Fully 3D Printed Enclosure

[keenanjohnson]

I would really like the sensor to be green-colored, since it is a frog after all.

The enclosure is made of an unknown plastic.

Spray paint would probably be the easiest, but given that this is a greenhouse gas project, I'd like to avoid the aerosols and the nasty volatiles contained within.

[keenanjohnson]

No worries! It's all good because the USB connector wasn't in the model previously, so I didn't see it until I did the print. You did the heavy lifting with the initial concept!

I would like to make it smaller just to minimize print time, but unfortunately, we have a USB cable coming out of the top of the board as well to connect the GPS Unit (That's also not modeled, but I should fix that haha). I'll post a picture soon with the electronics in the box and you'll see what I mean. Once I put it together and get a better sense of the bend radius for that cable, I'll have a better sense of where it can shrink I think.

[keenanjohnson]

Ok here are the pictures of the build of my cad above. You can see all the space that the stupid GPS USB cable takes up haha. I think there is room to move the GPS antenna above the USB cable and shrink the whole box horizontally, but there isn't really room to shrink vertically. I just ordered a right angle USB cable I'll try to see if that saves us any vertical space.

[eaudiffred]

Those cables take up way more space than I thought! What do you think about shifting the GPS antenna down to the bottom, and then sliding the wifi (?) antenna back up into the box along that whole right side? Is it astatically an issue to have that antenna hang out?

Hopefully the right angle cable buys some space. There is no way to change the Pi layout, is there? Move the power port to the other side with the other cables?

The external components on top seem to fit in-between the fins. Any issues up there?

I haven't had a chance this week to play around with it. Hopefully I get some time this weekend.

[keenanjohnson]

All right. I think the move is actually to move to a different GPS with an integrated antennae and place it in the top area with the other sensors. This is a little bit more costly, but I was having poor GPS signal issues with the other sensors being between the sky and the GPS antennae. I haven't finished modeling that in 3D but will tomorrow. Link to assembly

I've updated the CAD to move the wifi antennae back inside the enclosure now that the GPS is gone, there is plenty of space for that. Clearance in the top area is perfect right now.

I fixed a few small things like increasing the length of the slots to attached the bolts in and some other minor positioning tweaks.

The main issue now I believe is that the seal is really small and it's really challenging to get it to stay in place. @eaudiffred ow did you imagine keeping the seal in place while the lid is screwed in? Should the seal be glued to the lid or enclosure first?

[eaudiffred]

Great updates! Since you've gotten through a ton of fixes I just copied your most recent file and then made some changes to the oring cut out path. Assembly

Generally orings don't have much, if any memory so they just kind of lay in place. Maybe this is a different story because its so small. You could try glue for this one. Generally its not good to glue the seals in place though. You could also try heating it with a hair dryer to get rid of the shape it was in the package and see if that helps it become less resistant to taking a new shape.

The change I made to the slot path shifts the slot down so that the oring can be pressed into place and retained by the part of the slot that starts to come back together. It's hard to write into words. Hopefully these pictures make sense.

The other thought I had is shrinking the slot diameter to 0.9mm or 0.95mm to make it tight.

If none of that works we could also try a flat seal or gasket. The trick with that is getting it cut to the shape you need. There are also liquid gasket options. Basically like silicone. That isn't reusable though, so every time the lid is removed you have to scrape off the old and apply new.

[eaudiffred]

Another thing we could try is a "tooth" design. That would eliminate the need for any rubber material. I think it would be less sealed, but would still do a good job of keeping out water.

This basically builds in a ridge on the lid that slides into a slot on the enclosure. Just makes it that much harder for water to work its way in.

[keenanjohnson]

Awesome! Yeah I think that this o ring must just be so thin that the surface tension causes it to have some memory. Perhaps that would change with heat, but I really like your other ideas also! Trapping the o ring would definitely help to align it.

For reference, I took a look at an Off The Shelf outdoor enclosure I had laying around and here are some pictures of what they did. It seems that they did both of your ideas (the tooth and a trapped seal in the tooth slot. See the pics below. I really like the tooth idea because it simplifies the construction for nonexperts who might want to build this thing (school kids, etc) and limits the part count.

Perhaps the seal isn't required with a tooth design since we have a vertical enclosure? Since the slot is vertically aligned, the water may enter the slot, but should flow to the bottom of the enclosure where it's pretty much harmless since it's away from the electronics.

[eaudiffred]

I agree. I think the oring/seal is over kill for this specific situation. The tooth design should be plenty protection and simplify assembly. I'll draw it up this afternoon.

[keenanjohnson]

Sounds good! I plan on giving the first test item a good soaking with the hose which I think should give us a good amount of information if that is adequate or not haha

[eaudiffred]

Ha! Yes, the hose soaking should tell us what we need to know.

Here are a couple of pictures. What do you think? Assembly

[keenanjohnson]

Looks awesome to me!

On Wed, Oct 6, 2021 at 2:56 PM Eric Audiffred @.***> wrote:

Ha! Yes, the hose soaking should tell us what we need to know.

Here are a couple of pictures. What do you think? Assembly https://cad.onshape.com/documents/d3d25a3218219164c6876e91/w/0529d3300c2829e6698f062d/e/e1be5b132f4cd14bae918704

[image: image] https://user-images.githubusercontent.com/88562047/136288819-fd2ea40e-a3e8-4e8f-a874-527c1cf4a2c0.png [image: image] https://user-images.githubusercontent.com/88562047/136288860-bd7a986c-10e6-43b6-8aad-037a65ca42bf.png [image: image] https://user-images.githubusercontent.com/88562047/136288904-d7ebfad3-e470-42b1-8c3e-9f3774f73414.png

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/Ribbit-Network/ribbit-network-frog-sensor/issues/18#issuecomment-937218536, or unsubscribe https://github.com/notifications/unsubscribe-auth/AATQ3FXTVDISUWYVS6NRICDUFTA2HANCNFSM5BYGC5CQ .

[keenanjohnson]

Let me know when it’s ready for a test print :)

On Wed, Oct 6, 2021 at 3:03 PM Keenan Johnson @.***> wrote:

Looks awesome to me!

On Wed, Oct 6, 2021 at 2:56 PM Eric Audiffred @.***> wrote:

Ha! Yes, the hose soaking should tell us what we need to know.

Here are a couple of pictures. What do you think? Assembly https://cad.onshape.com/documents/d3d25a3218219164c6876e91/w/0529d3300c2829e6698f062d/e/e1be5b132f4cd14bae918704

[image: image] https://user-images.githubusercontent.com/88562047/136288819-fd2ea40e-a3e8-4e8f-a874-527c1cf4a2c0.png [image: image] https://user-images.githubusercontent.com/88562047/136288860-bd7a986c-10e6-43b6-8aad-037a65ca42bf.png [image: image] https://user-images.githubusercontent.com/88562047/136288904-d7ebfad3-e470-42b1-8c3e-9f3774f73414.png

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/Ribbit-Network/ribbit-network-frog-sensor/issues/18#issuecomment-937218536, or unsubscribe https://github.com/notifications/unsubscribe-auth/AATQ3FXTVDISUWYVS6NRICDUFTA2HANCNFSM5BYGC5CQ .

[eaudiffred]

Good to go.

Where there any other tweaks that need to be made?

[keenanjohnson]

I don't think so no! But I'll let you know! You rock for helping out on all this!

[keenanjohnson]

I added the 3D model for the GPS Unit and USB cable into the 3D model. I had to move the mount holes over slightly and adjust the opening, but everything fits just barely haha. You can see the sensor stack up I have, which I've tested physically and seems to work nicely. Let me know if this workflow of copying the assembly each time works for you. We could agree on one document and both edit it, or continue with this workflow while we're making lots of changes.

Here's the new CAD Assembly.

[eaudiffred]

Looks good. Close fit, but still clearance.

Up to you on the work flow. I think it would be easier if we could both work off of the same document. However, we might loose some of the version history that we create by creating these copies. What do you think?

[keenanjohnson]

All right the print is finished and things look generally good! The fit is really snug and I had to press on it quite a bit to get it to pop in, so we might want to look at increasing the tolerances in the groove slightly I think to account for a bit more variation in the production.

Just one issue I noticed when screwing it together: because the bolts for the head and sensor shrouds before the box lid goes on, there isn't enough room to get a standard Allen Key to put in the top two bolts for the electronics box (hopefully it's clear in the pictures below). Perhaps the best solution is to move those top bolts down so they aren't in the corners, but instead in the upper-mid section below the sensor shroud so they can be easily installed (see pictures below)

Let me know what you think! As for the OnShape Assembly, yeah it could be nice to just work in one Document. That's how I've worked at companies with OnShape before. To keep the history, we can each just create a "Version" of the document each time either of us makes a change which will allow us to keep all the revision history! I believe my assembly here is the most recent, so I'll add you as a collaborator there and let's just create the versions in that assembly going forward. Shoot me an email or text if you have any issues or need help figuring that out!

[eaudiffred]

You are getting these turned around quick! I'll increase the tolerance by shrinking the tooth. I think its only 0.05mm smaller right now. I'll increase it to 0.2mm tolerance total, so 0.1mm on either side.

The version history is cool. That's defiantly the way to do it.

The bolt position is a bummer. I'm afraid that moving them lower will cause space issues with the antenna or the Pi. The groove for the lid tooth takes up a lot of that edge space. What do you think about widening the shroud, or modifying it somehow to allow the tool to fit?

I'll explore both options (moving the bolts and expanding the shroud) and see what I can come up with.

[keenanjohnson]

Awesome! I wonder if we even need top bolts? The lid is pretty stiff and just the two bottom ones seem to retain it just fine. Perhaps if we just added some sort of clip or feature on the top of the lid enclosure and just used the two bottom bolts that would be adequate? Maybe not even that.

[eaudiffred]

I added two posts instead of bolts. I loosened the tooth/groove tolerance too.

[keenanjohnson]

Awesome! That seems like a good compromise to me!

[eaudiffred]

It may not be totally necessary, but for some reason I was nervous about relying only on the tooth and groove to hold it in place. We can always tweak it in future versions if it proves to be unnecessary.

[keenanjohnson]

@eaudiffred the new design seems like the best version yet! The tolerances are much better on the groove making the lid easy to snap into place now.

The pegs work great however, when I tried to remove the lid I tried lifting it out at an angle and the angle caused one of the pegs to snap, which I imagine might be a common occurrence with this design. I'm not sure what the best way to fix that is, perhaps you might have some ideas?

[eaudiffred]

If you break off the other peg and then reinstall the lid does it seem secure enough? Meaning, do the pegs serve a valuable purpose?

If they do add support, rigidity, alignment, etc. I could try something like below. It'll then open more like a hinge and allow the door to be opened by the bottom, tilt outward, and then slide off/out.
rough looking drawing, but hopefully it shows the concept, lol.

[keenanjohnson]

It honestly feels pretty secure without it. I had to pry pretty hard to get it off, so my guess is that it would be fine without it. The nubs might be nice just to help locate the lid and guide it into place, so maybe the quarter circle design you have above could be nice! What do you think?

[eaudiffred]

Updated with the quarter circle design. I think having some alignment is a good thing.

[keenanjohnson]

As we discussed @eaudiffred let's release this as a version 2.0 of the Frog Sensor alongside the new compute (discussed in #62). I've added it to the Frog Sensor V2 Milestone.

[eaudiffred]

Sounds good. This is the board, right? BEAGLEBONE_BLACK.pdf

Where with the USB and power cords be plugged in? Is it just these two that I've circled?

[eaudiffred]

The boards are very close in over all size. The bolt holes will have to be shifted a little, but it won't really make a huge difference. The biggest change will be the location of the plugs/cords and the larger wifi dongle.

[keenanjohnson]

Yup! The new power jack is going to be the barrel plug on the right.

On Tue, Nov 2, 2021 at 7:44 PM Eric Audiffred @.***> wrote:

The boards are very close in over all size. The bolt holes will have to be shifted a little, but it won't really make a huge difference. The biggest change will be the location of the plugs/cords and the larger wifi dongle.

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/Ribbit-Network/ribbit-network-frog-sensor/issues/18#issuecomment-958610466, or unsubscribe https://github.com/notifications/unsubscribe-auth/AATQ3FQLBOZGADAHHJQNOXDUKCVZTANCNFSM5BYGC5CQ .

[eaudiffred]

Cool. I made the mounting hole location changes and moved the power cord entrance hole. It looks like there is a lot of space in there now, but I know the cords take up room and so will the dongle. Like you said, I think there is space to mount the dongle vertically next to the board. Is the Tenda W522U what you plan to use, or was that an example?

[keenanjohnson]

That was just an example. I have a few different wifi adapters I'm trying to test tomorrow and I'll know for sure which one to go with!

Thanks for all your help!

On Tue, Nov 2, 2021 at 8:33 PM Eric Audiffred @.***> wrote:

Cool. I made the mounting hole location changes and moved the power cord entrance hole. It looks like there is a lot of space in there now, but I know the cords take up room and so will the dongle. Like you said, I think there is space to mount the dongle vertically next to the board. Is the Tenda W522U what you plan to use, or was that an example?

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/Ribbit-Network/ribbit-network-frog-sensor/issues/18#issuecomment-958631746, or unsubscribe https://github.com/notifications/unsubscribe-auth/AATQ3FWV7WOBAL2AYHGCUQLUKC3QFANCNFSM5BYGC5CQ .

[keenanjohnson]

All right I've tested a bunch of Wifi Adapters and the only ones that seem to work reliably with the wifi hotspot are these ones from Panda Wireless (@dtischler and @mpous for reference).

Here are some dimensional pictures. I think the best approach might be to mount it like this with the antennae coming out of the bottom of the Frog again (see last picture)? What do you think @eaudiffred ? Perhaps we could design a clip into the enclosure to affix the wifi module in?

[eaudiffred]

I like that Idea. The antenna will plug the hole, so I don't think it will be too much of a concern.

Two other items:

1. I think getting rid of the cord grip makes things easier. A grommet can be used instead. Might not need anything.
2. Can a 90deg angle USB cord be used to save some space up top?

What else needs to fit in the enclosure? Is it just a few other wires running from the board out through the opening on top?

[keenanjohnson]

• Cord grip - Yeah I agree. it might actually be nice for there to be an opening in the bottom to allow any water that did somehow enter the enclosure to drain!
• USB Cable - Yes! I hadn't been able to find one, but yeah that would definitely save some vertical space!

Yes, I believe that will be it for now!

[eaudiffred]

Printed the lower shroud. After seeing it I decided to make some changes to the nail hook location. Moved it inward to eliminate the need for support structure and to get the nail closer to the center of mass for better balance...hopefully!

A few of the main Cura Settings Used: Head Temp: 220F Bed Temp: 85F Layer height: 0.2mm Wall thickness: 0.8mm Infill: 10% Print Speed: 50mm/sec

I think that increasing the wall thickness to 1.5mm and increasing the infill to ??% (need to test this one. Plan to try 50%) will be much better. It will make the walls completely solid, and give the nail hook and mounting blocks more material.

[eaudiffred]

@keenanjohnson How have you been printing the heads? The way I have it positioned in Cura right now I am going to require quite a lot of support material. This is about as lean as I can get it, and I'm not entirely sure it will work.

[keenanjohnson]

Yea I haven’t printed one of these in a bit, but that is basically how I did it previously.

I set the support to be only touching the building plate and the overhang angle to be like 80 degrees I think. It was quite a lot of support, but I couldn’t see any other way to print it.

I’m open to any ideas on changing the geometry ti make it easier?

On Sun, Nov 7, 2021 at 7:52 AM Eric Audiffred @.***> wrote:

@keenanjohnson https://github.com/keenanjohnson How have you been printing the heads? The way I have it positioned in Cura right now I am going to require quite a lot of support material. This is about as lean as I can get it, and I'm not entirely sure it will work.

[image: image] https://user-images.githubusercontent.com/88562047/140652134-e781ae8b-dbc5-4e0f-889c-30b7866e419d.png

[image: image] https://user-images.githubusercontent.com/88562047/140652103-b0546954-a92f-43e3-855d-bae88a05e4ac.png

— You are receiving this because you were mentioned.

Reply to this email directly, view it on GitHub https://github.com/Ribbit-Network/ribbit-network-frog-sensor/issues/18#issuecomment-962634763, or unsubscribe https://github.com/notifications/unsubscribe-auth/AATQ3FVMSDIN6NBZQGSPAWDUK2ODPANCNFSM5BYGC5CQ .

[eaudiffred]

I'll look into geometry options that are a little more printer friendly. The less supports I can use the less time and money is takes per part, so it seems worth the effort.

[keenanjohnson]

@all-contributors please add @eaudiffred for his heroic contributions here to the design, ideas, and testing.

[allcontributors[bot]]

@keenanjohnson

I've put up a pull request to add @eaudiffred! :tada:

[eaudiffred]

Quick update:

I printed out the base and lid. Defiantly makes a huge difference to have them in hand!

I think the alignment extrusions at the top of the lid are unnecessary. Even with the new style one of them broke off when I tried to remove the lid. The tolerance between the groove and the tooth was much tighter than I thought and kind of snapped into place. I think that's actually a good thing, it just makes it difficult to remove the lid. I need to think about that a little more.

I slimmed up the base a little bit to save on material, it was much heftier than I anticipated.

The lower shroud is in good shape.

I'm still working on ways to tweak the head to make it more printer friendly. No break throughs yet. There's pretty tight tolerances between the electronics on top and the inside of the head.

[keenanjohnson]

Sounds great @eaudiffred ! I really appreciate it! Let me know how I can support you!

[eaudiffred]

@keenanjohnson when would you like the enclosure design complete and ready for the first production runs?

I printed a head over the weekend and the total support material needed wasn't too bad. There are some tweaks I can make to the design to reduce the print time too. For now I think its best to leave the overall design of the head as is and print from the bottom up.

[keenanjohnson]

Awesome!

I think there are two things we need to test/finalize before starting to build sensors with this new enclosure, but we are super close!

• [x] Did you get a chance to test the screw attachment point? I think that was one last question mark on the general design from earlier?
• [x] I need to finalize getting the software to work on the new compute board and then do a test print here to make sure the electronics all fit in nicely. I'm still working this and will keep you posted as I finish it! #62

[eaudiffred]

See the pictures below of the hanger point. If the screw/nail is in the perfect position the whole unit will hang vertically. If it's a little loose the unit will tilt forward and lean against the post/wall. I don't think that's a huge problem. The only way to fix that would be to add an extended support to the bottom of the enclosure box. I don't think that's necessary. What do you think? I could try moving the mount point to the head, and then the lower shroud may help prop it up a little bit, but it all depends on how the nail/screw is located. I'm probably over thinking this, lol. It may be a non-issue as is.

[keenanjohnson]

Yeah I think that's fine. If it becomes a problem I'm sure we'll revise the enclosure later :)

We're getting close to validating the software on the beaglebone black, so I expect to be ready to commit and commit to the interior cad either today or in the next few days.

[eaudiffred]

Great! I have a little more work on the mounting (see picture), but should be able to complete that tonight or tomorrow. Then I think we are ready to go!

[keenanjohnson]

Awesome! Yeah let me know when you think it's ready and I'll try printing / building up the first new version here!

[eaudiffred]

Its ready! I'm printing a copy now. The two bolts that mount the shroud and head need to be a 25mm now, or stack a couple washers on a 30mm bolt.

Do you have the Panda wifi stick? The mount and hole cut out for that antenna is one thing that may need to be tweaked after a test print.