I started the week off strong by printing and cleaning up all of the rims for the thruster guards, which are in place for the safety of anyone who finds themselves in a position in which they are underwater and handling the ROV, as skin becomes more delicate and permeable underwater; the rims for the thruster guards serve the purpose of protecting those delicate personnel from sharp edges that can result from the process of cutting the mesh.
Despite completing the manufacture of the thruster rims, and carrying them with me in a string bag, to and from school, every day, I had not found a time to install them.
We made a decision over the week, due to a numerous variety of variables, that we will re-do the wet link penetrators. And what I mean by “we will re-do the wet link penetrators” the freshman will re-do the wet link penetrators.
This will serve numerous purposes. It may solve a major issue that has cropped up this week: the ROV leaking. It will also allow us to re-organize the thruster and tether cables on the outside of the ROV.
Another change we are considering making to the ROV is shifting around how the interior electronics interface with the exterior cables. We are considering adding an in-between PCB, with ports on it for interior cables, which are currently hard-wired to the outside cables. Making this change will allow for Much easier testing of the circut boards, and removal of the main PCBs from their places entirely.
On Wednesday we set up for more pool testing. I removed the new tube clamps and replaced them with the old ones, keeping a baseline for my data. I then installed the 80-20 shrouds with the best pilot in the world: Jonas and the best electrical engineer in the world: Miles.
When we put the ROV into the water, we encountered a massive, gigantic, humongous, catastrophic, apocalyptic issue: there was a tiny leak in the ROV. On the pool deck, we tried to re-lubricate the rear flange and Jonas tightened the screws holding on the front plate. Even after we did this, the ROV was still leaking.
We ran out of time on the pool deck, so we brought the ROV back to the shop. In the shop we deep cleaned the sealing components such as the O-rings and flanges. After re-assembling the ROV found that there was still a leak by pulling the tube out without removing the vent cap. If there was no leak, trying to do this would cause a vacuum preventing the end from being removed.
From this, I came to the conclusion that one of the wet link perpetrators was the issue, contributing to the need to re-do them.
Also this past week I cut claws for the mechanical gripper out of aluminium. The design of the claws consists of 2 1/8 inch claws with a 1/4 inch geared spacer in-between them. this side is mounted directly to the servo. The other side has a 1/4 geared claw that is attached to a custom 3D printed mounting plate attached to the servo, allowing for free-swivel movement.
After I cut the claws, I spent around an hour with a file and de-burring tool, cleaning up the parts. Then I assembled the gripper.
After some quick testing I realized that I needed to install washers in-between the spacer and the drive claws. After I installed those and set up the electronics, the gripper worked excellently.
While doing some testing, Miles and I brainstormed a better current limiting system. We decided on switching out the linear voltage regulator out for a buck regulator. Switching them will increase the efficiency from around 60% to around 90%. We also decided to add in a current sensor so that if the gripper still consumes too much power, we could use the sensor in software to reduce the power consumption.
This software would monitor the current, when the current spiked past a limit we would know that it would be holding something. The software would then back off the gripper by single degrees at a time, relying on the compression from the rubber on the claws to hold the object while using limited power. I manually tested this using a power supply and a rotary encoder, and it worked excellently.
During the whole week I have been working on the design for a new electronics tray. I decided that this was necessary due to all the unexplained electronic failures the ROV experiences. It has been my theory that some components are loose or touching, or some other physical issue has been the cause of weird failures. It is also generally good to remove a point of failure: the weak and overly flexible electronics tray.
I started by sketching out some basic designs and iterating as I went along. I find this to be a cathartic process and a great way to work through complex problems.
I decided to replace the current thin threaded rods with thicker non-threaded rods to increase the rigidity. In order to increase the rigidity further will replace the acrylic mounting plate with 2 stacked water jet cut 1/4 inch aluminum panels with tapped holes in them. I will thread the ends of the rods to screw them into the plate.
The slots for the PCBs will be 3D printed and will slide onto the rods. They will be adjustable, locking into place with set screws pushing on an interfacial component. One side of the slots will have a more open design to allow for the ports coming off of the boards, with the other side being more significant for structural integrity.
I measured dimensions of the electronics and consulted with Miles to determine that the best length for the rods would be approximately 115 mm or exactly 7.5 in.
I found that 1/4 inch rod fit best into the slots of the PCBs. I searched through the stock and found the perfect pre-precision-ground steel.
I marked the steel at 7.5 in and cut it using the horizontal band saw. I used a file to clean up the end of the bar.
I asked Aaron for help tapping the threads on the bar. He tried to help, but the die wasn’t working for some reason. Mr. L didn’t know why it wasn’t working either.
Because we were so confused I decided to just use the lathe to thread the ends of the bars next week.
On Friday I plan to be put under general anesthetic for a couple hours while a medical professional sticks a camera down my throat. I will greatly benefit from this because I will regain the sleep I lost making this blog post.
