Before leaving the Foundry at 8:30 AM today, all sensors and thrusters on the robot were 100% functional.
We put the robot in the water while the robot was still off, and we quickly noticed many bubbles coming out of the main capsule, so we took the robot out. The main capsule had not formed a watertight seal, especially in the back right corner. The capsule seemed slightly deformed on the side with the tabs.
Quite a bit of water was in the capsule. We drained it and dried off the plate. When we attempted to put the capsule back on, we noticed the robot was turning on and off in spite of both switches (red and blue) being off. The voltmeter and other lights in the electrical stack were turning on and off randomly.
In looking at the robot on land, we could see that the red switch was a bit unreliable, but after some time (presumably to dry off), it worked fine. We should probably replace it soon, but not an immediate issue. We already had new switches ordered.
As far as the capsule, there was certainly some water on the base of the plate. We suspect it because there were a lot of cables jammed along the sides, and something may have not aligned on the bottom of the capsule. Upon further inspection, there was also water around the powered USB hub. When testing USB devices, it appeared as though the 3 USB extension cables were all bad.
We were able to do some debugging and got all the USB devices to work if we went without those extensions. This is except for the thruster Arduino and sonar. For the sonar we need to look at more in depth. It is not an immediate problem, but we suspect it is because the little sonar conversion board is on the bottom of the stack. As for the thruster Arduino, it sometimes connected to the computer. When it did connect, the thrusters made weird robotic sounding noises we have not heard before. The voltage would drop to the 14s when running (from around 15.5). After some more testing of trying to control the thrusters, we could hear some fun electrical noises indicative of something dying and then we saw a lot of magic smoke come from what appeared to be the multiplexer. We tried to determine exactly where it came from while it still smelled, but it was really hard to tell.
Okay so going forwards:
- We should not turn on the entire robot until we have found and replaced the fried part(s)
- We need to look at bad USB. Specifically, replace the powered USB hub since that was wet
- DVL capsule connector is corroded. It did not affect performance, but we should look at this
- Replace the switch when we get it
- Debug the sonar and its connection board