Environmental Resistance - nolanhergert/EasyLED GitHub Wiki

Case / Enclosure

Requirements

Needs to be robust to dropping from 6 feet with the weight of a helmet. Ideally much heavier too, but that's a lot to ask.
Ideally something that people can 3D print. But for max strength and durability and surface finish, I probably want to injection mold it, so I need to be able to create a "production" mold easily.
Don't need to take it apart, generally. However, I don't necessarily want to make it difficult to modify or repair (clean up with flux), or replace. Gluing is out.
Don't really care about no screw look at this point. Don't need side snaps.
Needs to not vibrate (have play). Hard to get to this point with snap fits? Not sure.
Easy to replace case if needed.

Notes v2

What about sealing around the USB cables?
- Can't use typical cable glands, as the USB micro connector is wider than the cable
- Can use injection molding to overmold a gasket directly on the cable, but is not as flexible to moving around.
- Can I get enough squish on the side by undersizing the hole and making it push out? I think so
- How to seal unused holes?
What durometer do I need to make a good seal on 3d printed parts?
- Need enough tension to make a squish. Can use bolts and nuts as side "tension holders" that require the user to press down first and then insert the bolts as holding the tension. Or try for snap fit or a project box snap, they seem really bulky though and still require bolts.
  - https://www.thingiverse.com/thing:4555276 is an example waterproof case with snap fit-ish clips that I don't really like. Easy to open though.
- Can test hot glue, acetic acid cure silicone, and non acetic acid cure silicone. Need to make gasket flat using a dragger ala https://youtu.be/76KV98fSzF8?t=110

Notes v1

https://www.reddit.com/r/product_design/comments/7tb9b0/screw_vs_snap_fit_vs_adhesives_vs_heat_staking/

Making a snap fit is hard. I don't want it to be too tough to remove. But it needs to be strong and last.

https://www.youtube.com/watch?v=SvI8qrdAr1w
An easier one? Sliding case (will work for the back side of PCB, where there are no parts). OpenSCAD. https://www.thingiverse.com/thing:201304
- However, not sure why the snap is at the front instead of the back...
- Also, impossible to injection mold.
Nice ideas on thingiverse though. "snap fit box"
This one looks good! Vertical lid w/ snaps. https://www.thingiverse.com/thing:2866563. Would like a coin side opening though. Test 3D print at robotics?

I'm really feeling screws at this point.

USB Micro dimensions

Page 20 of this USB spec doc: https://raw.githubusercontent.com/hellamps/Kindle3uart/master/non-free/docs/Micro-USB_1_01.pdf

Electrostatic Discharge

The ESP32 module ESP32WROOM32 or whatever, seems to test for ESD, and has a TVS already built in, at least for the 3.3V and GND rails. Awesome! Hopefully I can test it... https://www.espressif.com/sites/default/files/documentation/esp32-wroom-32_datasheet_en.pdf

Article explaining TVS thing: https://blog.upverter.com/2019/11/29/adding-pcb-esd-protection-to-your-design/

Part is $.05 on JLCPCB, but probably need one for each line?

Water Testing (No thermal cycling)

Bare 2-LED strip lasted several days, but curiously wore out on the 5V wire and didn't break on the PCB anywhere (despite wear). Maybe the solder joints last longer than bare copper?
With one q-tip of vinegar dipped into a small container of water (about .1% overall acetic acid concentration), the wire for the bare LED strip wore out within hours.

Tap Water Tests

Material	Tap Water
Bare	Lasts 3 days, infrequent bubbles
UV cure "resin" (cyanoacrylate style)	Etches quickly, visible frequent bubbles
Super Dry clear nail polish (dot 1)	Frequent bubbles
Fastdry wet n wild clear nail polish (dots 1 and 2)	Infrequent bubbles, but flickering by end of day
Acetic acid curing silicone enveloping bare pcb	Very infrequent bubbles, but only on very tip, which was uncovered. No issues
Silicone sleeve with acetoxy silicone as gasket	No bubbles, no issues
IP65 LED strip with hot glue gasket	No bubbles, cloudiness due to trapped water vapor. But resolved when dried

Flickering: https://www.youtube.com/watch?v=RjfsMdw_7rw

Silicone offgassed within 1 day uncovered by a silicone sleeve, two days with a sleeve.

Permeability

Is immersion in liquid water the same effect as a cloud of water vapor? I would think so, since vapor can escape from the water into the object, right?
Can dissolved salt go through silicone?

Thermal cycling

Changing the temperature of air causes the internal pressure to vary widely! I think it's inevitable that water is going to get inside...
Some bonding materials are not flexible, so the expansion and contraction can lead to delamination from the surface.

Do I have to test all the variables at once? How many cycles?

If only 10s of cycles are needed, maybe I can just manually move them back and forth between a hot, humid chamber and the fridge/freezer. Sounds good to me!

Youtube Comment

All great points. Nice hack that you're using a combination of both.

Because of this conversation already, I am throwing together a temperature/humidity/salt fog testing chamber to see what I find out...some principles:

Silicone is not liquid permeable but very vapor-permeable. While salt might not be able to evaporate off a board over time, maybe various types of acids can since we can smell the acid when it's curing? (your anecdote of a cap failing with a permatex joint is contradicting this). I think I also need to wash my PCBs since the flux can have acid too.
- I've noticed the IP67 rolls I receive take pains to not have the silicone touch the board (PCB is an inch inside from the side of the sheath). Seems like a good compromise that might work.
Why is silicone used for IP67 strips in the first place? What do you think about the "IP65" strips that don't use silicone and are much easier to adhere to? I feel like they will yellow over time, but it might be worth the tradeoff? Water intrusion might be a problem too.
My application is completely outdoors (bicycle helmet), but just needs to be rain resistant. My current design is using cheap snap fit cases, but they'll probably have gaps for water vapor to get in. What is a good long-term conformal coating for the PCB? Hoping that nail polish or a cheaper equivalent will work, but I'm skeptical.

I have a few more questions if you're interested. We can call if you'd like. ([email protected])

It seems like you are using inter-strip clamping connectors right (no soldering)? How are you waterproofing them? Hmm...maybe marine heat shrink?
Have you tried open-to-water-vapor connectors, something like USB? My prototype USB power pack has done fine over two years with an "exposed" USB connection, other than a bit of rust in places. But then again I don't leave it out in the rain either. Curious as to your opinion.