pond diy PoE camera housing issues

[brokensword]

Hello,



I have  a pond and once had an underwater PoE camera hooked up.  Well, it developed two problems in the year and a half and then the camera died.  I don't want to keep replacing the unit as it's not cheap, so I've been working on a 'reusable' underwater housing where I can put in inexpensive PoE cameras as they break, which seems likely.  In working out the design, I'm using pvc, glass for a front lens, and a customized endcap with rubber o ring + latches.  Now, I'm using flex pvc (3/4") as a shield tube at that end port as I need to get the ethernet cord from the cam and up to the surface where it connects with a coupler and then to the feed to the PoE injector.


The camera will be in a freshwater pond where water temps will range from 35F to 85F, approx 20" down.  The camera is a Reolink 5mp security PoE cam which gets power from a PoE injector, then on to the computer.  Since ruining the original ethernet connector (looks like water got in), I've now got a waterk-proof taped solid lead from cam out of the housing and to the coupling above water.


I've had multiple issues with leaking and couldn't understand until maybe, just yesterday, had a thought; I'd put silicone on every seam, even the ones where I glued pvc to pvc and where the flex pvc attaches, then siliconed the end of the plex pvc to make it air tight.  (I'm aware silicone isn't a good 'stick' for pvc but have it so the outside water pressure should push against any silicone and force it tighter.  If  I don't mess with it, it stays attached and functionable.   The one time I DID mess with it, I got leakage).  I still got water, not a lot, but apparently enough to ruin an ethernet cord end.  Fixing this, I then mounted the cam back in the case and let it do it's thing without any silicone to cap off the flex pvc end, effectively allowing air to get in and out.  Yesterday, I was going to rework the silicone and when I opened the diy case, I had a teaspoon of water come out.  Now, this is OUT of the water.  Sooo, long story short, I think the cam is taking humidity out of the air and it's pooling up inside the case.  I expect this may account for my earlier water inside, thinking then it was from faulty joints/sealing.  I DO have a packet of silicone absorbent granules inside, just for the purpose of taking out any humidity.  I assumed once gone, the air would then NOT deposit out and ruin the cam/cam cord connection.


So question for anyone having dealt with moisture in this sense, how do I insure I won't get such a buildup of water?  Do I have to manually extract the air right after gluing/latching?  I've heard to put in CO2/Nitrogen but I don't have easy access to this and am just looking for some sort of 'trick' that will keep the camera inside dry for years (or yearly when it dies on me, again!).  I need to be able to get in and out of this housing, hence why I'm trying to overcome the problem.  I know I could seal it up permanently but I might as well just go the expensive UW cam route in that case.
I've also wondered if this ethernet cable (supposedly exterior/bury usage) is letting water in through it's jacket and if that is something I need address.
Any ideas?  I know it's not a typical subject--I've checked and googled ad nauseum--but any insights to either my build or my current issue would be much appreciated!  


Thank you in advance, esp for anyone who read all the way through!



 

[TugboatEng]

Plastics and rubbers are permeable, some more than others. Silicone is the worst for permeability.

You also have very high amounts of differential thermal expansion between plastics and glass. Consider making it all out of one material such as acrylic or make it out of metal and glass and use resilient seals.

Add a dessicant pack to deal with small amounts of moisture or add a snorkel with a Gore-Tex vent.

Use a space heater in the enclosure.

 

[brokensword]

Thank you Tugboat, for all the advice. I'm going to pick your brain further and elaborate at the same time.

I'm trying for something inexpensive that can be reusable. I don't have access/ability to making anything out of pure acrylic/plastic/steel, but wish I did, though I'd still have the issue of how to seal up the ethernet cord that must exit the pond and be connected on dry land.

I've had aquariums for a long time, so know that the water pushes against the glass seal, keeping it secure. I see what I'm doing as similar, no? The water will be pushing on my diy housing (approx 4" pvc tube, one end with glass siliconed against the pvc ridge and the other with a cleanout that has been altered for latches that pin/crimp the rubber O ring between. I have flex pvc as an exit point. All pvc has been glued using pvc solvent, including the flex. The end of the flex exits the pond above the water surface and will be filled with more silicone to keep air inside stable.

So, are you saying water is getting THROUGH the silicone points of sealing? And you say plastics are permeable too; so water can get past the pvc casing? I DO have a pack (3" x 1") of dessicant inside with the camera. I'll research the snorkel with gortex vent idea, thank you.

I don't see the option of a space heater, especially underwater; just more cords to seal if I even could fit some sort of small heating device in there. The camera should provide heat, in any event.

Any further thoughts such that I can make something of this? I've seen vids showing underwater casings made this way that were successful, except they did not have the need for a cord coming out and up, just a camera housed inside. That would be easy. I feel confident the pvc solvent seals are working, wonder a bit about the glass to pvc, but it's butted with silicone against the ride of a 4" pvc coupling and then siliconed on the outside, where I expect water pushes against it. And I know there's push because I have to weight it down just to keep it from floating. It's at the ethernet-inside-the-flex and/or condensation that I'm fighting, I think.

Appreciate all your time, very much!

 

[TugboatEng]

These are the vents from Gore. They are available individually for electrical enclosures.

I suggest moving away from PVC, however. It is difficult to bond by methods other than chemical melt.

I am not knowledgeable on this subject, I just know of a few of the complications.

PVC isn't an inexpensive plastic. You must be using something commercially available do I assume it's pipe. Consider using ABS pipe and polyurethane sealants as they should (I haven't double checked this) adhere well to glass and ABS. Acrylic adhesive are generally the best for bonding a wide range of materials but polyurethane is more flexible and may be more tolerant with the differences in expansion.

This part is important: your adhesive needs to be in a thick enough layer to allow for that expansion. If you set the glass tight on to the plastic there will be no ability to absorb displacement of the materials. You need to leave a gap and then fill it with adhesive. This is true for your rubber o-ring seals as well. You're trying to make a bellows.

 

[brokensword]

Tugboat; thanks for following up! I had researched ABS and was going to go that way but not nearly as easy to find parts in the sizes I need as moderate prices. The idea here is to make something inexpensive so I can replace cameras as they die. If it's going to get costly I might as well save myself headache and time and just get the commercial camera again. I just never thought it would be this frustrating to keep something from leaking water IN. I've done enough plumbing projects and with success at keeping water,under pressure too, from getting out. I can't see much pressure from 20" down surrounding the 4" pvc tube. I DID leave a bit of room for the glass but not exactly sure how much. It's not tight, that's for sure. Probably 1/16" all the way around as the gap. I figure the heat from the camera is going to offset some of the pond water cooling temps during the winter and in the summer, the water again will ameliorate the temp to its surrounding. Was more worried about the winter, though. Haven't got it out in the pond yet to test though, still working on a tub of water in the basement as my test site.

Thanks again for all the help. Working on another test with yours and others' advice to see if I can stop the minimal but undesirable leak.

 

[SwinnyGG]

Simple changes will make a big difference.

-Use a piece of acrylic for the window instead of a piece of glass. You'll get less differential in CTE between the window and the housing, which will make your seals more reliable.

-When you say 'silicone' I'm assuming you mean household caulking which you probably got from home depot or whatever. Those materials are not intended to be fully waterproof under long-term immersion, and they will never handle that exposure very well no matter how you detail your joints. Ditch the household products. Sika makes a wide range of sealants for many applications, including many designed specifically to handle constant immersion.

-For your cord exit, use a cord grip. You'll never get long term durability with a length of flexible PVC filled with sealant. Honestly your best option for durability would be a solid PVC stub to get your cord above water, with a cord grip threaded in the end to close off the opening and pass the cord. This would allow your joints to the housing to be glued, and very strong.

This stuff is more expensive than your housing v1.0, but if you're only making 1 it's not a big cost overall. A good quality industrial cord grip for a single cat6 is like $4. A tube of Sikaflex-11FC and a bottle of primer is maybe $15. You're not talking hundreds of dollars to make something robust.

 

[brokensword]

Thanks for replying, SwinnyGG; I appreciate the help!

You're right on the silicone. And, I DID have a solid pvc piece up to the surface but it was unweildy at 30" (max depth I'd use it), so I switched to the flexible version as I also have to 'hide' this exit tube after breaking the surface. I don't think there's a difference for this purpose as I've been using flex pvc for over 10 years, and never had leaks, so I figure it should keep water out without any pressure inside, as well. I guess time will tell about that as I just don't think the solid was going to work. That said, if I get more leakage for this test and the next, I'll switch it out.

Since I posted this problem, I've also had scrutiny re my main seal method (O ring around a pvc piece, sandwiched top and bottom by lip of the two pvc pieces forced to close using clasps. I saw this method via youtube and tried it, but I feel the clasps were not strong enough to compress the O ring properly, or the O ring is not flexible enough). I changed this out to have two pvc flanges compress a flat rubber gasket between them, hed by 8 stainless steel bolts. I just put it under water today after also fixing the pvc cord exit end with new pvc and solid glue joints, including the flex pvc piece once again. If I get water, I'll start then to inspect the lens, but I've inverted the unit to see if any air escaped out the lens side (which I believe it would if glass side up--correct me if wrong). I'm using glass instead of plexi as it won't scratch as easily because algae will grow on it and I have to regularly clean for a clear view.

I'll definitely look at the cord grip you suggested as I didn't know it even existed. And I've heard of the Sika; will get some of that too, though my current design has silicone only at the glass lens; inside and out. I DID get some locktite marine glue, but haven't used that yet. I'm hoping I don't need to.

[edit] will this product do the job for a cat5E cable? From amazon;

https://www.amazon.com/gp/product/B08ZRQKVLG/ref=ox_sc_act_image_1?smid=A2S46TP95HIG1&psc=1

And I'm assuming I have to cut the end and slip this over, then re-crimp a new ethernet cord end.

Let me know if anything else looks amiss to you with this new design and other better suggestions. I'll know in short order if I still have issues. This test has no cam inside, the next will.

Thank you, again!

 

[SwinnyGG]

What's critical with regard to the cord grip is that you measure the actual diameter of the cable you're using, and buy a cord grip that's the correct size. Yes, you will have to cut the cable and crimp on a new connector.

I'd suggest shopping via McMaster as opposed to Amazon. You're going to get real quality components. Amazon for industrial parts is basically whatever offshore manufacturer makes them the cheapest. The cost difference might be a dollar.

When you say 'I'll get some Sika'... Sika makes probably 1,000 different product for wildly different applications. Make sure you get the right one. Immersion and what materials you need to bond are critical. High quality products, correctly applied, used in the wrong application will still fail. And whatever product you select, do not buy it on amazon. It is impossible to know if you're getting an actual sika product or something counterfeit.

 

[brokensword]

okay, will do. I've bought from McMaster before, so am familiar. I'm assuming they'll have the Sika product as well. Thanks!

[edit] I went looking and it was fairly confusing. I calculated I need something to hold .215" cord, but am having issues with exactly which unit to use. They all talk about 'knockouts' and since I'm putting this on a pvc pipe, I can't even figure out how they attach, the ones that come up at McMaster-Carr. Could you shoot me the link to the exact one you'd use with a 3/4" pvc pipe? Or let me know I'll have to use bushings/adaptors to get down to whatever size cord grip you're suggesting? Thanks! And since this will be above the water, IP67 should suffice, no? That said, I DO have an underwater cable from a dead camera I could use and that might be an option as then, I could eliminate the pvc 'stack' pipe to the surface and mount it directly to the pvc exit on the housing. Do you think this would be a better idea? That said, I'd still need the correct cord grip and how to mount it to the same pvc piece, no matter what size it ends up being.