Water Tight Wire Harness

Any chance you could post a photo so we can get a better idea of the problem area?

These guys should be able to help. They supply the best water tight connectors in the world.

http://www.weatherpack.com/

I have used the liquid tape successfully for connections on my boat and other home uses. You may want to contact these people directly for info on your specific application needs.

http://www.plastidip.com/

I ran the maint dept. at this place once upon a time. They do great custom harnesses and mold their own stuff. Just in case you decide to outsource. (usual disclaimer)

Oops, Forgot the Hyperlink!

If there is no pressure differential across the location and you only wish to stop capilliary action, automotive companies (Read this as GM and Ford) have used passive grease in rear lamp and vehicle exterior wiring harness connections for decades.

Cheap and effective and repeatable process that doesn't require "technique" or significant capital expenditure for potting compounds and fixtures.

Careful selection though, some greases have issues with solder joints.

In some cases it can be more important to give the water a way to escape from the enclousure, than to try to exclude it completely :D

If you are trying to prevent "wicking" you may be able to pull a vacuum on one end of the harness to draw in a waterproofing compound.

Some images, as several have suggested, would be helpful.

I thank you all for your suggestions, I am not ignoring. I had several fires break out after I posted and am tending to them. I am inserting a picture, these ends are inserted in terminal blocks on the PCB and are the source of the leak, we do use many plugs on the outside end, weatherpack included. While plugged in the leak is controlled, and to be honest I am not sure this is the problem we are making it out to be, but I wanted information and I know the CR4 board is a great place to get information.

Is it that water is getting between the (conductive) core of each wire, and it's insulation? I'm still trying to understand what the actual problem is, here.

[Edit: Think this is what Unredundant was getting at]

I have seen occasions where the water passed through the "inside" of the wire. Passing through the strands inside the insulation. This could also be possible as you have suggested.

Yes thats exactly right, the water is passing through the core of the wire. Our testing procedure was very simply to blow nitrogen through the wires, 98% of the time the wires passed. Upon further review after an incidental finding of a leak we have determined that the tinning of the wires work to seal the wires until they are manipulated to be installed in the PCBA, once they are curved to fit the seal is broken allowing a leak path.

Think you'd be best to look into the wire suppliers. Some of the stuff I've used from e.g. Igus and Leutze seem to have the insulation material very closely moulded over the conductors. I haven't looked into water permeability, but the suppliers should be able to advise.

Another alternative would be potting the wire ends, with the compound sealing the interstices between the insulation and the cores, and between the strands of the cores, but this is starting to sound messy.

Keep us updated.

[Edit: I think the tinning is just melting the insulation locally, so that it makes a reasonable seal until flexing breaks the seal.]