| << return to blog entries
2010-03-24 10:30:07 (3912 views) Some breakthroughs with waterproofing after we posted the last entry...
1) It seems the O-rings are all ever so slightly different in size. Perhaps this is because they have to be pulled out of the mold and they get deformed when this happens. So we found the best sized O-ring out of the bag of samples and tried that. Immediately the quantity of water leaked was reduced. We need to verify with the manufacturer why there is a size discrepancy and fix it.
2) To better test the O-ring we replaced some "filled" grommets (without holes) to the real grommets so that the O-ring and screw rings would be the only parts which could leak. We're 100% sure the screw rings do not leak. Therefore we are singling out the O-ring for testing.
3) The bottom of the enclosure has screw holes that are too small, and for this reason the screw has to tap its way in rather than just sliding in effortlessly. This has the side effect that as the screw comes out of the bottom and enters the top of the enclosure, the grooves created in the top are misaligned if the halves aren't pushed together with great strength as the screw is going in. The thickness of the O-ring when it is not being compressed is obviously greater than its thickness at maximum compression. Because of this thickness, the grooves left in the two halves are misaligned and so the enclosure never tries to properly compress the o-ring. We were able to improve this situation by pushing the two halves together very hard as the screw was being inserted. Doing this, and applying a great deal of strength to tighten the screws, resulted in very very little water leaking. We were able to submerge the enclosure for half a minute with the PCB. This is therefore the first real underwater test.
Having never given the topic any thought, we naively expected to be able to communicate with the receiver as it was submerged. This "sort of" worked - the range was reduced drastically (to 10 cm away from the sink) and sometimes packets were received with a CRC error.
Looking it up, we found that this is entirely normal. Water, unless it is distilled, does not allow RF to pass - it turns RF waves into heat very efficiently (which is why your microwave oven works). Aside from that, there is refraction both when the signal enters the water and when it enters the inside of the device through the enclosure away from the water.
This is why underwater communication normally is carried out with sound waves!
Comments
Post new comment
 |
