Water ingress into potted electronics via the cabling

[DTJ]

I've got some issues with water ingress into some potted electronics.

The immersion depth is only 30cm or so.

The electronics is potted in epoxy with the cable sheath and then the insulated cable conductors inside the epoxy.

I trust the epoxy as it has been used hundreds of times in a very similar application.

I'm still in the early stages of this but I suspect the water is getting in between the epoxy and the outside of the cable insulation.


Has anyone come up with a simple effective means to fix this?

I know fancy ozone (or something like that) etching of the cable insulation can be done to improve the epoxy bond but I'm hoping there's a simpler primer / sealer that can be used.


Cheers.

[mzzj]

This would really depend on the cable sheath material. PTFE vs PVC needs totally different approach.  Epoxy in general bonds poorly to plastics.

[redkitedesign]

Water ingress in submerged assemblies is really hard to prevent, especially along openings in the assembly.
Whenever possible, the easiest way to deal with it is to let it happen and remove the water in an other way (pump it out). Its what boats do :-)

However, these people have posted some experiences you might find usefull:
https://thecavepearlproject.org/2023/03/17/waterproofing-your-electronics-project/

[PwrElectronics]

Even if the cable jacket is not the source of leak, fluids can wick on the inside along the wire as well.  So, the other end of the cable must also be sealed.

[jonpaul]

we use Gardner Bender conduit putty.

Also RTV 100% Silicon rubber calk

j

[2N3055]

Even if the cable jacket is not the source of leak, fluids can wick on the inside along the wire as well.  So, the other end of the cable must also be sealed.

Yep.
I had water ingress into sealed sensor box through 10m length of cable via capillary action because box on the other side was damaged and water went in... 

[John B]

If you've ever used untinned copper wire outdoors, you'll probably have seen the creeping green death inside stranded wire.

I would think you'd need a solid core wire bulkhead on the enclosure to ensure water can't penetrate into the enclosure.

[MisterHeadache]

I work in the OEM auto sensor industry. There is no economical adhesive method or material that makes a reliable bond to low surface energy materials, which most wire insulation and jackets are made from.  So the common solution is to use a silicone rubber gasket, either an individual one for each insulated wire (a sealing 'donut') or a single-piece with holes for each wire (a 'manifold').  Look under hood in the engine compartment - all the sealed connectors use one of these approaches.

[ajb]

So the cable jacket is embedded in the encapsulant? Could you fit a sealing bulkhead gland to the box before potting? Or address the issue at the other end of the cable?

I work in the OEM auto sensor industry. There is no economical adhesive method or material that makes a reliable bond to low surface energy materials, which most wire insulation and jackets are made from. 

I've been looking at automated wire processing equipment lately, and was initially surprised how big a deal seal handling was.  Makes a lot of sense for the automotive market!

[DTJ]

I work in the OEM auto sensor industry. There is no economical adhesive method or material that makes a reliable bond to low surface energy materials, which most wire insulation and jackets are made from.  So the common solution is to use a silicone rubber gasket, either an individual one for each insulated wire (a sealing 'donut') or a single-piece with holes for each wire (a 'manifold').  Look under hood in the engine compartment - all the sealed connectors use one of these approaches.

Thanks - that's interesting.

I doubt I'll find an off the shelf seal to suit. Would a single small tight fitting O ring on each of the wires work?

I assume this works by the epoxy bonding to the O ring with the O ring squeeze on the wire providing the real seal?

Cheers

[DTJ]

So the cable jacket is embedded in the encapsulant? Could you fit a sealing bulkhead gland to the box before potting? Or address the issue at the other end of the cable?

I work in the OEM auto sensor industry. There is no economical adhesive method or material that makes a reliable bond to low surface energy materials, which most wire insulation and jackets are made from. 

I've been looking at automated wire processing equipment lately, and was initially surprised how big a deal seal handling was.  Makes a lot of sense for the automotive market!

Yep. The sheath penetrates about 15mm then there's about 20mm of wire where the cable sheath is stripped.

There's no opportunity for a gland or anything like that.

[dmendesf]

 Seems that a cold plasma pen can activate surfaces and improve bonding. Problem is that they are expensive. Someday I will try to make one.

[tooki]

If cost is no factor, how about commercially available vacuum feedthroughs? For high vacuum and ultra high vacuum applications, there are companies that make feedthroughs — metal flanges with integrated wires. I have no idea whether these would work to keep water out at pressure, but I imagine that water pressures can be significantly larger. But ultra high vacuum requires essentially no leaking no matter what, so maybe it’s robust enough anyway.

Or maybe there are equivalent manufacturers of submarine feedthroughs.

[AndyBeez]

I have no idea about the mechanical design of your enclosure but, one extra barrier to ingress is to place the component inside chandler's or maritime grease. This is the salt water resistant and repellant grease as used on prop shafts. It's sometimes branded as "stern grease" or "K99". About $20 a kilo.

[TZ400T3B]

Search for dielectric gel used to secure cable connections in boxes buried in ground. It sells in tubes same as silicone. It adheres to everything and is rubbery, exactly opposite to very unsuitable epoxy. Long ago it was plain asphalt used for this, heated and poured over. Worked too but the gel is much better.

[dmills]

This is a well known issue in the sonar industry, but I am a little surprised to see it at such shallow depths.

The cure is generally something like:

Encapsulate the end of the jacket using something like a Pur resin that will bond properly to both the jacket AND the insulation of the individual conductors, this may involve some chemical/mechanical fuckery if PTFE wiring is in play.

Then encapsulate the ends of the individual conductors, we used to do this by designing the boards with a solder pad and strain relied hole, conductor end is prepared, bought thru the strain relief hole, soldered to the terminal, inspected and then the end of the insulation encapsulated with a drop of epoxy resin.

Finally pot the whole thing in something that will bond to the PUR encapsulation and the epoxy.

Yea, it is a faff, but particularly cables not made for underwater service tend to leak. 

[DTJ]

This is a well known issue in the sonar industry, but I am a little surprised to see it at such shallow depths.

The cure is generally something like:

Encapsulate the end of the jacket using something like a Pur resin that will bond properly to both the jacket AND the insulation of the individual conductors, this may involve some chemical/mechanical fuckery if PTFE wiring is in play.

Then encapsulate the ends of the individual conductors, we used to do this by designing the boards with a solder pad and strain relied hole, conductor end is prepared, bought thru the strain relief hole, soldered to the terminal, inspected and then the end of the insulation encapsulated with a drop of epoxy resin.

Finally pot the whole thing in something that will bond to the PUR encapsulation and the epoxy.

Yea, it is a faff, but particularly cables not made for underwater service tend to leak.

I too cannot understand why its an issue in such low pressure situations.

I like that approach, I've been searching to find a compound that will bond well to PVC cable sheath AND to epoxy. Then I could paint a little onto the wire insulation and it would be ready to pot.

[thermistor-guy]

I've got some issues with water ingress into some potted electronics.

The immersion depth is only 30cm or so.

The electronics is potted in epoxy with the cable sheath and then the insulated cable conductors inside the epoxy.

I trust the epoxy as it has been used hundreds of times in a very similar application.
...
Cheers.

Has the same cable been used to the same extent, in a similar application? Has the cable changed? Has something else
in the assembly or assembly process changed?

Two thoughts come to mind re the cable. I don't know if they are relevant here.

First, there is a phenomenon in watercooling PCs, where water gradually diffuses (permeates) out through some types of
tubing. To the uninitiated, it looks like a problem with seals and couplings. But no, it's just a property of the tubing material.

Second, it reminds me of a problem I've seen in a ship-to-shore power cable, carrying 100s of amps. The cable jacket was
supposed to be impermeable to water, but it wasn't. The problem was not in the synthetic rubber material itself, but in the catalyst
used to cure it.

The catalyst was hydrophilic, and there were small amounts of it left in the cable jacket, enough to compromise insulation resistance
over time. Nowhere was this catalyst property mentioned in the cable data sheet.

[ajb]

I too cannot understand why its an issue in such low pressure situations.

Is there perhaps some air space enclosed by the encapsulation?  As the temperature of an enclosed air volume changes so does its pressure, which could contribute to the problem.  The air space wouldn't necessarily need to be connected to the cable entry if the encapsulant is soft enough.

[redkitedesign]

I too cannot understand why its an issue in such low pressure situations.

Given enough time, water will (slowly) leak through every material discontinuity. Tightening the seal will only limit the rate. Increasing the water pressure will increase the rate.

Air inside the product will expand/contract with temperature changed. That will give pressure changes that can influence leakage.

And very few materials are really impenetrable for water. I think only metals and glass are really impenetrable, the rest are just more-or-less leaky. It's just that above the water, the leakage is small compared to evaporation, so it is never an issue.

[dmendesf]

Some pointers about vacuum seals... Just remember how tubes are made:

https://en.wikipedia.org/wiki/Glass-to-metal_seal

I can find a bunch for RF:

https://www.alibaba.com/product-detail/Glass-to-Metal-Hermetic-Seals-RF_60807284004.html

But nothing cheap related to water. I'm also interested in products in this field. Maybe it's time to melt some glass...