WATL_Hardware (SD) Core Concept

[Artefacted]

Hi everyone,

Please find first concept attached.

This is the propsed design for the core hardware, which facilitates rigid connection to the star picket and will hold modular sensors using a standardised mounting adapter.

The core allows up to 4 modules to clipon at once and 90° separations eliminate issues from equipment stacking and allows our Pi Cam to easily obtain its 270° view field. The core includes prongs that can gently hold ribbon cables between modules. These prongs also mark slots which run to the interior of the core, which allows us to simply zip tie the core to the picket.

The total height of the core is 140mm, with the top 20mm forming a cap over the top of the picket. This means a paint mark 1880mm down the picket could be used when driving it into the ground to calibrate the top of the WATL to a z-height of 2000mm from natural ground level.

The module adapter (shown in black) has vertical slots and top corner holes which will hold an m3 nut captive and allows us to flexibly bolt sensors onto the WATL core, and quickly swap sensor modules. Pointing the camera at the fire is as easy as rotating the whole core to the nearest 45° to the fire before firmly placing onto the picket.

Not yet included are the individual module designs for the camera housing, etc. These will incorporate a modular clip into a unibody component.

I will post the .STL shortly I need you to 3D print the core vertical orientation up and the 4x modular clips on the flat. This will use the step lines in the print to friction hold the adapters to the core. Use your PLA printer.

Comments/questions welcome,

D

AS0136_WATL_SD01.pdf AS0136_WATL_SD02.pdf

[nicholasfm]

Hi Dan,

Excellent work, loving the concept for the modular design. This too would allow custom deployments where additional instrumentation can be deployed (or not deployed depending on the objectives).

I have a comment that I'm not sure of the solution to, and could be something to think about @uqsbadat . Long term we'd want these to be weather proof, which mainly means weatherproof sockets. The GPIO ribbon cable (which you've incorporated guides for) will most likely be used for the prototyping phase, but eventually we will want a way to have a cable suiting such weatherproof sockets running between modules. This was the only advantage to having a stacked design is that they the connection between the sense hat and the Pi needn't be anything other than a ribbon cable. It could be a simple matter of finding an adaptor for a more robust cable to run through weather proof ports from module to module. Or getting creative with heat shrink tubing in the interim?

The anemometer with have a waterproof RJ11 socket such as this: https://www.adafruit.com/products/827 though that's for an RJ45, not RJ11.

[joshua-wx]

Clean design! Great starting point for this project.

I don't this it needs to necessary have waterproof sockets for the ribbon cable, but rather just have a weather proof design that minimised water intrusion (this will also help minimise costs). The Pi camera on Camberlain has had water in it for over a year and it still works fine :)

J

On 14 October 2016 at 12:23, nicholasfm notifications@github.com wrote:

Hi Dan,

Excellent work, loving the concept for the modular design. This too would allow custom deployments where additional instrumentation can be deployed (or not deployed depending on the objectives).

I have a comment that I'm not sure of the solution to, and could be something to think about @uqsbadat https://github.com/uqsbadat . Long term we'd want these to be weather proof, which mainly means weatherproof sockets. The GPIO ribbon cable (which you've incorporated guides for) will most likely be used for the prototyping phase, but eventually we will want a way to have a cable suiting such weatherproof sockets running between modules. This was the only advantage to having a stacked design is that they the connection between the sense hat and the Pi needn't be anything other than a ribbon cable. It could be a simple matter of finding an adaptor for a more robust cable to run through weather proof ports from module to module. Or getting creative with heat shrink tubing in the interim?

The anemometer with have a waterproof RJ11 socket such as this: https://www.adafruit.com/products/827 though that's for an RJ45, not RJ11.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/uq-crg/watlnet/issues/7#issuecomment-253693425, or mute the thread https://github.com/notifications/unsubscribe-auth/APTMSymyDLe08NZMcQVsIRdvQi6_Wq4jks5qzue0gaJpZM4KWhy6 .

[Artefacted]

Hi,

Here are STL files. Time to organise to have these printed so we can check tolerances. Please keep the mesh oriented with Z-up as is.

Next issue from me will include revised adapter pieces for individual sensors.

Re: Water intrusion I think we can handle this with silicon/putty seals on the slot where the ribbon enters the sensor housing. I'm aiming to afford a certain level of ruggedness to these things, and a 'splash test' should perhaps be added to fire and wind tests.

20161017 WATL-01 3D.zip

[joshua-wx]

Everything looks good in blender. My only concern is the stability on the star pickets, but the variability in star pickets means we really just need to test a few.

Looks like we have a makerbot replicator at UQ. They charge $5 to CC per model with no limit on size. https://web.library.uq.edu.au/library-services/it/3d-printing

On 17 October 2016 at 18:22, Artefacted notifications@github.com wrote:

Hi,

Here are STL files. Time to organise to have these printed so we can check tolerances. Please keep the mesh oriented with Z-up as is.

Next issue from me will include revised adapter pieces for individual sensors.

Re: Water intrusion I think we can handle this with silicon/putty seals on the slot where the ribbon enters the sensor housing. I'm aiming to afford a certain level of ruggedness to these things, and a 'splash test' should perhaps be added to fire and wind tests.

20161017 WATL-01 3D.zip https://github.com/uq-crg/watlnet/files/532699/20161017.WATL-01.3D.zip

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/uq-crg/watlnet/issues/7#issuecomment-254143265, or mute the thread https://github.com/notifications/unsubscribe-auth/APTMSyoRNNmqbWhLqt2auCsjDhqARKUKks5q0zA7gaJpZM4KWhy6 .

[nicholasfm]

Hamish made the comment that we may experience star pickets getting a bit 'pancaked' from hammering in at the top. I wondered if the reason that the star picket edges don't sit in flush in the current design IE the little circular grooves at the end of the slots?

Anyway, simple testing will tell with tolerance. I've paid and sent them to the Library printer as I haven't heard back on from Advanced Engineering (cough).

Sounds good on the water intrusion issue. It's not a concern in the short term and not an issue for my objectives in the long term either honestly. More supercell proofing @joshuass ? Though rain may not be our issue there...

[uqsbadat]

Hi everyone,

A few comments about this almost perfect design to improve it for our application :

• The room designed to connect to the star picket seemed to be too big when we tried it on several ones. What dimensions did you use for the star picket ?
• We couldn't insert the modules in the core because they were too thick and with a too big width. Maybe if you could reduce this thickness and width by 1-2mm.
• I can't see how we can screw objects onto the black parts and not in the core at the same time. maybe the black parts shouldn't be symmetrical to allow us to fasten our devices on the outside part.

Than you for your work,

Stan

[Artefacted]

Hi All,

Responses to queries below I. the internal dimension was 40mm from memmory. I was working on a 2m sensor height. This means a 2.5m picket, obv. way larger than the average picket. If we want to use found pickets, I can incorporate little wings into the interior of the cap which could grip say 30-40mm picket widths

I I. This will continue to plague us with fdm printing. Next design will have 1-2mm larger tolerances. We'll fine tune as we go.

I I I. The intent is to screw into the black parts with them off the core, and so tighten with screw head facing in and nut captive on the outside. This means using screws with a small head (perhaps countersunk) but offers an advantage in preparing modules off the core, then sliding them on. The logic is that modules should not be screwed directly into the core so they can be removed/hot swapped /repaired/have data accessed independent of the rest of the assembly.

Next issue from me will review the shape of sensors and continue to assess the logic of how best to arrange fixings, and tune tolerances.

Thanks for the feedback

[nicholasfm]

Hey @artefacted thanks for the feedback. Happy to experiment with tolerances while still fdm printing, any chance of knocking these over soon? Any chance of updated timelines?

Code is basically done, apart from windspeed calibration which will be done in the field of tunnel (ie hopefully in soft prototype stage.

Also, is there an update on where the radiation shield is at? I reckon the mounts will be fine for now in terms of bolting with tolerances- covered this with Stan.

[Artefacted]

Hey Here are .stl files with revised tolerances. I am working on sensor modules, including radiation shield and have no major issues, just working through it. Will be very interesting to see how this thing handles the weight of anemometer, but it should be manageable by gripping the picket nice and deep. Individual module mounts next.

20161115 WATL-02 3D.zip

[nicholasfm]

Awesome, thank-you @artefacted, great info.

Can you organise the print @uqsbadat?

[uqsbadat]

Hi All, Just received the new design and the dimensions are perfect ! the module slides in the core, and the core and the star picket connect easily. thank you for this @Artefacted

[uqsbadat]

Hi @Artefacted ! the code is almost done, just need to do a few calibrations. It will be very exciting to see how it works with all the new sensors and your designs, but it should be great !

[uqsbadat]

Hi @Artefacted ! Sorry to bother you again.. What size of screws and nuts are the holes in the module made for ? Thank you

[Artefacted]

Hi @uqsbadat The sizing is to hold m3 nuts captive, the slots in the adapter piece will also stop a nut from turning, but the slot allows some flexibility attaching different shaped sensors. Sensor board designs coming soon.

[Artefacted]

Hi Everyone, Here is the sensor module. Two rhythms of Stevenson screen, designed to receive a SenseHat board. There is a separate file for a clip that should sit below the board and make it easier to insert/remove from the screen. The stevenson screen should be oriented in the .stl to print upside down, with the cylinder stopping it from toppling during printing. Let me know if you have any issues, we can either print with a raft or I can model some extra pads to keep the print stable. 20161115 WATL-03 3D.zip

[nicholasfm]

Hi Dan, Can you confirm the dimensions of this guy? Shapways is asking for ~$700 per print for the screen and 800 cm^3 of material?

[nicholasfm]

@Artefacted - sorry man, should have tagged you in the above to get the notification. The library (doing our soft prototype printing) informed us the dims were too large. Response was as below:

Unfortunately the dimensions of the file we received is too large. Can you please confirm the dimensions of your job so printing can begin. I have attached a picture of one as an example.

[Artefacted]

Hi The height doesn't need to be quite so tall. Please find sensor meshes attached which are resized to fit within the library's 252mm x 199mm x 150mm bounds. Shapeways quote sounds scary, I'd print at the library for $5 if you can ;)

20161213 WATL-03-rev1 3D.zip