Week 6: Validation Complete!

This week marks the end of our component validation phase and the beginning of our system integration phase.

Mechanical

For the most part, the mechanical body is done and put together. Our plans for the water tank still remain up in the air, but we hope to keep exploring this more with our new tank design. We're hoping that 3D printing it will ensure there are no holes for leaks, with the final wall as a separate piece of acrylic for the capacitive level sensing. We wanted to keep the wall with the level sensor acrylic since we weren't sure how the results would change with 3D printed filament instead, since we need to be able to sense the water capacitance through the walls of the tank. This does mean we have to be careful of leaks when attaching that final wall. We're hoping that since it's just one wall and the interlocking design has been scrapped that some Crazy Glue, gasket, or silicon sealant will keep everything dry.

Electrical and Firmware

We finished component validation with our last component, the power switch. We tested this out by hooking the switch up to a simple LED project on the Arduino and turning the light on and off with the switch. There were three terminals, which confused us a bit, but we figured it out and made sure to remember it.

Switch configuration (and LED in the back)

We also started integrating a few of our sensors. We started with the LDR sensors by hooking them up to some wires and attaching them to the mechanical body. It was pretty cloudy today so we had to work around this by testing with a desk lamp that had different lighting settings. Once we verified that the LDRs were connected properly and detecting light we integrated the sensor testing firmware into the finite state machine.

We also integrated our ultrasonic sensors in and tested the code using two ultrasonic sensors at the same time instead of one. We ran into some issues here where only the one ultrasonic sensor was giving us a reading. After some research we realized we couldn't trigger both sensors at the same time and instead had to do it serially, one after the other. The problem with this is we didn't want to use a strict delay from Arduino as this stop execution of the entire system, so we had to spend some time implementing this using time elapsed.

Once these parts were verified and confirmed to work, we began soldering them into place. We soldered three LDRs to wires that we can easily solder to our prototype board once everything comes together. We also soldered wires to the motors and the pumps to get them ready for integration as well. Since we did our soldering at home (beside the window to make sure we're safe) and we didn't have a clamp to hold our components in place this ended up being a two person job.

Cindy soldering the LDRs into place

The next step for us is to test the motors and IR sensors, and so we started with the IR sensors. Unfortunately, after wiring all 3 pairs of emitters and receivers and wiring them into our robot, we discovered that the IR sensors we got were far too sensitive to light. This meant that when we turned the lights off, we were practically unable to detect whether or not an edge was detected. Rather, the values would just stay the same as if there was an edge present when there clearly weren't. After finding this out, we were destroyed and heartbroken. But don't worry, we managed to find some IR sensor pairs online that came equipped with a board to handle noise from external sources such as light. Hopefully this would make it easier to detect an edge... right?

Stay tuned for next week where we'll see if these new sensors work, and we try to get our robot moving!

Other