When I was in college I worked for a Professor that did studies with loss coefficients in bifurcations (aka fluid pressure loss in flow junctions). To check the fluid pressure at different points along the pipe he used some analog pressure sensors that fed into a TLC2543 ADC chip which was then connected to a Parallax Basic Stamp and finally via serial connected to a desktop.
This worked, but it had its limitations:
1. You were limited to how many pressure taps you could have unless you setup a whole 2nd system.
2. You were limited to having a computer near by that could run the python script to take in the serial data and parse it out properly.
3. It used a Basic Stamp, which were expensive to replace.
After I had graduated and had some free time on my hands I decided to redesign part of the system so that an Arduino was used to be the middle man between the TLC2543 ADC and the computer. I also decided that I had to figure out a way to make the system expandable. The Arduino part wasn’t very hard to figure out, and as for the expandability I figured out I could use the chip select line on the TLC2543 and using a multiplexer (mux) I could then control 8 different ADC chips, which means 8 times as many sensors!
Here is the code for the Arduino:
The format of the serial output was determined by the python script running on the computer.
I also decided to redesign the board that all of this was hooked up to!
(The only fully populated board resides at the University. I couldn’t afford to fully populate my other prototypes because the pressure sensors were so expensive.)
You can see the headers that the Arduino attaches to just left of center in the top half. The TLC2543 sits right in the middle of the board. Under the Arduino was the mux chip and a set of DIP switches that allowed you set the “address” of the board. This design didn’t support the boards being chained together, but it was a proof of concept that it was possible. I had planned on leaving up to his student team to redesign the board with the connectors that allowed it to be daisy chained.
Unfortunately after I had given the fully populated board to the professor nothing ever came of the system. I have since tried to find the original design files but have been unable to locate them.
After thinking about it more I then decided, for fun, to step it up another notch and use a Raspberry Pi to interface with the TLC2543 (instead of an Arduino) and then upload the test results directly to Dropbox! You can head over to my Raspberry Pi pages to check out that project!