Protection of electronics which was flooded with seawater

Community General Discussion
21

to elaborate: the Swagelok solution is not really a form of connector. The bulkhead portion penetrates and seals to the inclosure while the tubing portion provides a sealed passageway for the cable from the surface electronics to the underwater unit. It is most practical for a system that is attached to some fixed object at each end.

Screenshot 2026-06-07 at 3.39.51 PM
Screenshot 2026-06-07 at 3.39.51 PM1150×692 51.9 KB

Here is an example of what they look like. It is not necessary to have tubing at both ends. They are easy to seal to the enclosure with washers or grommets and can be used with rigid or flexible tubing. Available for tubing that will accommodate a range of cable diameters.

Also, they have the additional advantage of allowing a small positive pressure in the enclosure.

22

How big of a PCB do you have? and How long do you need it to be underwater?

If they are small enough, you can use a balloon or condom, and worry about sealing up the neck.

23

I have used a lot of these in the past. It's a bulkhead fitting for passing tubing into enclosures etc.

I would be concerned about the water tightness of the fitting if used under water.
As you can see, it's a threaded body with a fixed flange nut and a retaining nut.

They aren't designed to be waterproof at the point where the body goes through the bulkhead or enclosure. It's difficult to get a watertight seal on a parallel thread. It would need something like an "O" ring on the flange side to achieve that.

Having said that, it could be a good way to pass a tube through a surface.

One thing to look out for is whether or not the fitting has a "stop" in the middle to butt the tube against. That would mean cutting the tube.

A little story. I worked a lot with compressed gases, including hydrogen. Hydrogen is notoriously leaky and will find ways out that other gases won't pass through.

No way will I let hydrogen into my property for heating, relying on traditional plumbing.

We had big problems with leaking, especially with fittings. Not one UK fitting would pass leakage testing. Then we found Swagelok, Whitey etc. Problem solved. Expensive, but the best fittings. Why Swagelok? Because the fitting have a 2-part ferrule, which is swaged or compressed onto the tube. When you tighten a Swagelok fitting, you use a distance gauge to get the correct gap between nut and fitting body.

The most reliable fittings ever, not one failed.

24

On the offshore platforms we used a Delrin washer on each side for a seal. The torque used on the fitting for a seal was important. Delrin doesn't "creep" at moderate torque and is very resistant to salt water and other chemicals.

For use as a cable "conduit", the tubing was only important on the seaward side so that is okay.

Swagelok, Whitey, was used a lot back in the pneumatic control days, where I first ran across it. That and for gas chromatography fittings. Good stuff but, yeah, spendy.

25

PCB itself is 15x17cm, it would be great if it can stay underwater all the time, except maintenance day when we go at the location and visually inspect everything

26

How will you know that water has leaked in?

27

How will you know => see post #14

28

That is one of your posts. I'm waiting for an answer from @vvb333007

29

Right now it is a construction of 2 mesh-like electrode, located really close to the wall of the enclosure, where cables enter and along the upper door of the enclosure (it is a box with one wall having []-ring). This is the locations where water can enter. Around 2mm distance from walls. They detect if there any electrical contact between them and the enclosure. But right now all it does is sending a message to the mainland. By the time people arrive it all dead , unrepairable

30

Do you need to have the atmosphere immediately around the electronics dry and room pressure?

I would consider using a polymer enclosure and filling the whole thing with castor oil. This way there won't be any pressire differential on the wires which may cause wicking.

That said castor oil is likely to have very different thermal properties so I'd look at getting any spare I/O inputs reading the temeratures of processors or other hot spots.

31

You could use humidity sensor instead, to get earlier alarm. Even fraction of a drop would raise RH noticeably (if initially filled with dryish air).
But if the leakage event is sudden it really doesn't help, so I would still focus on protection instead of detection.

32

Just from a logical standpoint:

The PCB is supposed to be protected from seawater at roughly +0.7 bar. The primary barrier can fail (and sooner or later it will - it's only a question of time), so a secondary barrier is added (gel, resin, whatever).

But if the consequence of a leak is still:

"By the time people arrive it is all dead, unrepairable"

then what is the actual purpose of the second barrier, especially considering the many disadvantages it introduces?

Is the secondary barrier intended to provide a specific survival time or service-life extension after a leak event? If not, I'm struggling to see how the added complexity is justified.

I'm trying to understand the engineering rationale here.

33

Both. If it can detect a leak and survive after that at max 24 hours - that would be just what we need.

The idea is to prolong the rime to react. It is just an idea, I am not saying it is the right way to go. I am asking for opinions :).

34

I was thinking that reconnecting powersupply in a special way will create a cathode protection , the aliminium case will start to dissolve while PCB tracks and vias will survive. When salt water touches PCB tracks it is a matter of hours even if it is powered off. When vias are corroded it becomes unrepairable.

35

Yes, gas chromatography was one of the areas I came across them.
With high pressure hydrogen lines around, you didn't want leaks.

If I was sitting on top of a Saturn 5 rocket, I would hope that Swagelok fittings were being used.

Some of the cheaper, often brass, fittings, were made of sintered brass rather than solid cast material. It was good enough for gases with larger molecules, but hydrogen will soon find a way out.

I was shocked one day when a brass bolt failed. Under a magnifier, you could see the minute sintered beads. The bolt was part of an earthing rod installation on a rural site with overhead power lines.

36

Although you don't seem to consider the solution from post #14 a viable option, even though it is an industrial-grade approach, there is another solution I have used successfully on small PCBs and sensors: full encapsulation (potting).

In this approach you don't really need a housing anymore. The critical detail is how the cable transitions are done.

The electrical connection between the PCB and the cable is not made directly. Instead, solid copper wires are used as adapters, and these are completely covered with leaded solder. Depending on the design requirements, the sequence can be:

  1. PCB → adapter wires → cable

or

  1. PCB → cable → adapter wires → cable

Cleanliness is absolutely critical. Flux residues, grease, fingerprints, or any contamination on the adapter wires must be removed thoroughly so the potting compound can bond properly.

The outgoing cable should also be embedded for some distance inside the resin, ideally in a serpentine path. This provides strain relief and creates a much longer sealing path. The cable jacket must also be perfectly clean where it is embedded.

The excellent sealing performance comes from the resin-to-adapter-wire bond. In addition, vacuum degassing is highly recommended. Either degas the resin before pouring, or preferably degas the entire assembly while the resin is still liquid. For this type of application I would consider the latter almost mandatory.

Of course, this approach only makes sense if heat dissipation is not a major concern, and serviceability is effectively lost.

From my perspective, there are two technically consistent options:

  1. A housing-based solution that is designed to reliably prevent water ingress.
  2. A fully encapsulated solution that accepts the environment and isolates the electronics from it.

Personally, I struggle with the intermediate approach, where multiple barriers are added but water ingress is still expected at some point and the electronics remain dependent on those barriers continuing to work.

In any case, those are my thoughts on it. Good luck with the project.

37

Won’t applying a voltage between the copper and aluminum cause harmful currents to flow through the electronics you are trying to protect.

38

@vvb333007
The best solution is just to prevent ingress of seawater as several have suggested. A semirigid HDPE conduit is a proven way to do that. You haven't described the installation sufficiently to indicate why that is not possible.

39

Well, my idea of using an ammo can is out. You could try finding a large Tupperware style container. Flip it upside down to create an air pocket.

40

Absolutely!
What we need:

Before asking for help, please try to provide as much relevant information as possible. The better the information, the more likely it is that someone can (and wants to) help.

This checklist is intentionally broad, but it is not complete. Depending on the problem, additional information may be required. If in doubt, provide more information rather than less.

Please make sure that a reader with no prior knowledge of your project can understand the situation without having to guess important details or search for missing information.

1. What?

  • What exactly is the problem?
  • What is the expected behavior?
  • What happens instead?
  • Is the problem permanent, intermittent, or random?
  • When did the problem start?

2. Where?

  • Where is the system used?

  • Indoor, outdoor, vehicle, industrial environment, underwater, etc.

  • Relevant environmental conditions:

    • Temperature
    • Humidity
    • Dust
    • Vibration
    • Pressure
    • Saltwater
    • Chemicals
    • Weather exposure

3. How?

  • How is the system built?
  • How are the components connected?
  • How is it powered?
  • How is it operated?
  • What has already been tested?
  • What were the results?

4. Hardware / Materials

  • Exact part numbers and model names

  • Manufacturer if relevant

  • Links to products or datasheets if available

  • Materials involved:

    • Metal
    • Plastic
    • Rubber
    • Wood
    • Composite materials
    • Other

5. Dimensions / Ratings

Provide numbers whenever possible:

  • Voltage
  • Current
  • Power
  • Pressure
  • Temperature
  • Speed
  • Torque
  • Weight
  • Dimensions
  • Distance
  • Load

Always include units.

6. Documentation

If available, provide:

  • Schematics
  • Wiring diagrams
  • Circuit diagrams
  • Datasheets
  • Relevant code snippets
  • Error messages (complete and unedited)
  • Log files

7. Visual Information

A picture is often worth more than a page of text.

Useful examples:

  • Photos
  • Screenshots
  • Sketches
  • Drawings
  • CAD models
  • Block diagrams

8. System Context

Help others understand the overall objective:

  • What is the device or system supposed to do?

  • What is the purpose of the project?

  • What requirements must be met?

  • What constraints exist?

    • Cost
    • Size
    • Weight
    • Power consumption
    • Reliability
    • Maintenance

9. Consequences of Failure

This is often one of the most important pieces of information.

What happens if the system fails?

Possible consequences include:

  • Minor inconvenience
  • Loss of functionality
  • Data loss
  • Equipment damage
  • Damage to other systems
  • Production downtime
  • Financial loss
  • Environmental damage
  • Safety hazards
  • Personal injury
  • Risk to human life

Also indicate:

  • How quickly must the failure be detected?
  • How quickly must corrective action be taken?
  • Is the failure recoverable?
  • Is maintenance or repair possible after failure?

10. Research Already Performed

  • What information have you already found?
  • What solutions have you already tried?
  • What worked?
  • What did not work?
  • Why were the attempted solutions rejected?

11. Additional Relevant Information

This checklist cannot cover every possible situation.

Please include any additional information that you believe may be relevant, even if it does not seem important at first glance.

What may appear insignificant to you could be the key detail for someone with experience in that specific field.

12. Community Knowledge

No checklist can anticipate every problem.

If you are responding to a request for help and believe important information is missing, please ask for it.

Likewise, if your expertise suggests that this checklist can be improved, refined, or expanded for a particular field or application, please do so.

The goal is not to complete a checklist. The goal is to provide enough information for others to understand the problem and help effectively.

Remember:

People can only work with the information provided. Missing information leads to assumptions, unnecessary questions, and often the wrong answers.

The clearer the problem description, the better the chances of receiving useful help.