness was I in heaven then when I discovered this chip.
I immediately ordered five because we needed to operate 20 different solenoids. So I talked my senior project people into going into lab today and seeing if we could get one of the cylinders to move up and down. After testing the chip to make sure it outputted the correct 12 volts, we hooked the chip up to the manifold and tested it. NOTHING! Can you believe that? NOTHING! However, I practically guessed that we were getting nothing because something wasn’t grounded correctly. I tied all the common supplies on the manifold together and put it to ground. One of the solenoid LED’s lit up. After connecting an air supply to the manifold, we discovered it was working. We hooked up a cylinder and the piston moved out! WE GOT IT TO MOVE! So our goal of getting a cylinder to move has been complete. Our next goal is to figure out the type to get 4 possible positions on the cylinder. Stay posted.