Usually the electronics is contained in sealed and strong metal tubes and are not subjected to pressure. When these tubes leak, the electronic devices inside are not only wet, but also crushed by the extreme pressure. Everyone involved in this line of work has a collection of conversation pieces (assorted crushed items from a failed tube).
As VE1BLL sez you can't actually leave the parts to the horror of highly conductive and corrosive saltwater immersion.
In fact you can't raise the pressure much at all or you will drive moisture into the pin/package interfaces which will cause immediate failure even with distilled water.
I believe that underwter electronics are sometimes in an "Oil" filled container that has a pressure equalizing metal bellows. The "Oil" provides a thermally conductive heat path to the exterior sea water and the equalized interior and exterior pressure minimizes the wiring seal requirements.
When I worked for a small time player in the silicon chip business, we used to produce high rel parts for down hole well datalogging. (That's oil wells).
Some parts were tested to 200 deg C and were complete with a little hole in the top of the lid. (Package was a cermamic dil 16 with a gold plated lid as a seal).
This stopped the package being crushed by the pressure.
I assume that the electronics was surrounded by oil of some sort.
I forgot to add to the my posting. The electronics are housed in a oil-filled cavity. I have placed the board in a pressure chamber filled with oil and while my board is powered on, I am pressurizing the chamber from 0 psi to 6000 psi (which is roughly ~13000feet of seawater depth). So I was just curious if anyone else has had any experience with diodes, specifically schottky BAT54SW series, misbehaving under pressure.
IRstuff, Misbehaving meaning with no analog input to my microcontroller the output is at max. I am using a microcontroller to control dimming function for a underwater light. With 0-5VDC analog input, controls my phase delay or triac firing(using a zero-crossing input as a reference). Well with no analog input (0VDC) there should be no light output and there was. I have a dual schottky diode, BAT54SWT1, connected to the front end of my analog input to protect from ESD or voltage spikes. I was thinking with no analog input (0VDC), that BAT54 diode was leaking and setting the PIC analog input at 5VDC, which is the max analog input that gives me max brightness (triac firing right after zero crossing for each half cycle). This was a prototype board, PCB made using a local electronics shop copper board and etching chemical. Mostly all the components were surface mount, soldered quickly (not in a controlled environment) so I am thinking the board was being stressed non-uniformly. That's the only thing I can think of.
You realize that the PIC has THE beafy-ist internal Schottkys on any micro? You don't need to add an external leaky set! Ditch that puppy! And next time around just make sure there is SOME resistance on the input to reduce the current the Schottkys would need to handle. But for the moment just pull those external diodes.
Hmmm... The Microchip site has a ton of them. The diodes on the inputs are so large that you can run 240Vac mains into the pin if you just proceed it with a proper resistor!
Look for "line voltage input".
Have a good weekend too!
you should jack a picture or two in here that project sounds like a bunch of fun!
