Potted Pass-Through


#1

There are many instances when you may want to pass electrical wires (such as a tether, motor wires, communication lines between compartments, etc) through a waterproof barrier. A cheap, easy, and effective way to do this is with a potted pass-through. The basic idea of a potted pass through is that an initially fluid material (such as epoxy or potting compound) can be poured around electrical wires that are going through a tube. When the material hardens, it forms a wall that water can't penetrate. By attaching the potted tube to a second (dry) tube using vinyl hose material, the connection can be removed easily without having to destroy any parts or cut any wires. This diagram and photo should give you a better idea of how everything goes together. Note that in order to keep the pass-through from leaking, you should sand and clean the outside of the wires and inside of the tube so that the potting material bonds well with both. Many compounds will work well, but my personal favorite (due to having just the right amount of softness) is Loctite Hysol E-90 FL.


Trick 1, Figure 1: Diagram of Potted Pass-Through


Trick 1, Figure 2: Potted Pass-Through being implemented on ROV1 for motor power wires.


#2

Are there any resins you would recommend to encapsulate circuits in? I’m working on some underwater lights and experimenting with uv resistant resin.


#3

I'd recommend Hysol U-09FL if it needs to be clear, otherwise E-90FL is the way to go!

E


#4

That's fine, however one yank on the wire from an underwater tangle and you have broken the seal and are rewarded with a leaky dry-space.

Instead, use 1/16" brass all-thread rod (from the hobby shop) and cut a piece slightly longer than the thickness of the hull or endcap you wish to penetrate. Tap a threaded hole for each wire the size of the all-thread you wish to go through your hull and screw your allthread through leaving a bit on either side of the hull. Put a washer and two nuts on each end of the all thread and you have a nice, tough, waterproof through hull for your electrics! If you're worried about corrosion, a bloop of silicone sealer will keep the grey fur off your connections.


#5

The best solution is using a cable glad as shows in the picture, using brass on acrylic is a very very bad idea as the materials expand and contract at very different rates thus the acrylic will crack and damage your pressure housing, we often use cable glands for cheap effective penetrations on survey equipment or lighting etc, its far better than brass rod.


As you can see from the pic above these lights are set up for 500Mtr depth rating using a cable gland the rear of this gland is potted with Devcon 90 and using the appropriate tap for the thread will allow a leak free and tidy penetration.


#6

Tygon UV corrected (yellow hose) is the best this is used on oil filled commercial ROV cables no other works as well


#7

Reading the assembly wiki I was wondering how you make the epoxy work as a perfekt seal. For me, in this picture http://wiki.openrov.com/images/thumb/b/bc/Potting_25.png/500px-Potting_25.png, it looks like water could travel through the yellow netted wire sheathing into the dry area. Exactly how do you do to make the epoxy enclose all the wires so no water can penetrate the passthrough?


#8

The way we had it in the photo required us to wiggle everything around quit a bit to get the epoxy to flow into all the nooks and crannies between the yellow sheathing and the wires. In retrospect, it's better to cut the sheathing so that it does not follow the wires through the potted pass through. E-90 is very inviscus and cures slowly, so it's pretty good at filling small spaces.

E


#9

Best method is to apply a vacuum and the epoxy with flow perfectly no air bubbles, we do this with all our new potting


#10

For those of us relatively new to the world of DIY, how does one do this?


#11

Simple way is to use a large plastic seal-able box bought from any hardware store, then setup the cable end with a gland so that this can pass though the box and save you space, another entry into the sealed cavity will be done with a brass or steel 1/4 turn fitting and if you like a vacuum gauge, a shop vac with a suction setting and place on 1/4 turn and such down for 5 mins and your done.


#12

Hey guys - I just bought two "seal-a-meal" vacuum sealers from a thrift store for 10$ each last week. so far I have used them to vacuum bag a few small fiberglass parts. there is a vacuum port on the top for use with a tube :

like here on instructables .

i was fixing a surfboard fin for a buddy and didn't want airbubbles in the laminate, so i laid up 3 layers of light cloth and bagged it with a seal a meal bag.
I think you could use the tube and a jar to degasify the epoxy, then carefully pour it into your coupling.


#13

It's also important to address each wire that isn't fully insulated. The beldin wire we use has a jacket that could act as a hose for water to enter.

I made a small incision around the housing and made sure the epoxy made it inside the jacket.


#14

These vacuum ideas and set ups sound interesting. I'd love to see some pictures of how you guys have set this up...


#15

Great thread, David and all, I did the incision thing just now with my openrov potting step. See pic that i stripped a very small section of the wire that will be in contact with the epoxy. I was worried about the insulation jackets passing water through to the electronics tube. see pic


.


#16

Russell, would you mind further explaining your potting technique on the rear of the cable gland? Anything specifically different for potting the gland?


#17

Hi

I would be happy to explain

when you pass though a bulkhead or open to water like what we are trying to do with the glands the following steps make sure you don't get a leak.

use a sealing fluid on the glad threads is a PTFE or similar thread sealant

when you strip back the cable ensure the jacket is below the potting compound and not exposed into the electronics area as if you damage the jacket you flood the can.

ensure the oring on the gland is a good fit and jacket is secure i use silicone sealant when assembling this keeps glad closed and a o-ring on the base of the thread to prevent over tightening

if you need more help my email is info@interspec.org


#18

For potted pass-throughs I have had really good luck with Brass Thread Adapters. You can find these at my favorite store McMaster-Carr . Part Description: Brass Thread Adapter 1/4"-20 Male X 1/4" NPT Female Part Number: 4624K32.

I am an engineer and use SolidWorks for design projects. So, I like the fact that I can get a 3D model, part number (for my Bill of Material), Unit Price, and ultimately a physical part-all from one place! Also, I like the fact that I can download native SolidWork and STEP files -this saves me a lot of 3D modeling time.

Getting back to the water tight pass-through pass....all you need now is JB Weld Epoxy. I linked to the 24 hour cure stuff instead of the 4 hour. I have not had good luck with the adhesion properties of the 4 hour and I would not recommend using it.

Pass your wires through the threaded adapter, plug one end with clay or puddy, mix epoxy, and then pour the JB weld into the threaded adapter. Make sure to move your wires around as you pour-this will help to remove any air that is in the adapter and fully coat all your wires. The puddy on the underside of the brass fitting will form a potting well and will limit/stop the epoxy from running out of the pass through as it is curing. If you have a lot of wires going through you may need some plastic syringes to get your epoxy into the pass-through. I would recommend warming the mixed epoxy under a light bulb for just a couple of seconds. This will change the consistency and make it flow like honey into the syringe. Once in the syringe you can fill the pass through from the bottom of the potting well up.


#19

This is very interesting stuff... There appear to be several conflicting ideas about how to properly pot cables. I'm curious, have any of you depth tested your ideas? What kind of depth have you gotten without failure? (I guess this is also assuming the pressure vessel didn't fail in some other way).


#20

We've gone to about 20m using the potted pass-through with wires going through (during operation of a functional OpenROV). We've tested to a pressure of 210psi (equivalent to a little over 130m) in a test chamber with potted endcaps that had no wires going through. More tests are pending.

Here's a thread that gets into some of the pressure chamber testing:

http://openrov.com/profiles/blogs/pressure-testing-main-tube?id=6365107%3ABlogPost%3A36762&page=2#comments

Eric