Underwater case

[colotron]

Hi there.
I'm looking for information about case/enclosure design for underwater electronics (200 meters depth or more).
I've seen some with O-rings but they are too big. I intend to use something like 50x20mm (2"x0.75").
Then I saw this small case:

http://www.lotek.com/lat1000-fish.htm

Does anyone know how this cases are made? (what kind of plastic, how to seal it, suppliers).

Cheers.

[Rerouter]

at a guess that is not a case, but set in a hard corrosion resistant resin,

as for an actual case, 200M is deep, short of a similar approach where there is no free air to violently compress or expand, your going to be hard pressed finding small watertight enclosures that can withstand 20 atmospheres (2 Mega-Pascals) of pressure difference :/

[colotron]

Yes, a case seems to be unrealistic, I will read more about resins and potting, thanks Rerouter.

[sprocket]

that is going to be very difficult making something that small going that deep. I work as an engineer here at the the Danish Technical University Marine reseach faculty, and I work with alot of equipment hat have to go down to many hundreds of meters. And it's not only a problem from a engineering prospective, but as soon as we are talking equipment for deep water there is very little in regards of off the self products, and much of we have we have to build of self. So it wont be cheep either.

But a little more info on what it is supposed to do, what sort of sensors and so on would be helpful in regards to pointing you in the right direction. But you probably going to have to prepare you self that what ever you need is mostly like going to have to be custom built.   

[colotron]

Thanks sprocket. I was asked about the possibility to design an inertial tracking system for penguins, i.e. accelerometer + gyro + memory + micro + battery. And (if possible) pressure & temp sensor (meas spec have the MS5803-30BA, but this is not a priority right now). Without the pressure sensor there is no need of any contact with the external environment, so it should make the design easier.
But on the other hand, the accelerometer and gyro packages are affected by stress, they should be isolated of the external pressure (shield of some kind if potting is used, not for RF but for pressure).
And yes, it will be not cheap, but the cost is directly linked to solve the potting issue (not case anymore), and I'm trying to have an idea of the effort involved.

[fcb]

200M is not very deep, pretty straightforward to design a casing for that depth.  It's all about form-factor though, turned cigar tube with a flange/oring and lid would work fine.  You could even make something flatish & round at that pressure (if you design it right). What profile does your penguin prefer?

Potting is a good way to do it, if you don't need to service it at all and you can stand the extra weight, and can get the data out with IR/RF.

[SeanB]

As well you have to consider that you cannot easily use parts that have any air inside them. No crystals, no ceramic capacitors ( not air but compression causing piezoelectric voltages to appear) and no metal cased transistors or IC's. If the parts are small enough you can encapsulate them inside a test tube with clear low exothermal epoxy, and centrifuge it during cure to remove bubbles before it sets. Anything containing air pockets has to be able to withstand the pressure without collapse.

[fcb]

As well you have to consider that you cannot easily use parts that have any air inside them. No crystals, no ceramic capacitors ( not air but compression causing piezoelectric voltages to appear) and no metal cased transistors or IC's. If the parts are small enough you can encapsulate them inside a test tube with clear low exothermal epoxy, and centrifuge it during cure to remove bubbles before it sets. Anything containing air pockets has to be able to withstand the pressure without collapse.

It's only 200M, not 10,000M. About 20bar/300psi, and it's attached to a penguin...

[colotron]

SeanB when I read "no caps + no crystals..." and   
Thanks fcb. I guess the form factor could be tube-like if that resist the pressure better than other shapes, and yes, having a lid will be great. What kind of plastic could work?.
The MEMS accelerometers and gyro will need to be isolated of the wide pressure change on the case. So this means there should be a air space inside for the electronics. Is possible to have such a slot inside the enclosure?.
I think I will have to experiment a lot before have any definition to this respect.
Thank you very much for your answers, if this goes beyond speculation I will post a follow-up.
Cheers.

[fcb]

If it's plastic, you'll need to make it thicker than SS (stainless steel).  I would use SS and you'd probably get away with a screw thread on one end rather than flange/oring.

Your basic design will be at 1atm inside (think of it like a little submarine), and if your cute, you could design the water pressure sensor to screw in at one end of the tube (or into the lid).

Whatever you choose to do, design and pressure test your housing first.

[HLA-27b]

This calculator gives remarkably thin walls for 20 atm. Basically you have nothing to worry about.

http://checalc.com/calc/vesselThick.html#tab1

20 Atm. 30mm diameter gives you 0.3mm wall thickness for 316 stainless. It is hard to find pipes with walls this thin 



Another idea...

1 - Find a PET preform manufacturer
2 - Pick a preform to suit your project
3 - ?
4 - Profit

[colotron]

Ok, this doesn't look very bad now.
fcb you read my mind (sensor pressure into the lid).

Thanks for the calculator HAL-42b, you remind me of:

[HLA-27b]

Thanks for the calculator HAL-42b, you remind me of:

Just google "preform" and look at the pictures. It will make sense immediately.
You can not possibly go simpler nor cheaper than this.

[G7PSK]

Should not be hard to make a case waterproof to 200m you can get watches to that.

https://www.google.co.uk/search?q=watches+200m+water+resistant&aq=0&oq=watches+200m&aqs=chrome.1.57j0l3.21640&sourceid=chrome&ie=UTF-8#q=watches+200m+water+resistant&hl=en&source=univ&tbm=shop&tbo=u&sa=X&ei=-QEtUeDPE4PV0QWSwIAg&ved=0CFgQsxg&bav=on.2,or.r_gc.r_pw.r_cp.r_qf.&bvm=bv.42965579,d.d2k&fp=3168db55ceb29c02&biw=1280&bih=909

[dr_p]

No you cannot! A "200m" watch is NOT suitable for diving down to 200m.

http://en.wikipedia.org/wiki/Water_Resistant_mark#Water_resistance_classification

[fcb]

Just google "preform" and look at the pictures. It will make sense immediately.
You can not possibly go simpler nor cheaper than this.

  Genius HAL-42b! I watched a "Megafactories: Coca-Cola" a few weeks back, and they used PET preforms to blow-mould into coke bottles.

[G7PSK]

No you cannot! A "200m" watch is NOT suitable for diving down to 200m.

http://en.wikipedia.org/wiki/Water_Resistant_mark#Water_resistance_classification

Quite likely those I linked will not but your own link gives details for depths even greater where mixed gas is required, the thickness of case is not going to be that great for 200 meters, and watches have weak links in the buttons and winders a case with an o ring sealed top wont,it not as if its going to be strapped to the side of Alvin and taken for a spin.

[colotron]

Just google "preform" and look at the pictures. It will make sense immediately.
You can not possibly go simpler nor cheaper than this.

  Genius HAL-42b! I watched a "Megafactories: Coca-Cola" a few weeks back, and they used PET preforms to blow-mould into coke bottles.

Excellent!!!!, will be the first thing to try.
Thank you all for the good ideas.