Webcam troubles

So I set up a birdcam outside in early January and by the end of January it had failed.

Investigation showed that the ethernet connector had water in it and had completely corroded away the connectors.

To their credit, Amcrest (/Foscam) was a star and replaced the unit.

I think the problem is that they give NO instructions on how to assemble the cable gland that makes the end waterproof, and I put it together incorrectly. I have looked online for instructions and found enough similar ones that I now know how to put it together.

Today I put new ends on the cable and hooked up the new camera. Then a few hours later I checked and the cable gland was HOT. When I opened it up, it was actually steaming. In this photo you can see steam/smoke and condensation within the connector:

The ethernet plug was so hot it had carbon tracks burned in it and it was actually sparking.

Clearly a lot of water got into the cable, and when it warmed up a bit, it migrated down inside the sheath, condensed in the connector, started to conduct excess current, lather/rinse/repeat, runaway thermal event.

I cleaned the camera end and blew most of the water out with compressed air. Since I can't guarantee that water won't enter the outdoor cable through a critter bite or something, I've decided to make the last 2 feet of the run with a short ethernet jumper (with cable that's never been outside) and connect the two with a female-female waterproof gland (way cheaper than a camera) so that if this happens again, it'll smoke a $5 Chinese cable gland rather than the connector on the camera which is NOT replaceable or repairable.

I'm also running it indoors for a day or so with no cover on the gland to hopefully chase the moisture out. I cut 3 feet off the end of the cable before putting a new end on.

This is the gland I wound up with from eBay/China for < $6 each. They're quite good. I'm going to get a few more coming (they take the standard 3 or 4 weeks from China) You can put them on over the top of an existing RJ45 connector.

The other fun problem I have encountered is interior air migrating into the cable, and then condensing and running downhill to the device, or a low point in the cable. It wouldn't seem like a lot of air could get in, but it also doesn't take a lot of water to find a defect in the twisted pair insulation, or find it's way into the device...