CCD chip in mineral oil...

specialist companies that make lenses out of it. No chemical compatibility issues with oils either. If your project is interesting enough and there is any chance of publicity, you might even be able to blag a free sample. ;)

m:

camera, also submerged. secret, can't tell, etc...

rge the imaging chip in mineral oil?

acrylic.

it should easily, a quick google shows a document from the 70's about acryl ic windows for hyperbaric chambers

-Lasse

cks

I'd just clamp it to the outside out the container with an o-ring behind it

-Lasse

Yes, best solution. And you must also cut a retaining groove in the acrylic to hold the O-ring, say 40% in...That's a little tricky because you want a smooth acrylic sealing surface there - so should be polished a Q-tip in lathe plus abrasive or flash with a Hydrogen torch. Or you can cut groove in metal and weld it on. (or solder even.)

All this being said, I feel a little queasy about an O-ring at 3000 psi or whatever this is. :) Can always put a bead of silicone around outside edge. jb

And the space behind the lens (plastic

Well, there's your killer. The fill is oil, and the lens is basically oil (plastic). So, there's no difference in index of refraction, so it looks more like a sheet of flat glass than a lens. You are going to have to get a glass lens in there, and it is going to have to be extra high index of refraction glass and then you'll still have to adjust the focus.

Jon

I appreciate your train of thought, but the journey into high pressure windows removes that lens design problem. There is a cost of Fresnel reflection at the window interior - this is 4% by rule of thumb. Add say 1% at the outer surface and you have a total of 5%. This formula - (n1-n2)exp2/(n1+n2)exp2 = (1.5-1)exp2/(1.5+1)exp2 = .25/6.25 ~= 4%. This is for normal angle and describes light which has no coherence, so no interference effects. Multiple internal reflections are ignored. (1)

The point is that a steel box plus window allows you to enjoy your earthling optical design without re-designing a lens for oil immersion.

Without knowing how deep this device will go into the ocean, we can't know the pressure the ccd chamber will have to endure. But even at 2 miles down - 5000 psi, it looks OK.

And - regarding the oil immersion of the MoBo and peripherals - I'm beginning to wonder how "solid" solid state IS. Are there minute air pockets in capacitors, IC's, resistors, etc. that will saturate with oil at 5000 psi? It's not the environment they were designed for.

So worth thinking of the steel chamber for the PC as well - so that you substitute the minor hassle of a pressure chamber for the potentially bigger, and uncertain, problems of high pressure on the components. Better the Devil you know, etc.

just a few thoughts.

jb

(1) Interestingly, this interface reflection was recognized by Newton as containing a conundrum - How is it that a particular piece of light "knows" whether it will reflect or not? About 1/25 of the photons "decide" to reflect. Who makes that decision? At least this is Feynman's thinking as well...FLP

jb

Thanks everyone. It looks very much like you nailed my underlying problem. Since I hadn't considered RI, I wasn't paying attention to the right things when I did my "testing" of the camera. Now that I went back and checked wi th this in mind, I can see that I AM getting an image, it's just VERY fuzzy , as several of you predicted (maybe the camera is just acting as a pinhole camera, with way too big of an aperture).

As for many of the other questions/predictions, while I won't guarantee the re isn't something else I missed, I HAVE given this way too much thought...

Acrylic windows and O-rings: Probably out, or at least out of my budget, si nce the guesses as to pressure so far are way low... (Not for the immediate future, but that's the eventual plan.)

Oil penetrating the electronic components: I'm fairly confident this is not an issue. It certainly won't bother resisters unless they gets hot enough to boil the oil, which seems unlikely. I'm fairly sure capacitors will eith er be unaffected or INCREASE capacitance, and I don't think that will disru pt anything in this system (might even make it MORE reliable). Now that I s ee that even the CCD chip functions, electrically, with the oil in direct c ontact with the chip surface, my last electrical concern is put to rest. Ai r pockets in the ICs ARE a concern if they they DO NOT fill with oil and co llapse under high pressure, damaging the chips. Like I said though, even in the event of catastrophic failure, I'm out maybe $100-150 in materials.

This is really a proof-of-concept, and the camera is not essential, but wou ld be way cool. Even if I can't fabricate, or choose not to invest in an ap propriate lens system for now, if it works to a few thousand psi, there sho uld be money for fluorite lenses, if not diamond. If I have to turn it into a proper pin-hole camera and add more lighting, that would be sufficient f or now, and cost just a few bucks.

Oh, and someone mentioned the interface between the "window" and the water. That will be roughly flat, so refraction should be irrelevan t.

Currently, my biggest concern for failure is that I'm eventually shooting f or a range where the everyday definition of "incompressible" no longer appl ies. I haven't yet researched (assuming anyone even knows in this range) th e rates at which silicone compounds and plastics compress. If they are too different, that could eventually lead to chips cracking and failing. Again, spectacular failure IS an option.

Again, thanks everyone for the help, and I will update when/if I get somewh ere...or maybe come back for more help. :)

Nope, not if you're wanting to work off axis and using the smallest apertures (< f/16).

Cheers

Phil Hobbs

On Sunday, May 11, 2014 7:53:09 AM UTC-7, snipped-for-privacy@yahoo.com wrote: [about poor image quality in camera dunked in oil for work at some kind of (?)water pressure]

since the guesses as to pressure so far are way low... (Not for the immedia te future, but that's the eventual plan.)

for a range where the everyday definition of "incompressible" no longer ap plies. I haven't yet researched (assuming anyone even knows in this range) the rates at which silicone compounds and plastics compress. If they are to o different, that could eventually lead to chips cracking and failing. Agai n, spectacular failure IS an option.

Don't be so sure there will be a failure; there was a project some years ago (DUMAND, deep underwater muon and neutrino detector) that developed optics and electronics for use five kilometers under the ocean. A lot of research has been done on windowed pressure cells at 1Mbar (one million atmospheres pressure, about 15,000 kpsi) but that involves gem-quality diamonds for windows.

At the low-cost end, you could hacksaw a metal-case vacuum tube, to get a bunch of sealed feedthrough wires (eight of 'em, on an octal base tube), and weld/braze/solder any case you want, complete with thick glass window, for a video camera.

In that case then your biggest failure mode may be the oil getting forced into the packages, then blowing things apart when the pressure releases.

But if you're working in an area that's enough $$$, when you get that far you can consider bare dice bonded out to whatever substrate you want.

DUMAND wasn't as deep as that, I don't think, and they had a nasty experience with one of their expensive gigantic PMTs imploding, which produced a pressure wave that set off a cascade of them. Some major fraction of their nice 450 mm Hamamatsu PMTs were destroyed.

Cheers

Phil Hobbs

Sorry, I should have been more clear. The camera will likely be epoxied to the window, in as close to the same plane as possible. For the early runs, it will just photograph whatever happens to come into view. I expect it to be easier to maneuver the probe to point the camera in a desired direction than to have a pan/tilt camera behind a spherical window.

DUMAND was 5km off Hawaii but never completed. The DUMAND hardware was donated to NESTOR:

Everyone remembers the the Super-Kamiokande implosion that destroyed more than half of the 11,146 photomultiplier tubes. It was only 41 meters high:

What a mess to have to clean up.

I'm not really planning any trips to Jupiter, so I don't think I need to go quite that far. When I mentioned diamond leases, I just meant as a tiny, very high refractive index lens that would work well submerged and see no net pressure.

One classic maneuver is to combine the lens and the window, so there would be water outside and oil inside.

Joe Gwinn

maybe he wants to send it to the bottom of the ocean in a plastic bag...

Tanks and windows are cheap. Welders prefer thick (.25-.5 in) steel. jb

ould be way cool. Even if I can't fabricate, or choose not to invest in an appropriate lens system for now, if it works to a few thousand psi, there s hould be money for fluorite lenses, if not diamond. If I have to turn it in to a proper pin-hole camera and add more lighting, that would be sufficient for now, and cost just a few bucks.

ant.

for a range where the everyday definition of "incompressible" no longer ap plies. I haven't yet researched (assuming anyone even knows in this range) the rates at which silicone compounds and plastics compress. If they are to o different, that could eventually lead to chips cracking and failing. Agai n, spectacular failure IS an option.

Hmm that's a bit "weird". There is no such thing as "incompressible", ever ything has a bulk modulus. I'm sure you can look it up for plastic.

Seems you should be thinking about fiber optics maybe.

George h.

where...or maybe come back for more help. :)

I wonder if anyone has tried lower index concave lens in water... I'm guessing there are not too many good low index material either.

George H.

Cheapest off the shelf but not achromatic lens would be sapphire ball or half ball.

It all matters - the focal position with a boundary to water will be different even when that side is flat. Trace a ray at angle theta...

You are likely to have some interesting problems if there is any trapped air in the system. Above about 10atm the oxygen content of any bubbles will be sufficient to cause spontaneous ignition of mineral oil.

A German van der Graff accelerator machine once burned down as a result of using high pressure air to increase the electrical breakdown voltage. The investigation revealed that it was doomed from the moment of the first spark since there was more than 1atm pO2 available.