Gimbal Mount Investigation
- Austin Tailon Huang
- Gavin Mesz
Due Date: Aug 16, 2022
Contributor Log
Name | Contribution |
|---|---|
Austin Huang | Task Manager |
Gavin Mesz | Owner |
Basic manufacturing and manual machining
Mechanical assembly
Task Description
Currently, the mount to attach the gimbal to the rover has not been assembled
The status of the mounting design is also in limbo - needs to be confirmed
Picture of gimbal:
Physical mounting components in the bay:
The gimbal tower is mounted on two thin rods, on a large horizontal plate. This seems susceptible to vibrations and large loads. Need to confirm if this is the current solution for the gimbal mount in CAD.
The owner for this task would be responsible for assessing the current gimbal mount design, making all required improvements and changes, designing and manufacturing any new parts, and assembling the hardware on the drivetrain.
Task Breakdown
Prerequisites
https://uwaterloo.atlassian.net/wiki/spaces/UWRT/pages/43200839681
Drivetrain CAD should be cleaned first to get a quick basis on discrepancies between physical mounting hardware and CAD data
Gimbal CAD should also be confirmed and rectified to assess current status of baseplate design
Progress Journal
Sep 8, 2022 @Austin Tailon Huang
Task and ticket created, currently unassigned
Oct 1, 2022 @Gavin Mesz
Went into bay and tried to answer the questions about the gimbal.
Figured out what the gimbal does and formulated some constraints and criteria that I believe the stand should have.
Constraints:
Has to be under 1.2m from GND
Should be able to spot ARUCO tags from 20m away
Be able to locate equipment stand
Clear picture
Criteria:
Very little vibration
Not blocked by arm while extended
Still need some more info on the comm stand so I can formulate more constraints.
Oct 5, 2022 @Gavin Mesz
Add seeing obstacles as a constraint. Ask if there will be a camera close to the front. This will definitely be a problem if there isn’t as it will be hard to see obstacles in front of the rover. Another reason why putting the camera on back left would be great to get out of the way of the arm.
Oct 6, 2022 @Gavin Mesz
Talked to Austin about design process, who to go to for info on the comm stand, strategy for my weeks to come.
Also got more insight on the exact components that need to be designed. Gave a more clear direction of what I have to do.
Presented first very rough concept of mount. Needs more knowledge of comm stand components and placement of camera to work.
Oct 24, 2022 @Gavin Mesz
Setup a meeting with Logan and Billy to talk about the comm stand components. In the meantime, I’ll focus on trying to figure out the best placement for the camera.
Investigated placement of gimbal using FOV specs for the ZED2 camera. Seems like the FOV is wide enough to place the camera anywhere. Placement seems less like a vision issue and more of a software/fitting parts issue.
Middle placement
Side placement
Pros of middle placement:
Possibly less work on software’s end.
Cons of middle placement:
Arm will cover camera on most missions, possibly requiring the arm to be used as a hazards camera as well.
Comm stand parts will have to be placed on either end of the camera. rather than beside each other
I believe that the cons outweigh the pros. Like many other rover teams, we should mount the camera on the side. Not in the middle.
Oct 27, 2022 @Gavin Mesz
Had meeting with Billy, Logan couldn’t make it to the meeting
Some important info I got about mounting the comm stand:
Orientation of receivers don’t matter, said they could be placed in the rover if needed
Antennas plug into a frequency specific white box (receiver). All of these plug into a big black box (edge router). The data is processed then sent to the Robot Operation Station (ROS)
Constraints:
Needs to be able to rotate antenna at least 270 degrees if directional antenna is used.
Antennas need to be placed farther than (1/4) wavelength (avoid first 3-4 multiples of wavelength) to not have interference (5-10cm recommended with lowest 2.5cm)
Antennas are 1.9MHz (157m), 2.4GHz(12.99cm), 5GHz(6cm)
Needs to be able to house at most 6 antennas
Criteria:
Possibility to use directional antenna on rover. Needs own motor to turn.
Ability to be adjusted to 2.5 degrees above ground for best calibration with ground station.
Good damping on antennas
Nov 7, 2022 @Gavin Mesz Tried for first ever CAD.
Cannot fit gimbal under the stand, too wide. Would need to get rid of tilt in order to keep it under. Also have a concern with the field of view under the plate. Scaffolding 1-1/2 in diam as opposed to the 20mm diam antennas
Concerned about deflection of Al plate on each end of the scaffolding. Shouldn’t be too much but with vibrations it would hit some sort of resonant frequency and amplify the wiggling.
I think the biggest thing is getting rid of tilt on the camera if the vibration is too much. If design is okay for vibration then no changes need to be made in my opinion.
Final result of today
Nov 13, 2022 @Gavin Mesz @Thomas Valer @Austin Tailon Huang @Joshua Harper
Meeting about drivetrain tasks to sync up on ideas, get to know what everyone is doing.
If battery is put in between scaffolding, height clearance changes.
Check meeting minutes for rough sketch of what we thought drivetrain should look like
Add translucent FOV for gimbal to check minimum height/ position
Finish new CAD with just antennae
Nov 18, 2022 @Gavin Mesz
Looking at the current design with only antennas mounted on top of the scaffolding there’s some problems with clearance.
With the current spacing (according to the previous rover and the current comm antenna mount prototype) the design looks like this.
Recommend using 2inx0.125in wall aluminum square tube on top instead so that antennas can be mounted on top and scaffolding can be on bottom. Will be slightly higher weight (432g against 1kg MAX no holes) Will add holes similar to current ground stand or can switch to aluminum 1x2x0.125 aluminum rect. tube (793g MAX)
Could help with cable routing and mounting components like LED matrix under antennas on top too
Final design concept for the day, very similar to how the 2x2 on the ground station looks. Will work to make it easier to connect/disconnect coaxial cables. Should just be adding holes and slots. Will also work to make slots smaller on top for small antennae.
Nov 21, 2022 @Gavin Mesz
Looked at possible lowest placement of gimbal, the top of the gimbal should be around the bottom of the scaffolding MINIMUM to be able to see ARUCO tags in front.
This also assumes that the battery will not be placed in the place it is now. If it is the gimbal needs to be placed even higher
Ways like this however can make the camera move down even more
Really concerned to do this for science however. will be completely blocked.
Conclusion: gimbal should still be mounted at around the height of the scaffolding if placed in the middle for optimal sight. If placed on the side, as long as directly in front, the battery or the arm isn’t positioned there, the front of the camera should be able to see ARUCO tags. However a very large portion of it’s FOV will be obstructed hence why high up is the best idea.
Nov 26, 2022 @Gavin Mesz @Austin Tailon Huang
Had a meeting about my current prototype design
Takeaways:
Talk about risk of placing gimbal on the same plane as the antennae
Talk to Josh about what would happen if I were to place gimbal on lower plane. Give minimum height and see how he reacts
Off-topic: Work on asking good questions and framing problem really well.
Billy said risk is pretty low, not a big issue. Question now is which configuration is more mechanically stable and which one works well for josh.
Nov 28, 2022 @Gavin Mesz
Prepping for Monday meeting:
Worked on making design with box tube on top and Plate holding gimbal in front.
Problems:
In original prototype 2 design, made vertical scaffolding distance wrong (260mm instead of 280mm) and now an antenna is placed on top of a vertical square tube. This is a problem because there is no way to secure the coaxial cable to the antenna and properly tighten it without seriously modifying current assembly.
Arm will hit gimbal if it is mounted in front
Camera in current situation will only allow for around 180-190degrees horizontal FOV due to servo and frame obstruction with scaffolding.
Space below gimbal is 7.52 in / 191mm total
My suggestions:
Change limit switch on A3 since it only accounts for the arm hitting the chassis/drivetrain. Current configuration will most likely hit e-box and gimbal.
Suggest not letting the arm pass this angle as to miss the farthest extent of the gimbal (45 degrees back). Distance here is 8.3 in (more than 8.26 in). Can alos try and move camera off to the side a little so that the arm doesn’t have to be in a weird position (sideways or outstretched forward) every time we want to see something with the camera.
I believe 180-190 degrees FOV is not bad at the moment. Can always just turn the rover. It would be great to get more.
Options:
mounting closer to the front (less arm maneuverablility)
mounting on the same plane as the antennae (not a big deal according to Billy).
Can always either move to a plate design for the antennae such that they can be mounted on the back.
Looks a bit odd but this could also help balance the moment on the scaffolding.
My questions:
How to add prototype stand to mainassy without disrupting GrabCAD?
Is the height clearance okay with Josh? Possibly consider mounting gimbal on top of E-box again.
Thoughts?
Dec 17, 2022 @Gavin Mesz
this is the current camera setup with the arm. Pretty limited vertical FOV but camera on the arm will account for this.
When just driving around, this seems to be the best area to put the arm. Just a little off to the side from the gimbal.
FEA RESULTS FOR DEFORMATION
this works too (camera won’t have 360 FOV anyways
max deflection of the plate is 0.16mm if we use ALUMINUM strut channel brackets with a 1-5/8inx3.5in support on the bottom. In this simulation I assumed that the brackets were both fixed fully to the square tube on their entire back surface area.
Max stress: 7 MPa with yield at 90MPa (according to solidworks)
If we use cheaper and lighter ALUMINUM (will be steel) right angle brackets, the deflection is only 0.1mm with a high stress of 7MPa.
Need to take into account that this will be constantly vibrating and will increase stress and deflection. Will need to test once built
Jan 8, 2023 @Gavin Mesz
First update of the year: Worked on the chassis mounting interface and the gimbal plate supports.
Chassis mounting interface
Was given the idea to keep a square bar that would screw into both the scaffolding and the chassis with a slot in the chassis block to keep the bar parallel with the front.
Gimbal plate supports
I told @Austin Tailon Huang that I would be fine with him trying to work on the supports but I decided to get started on some design. I’m going for something similar to the angle supports in the wheel (MR22-DT-P0113) as I believe that will counteract the moment created by the gimbal best with the least amount of parts. Below is the result of my work.
The red thing is a tube (nylon or another plastic for weight reduction) that will be used to locate the plates.
Relevant Documentation Links