"Sea Trial" two has been going on since Sunday 14th of June. I installed two units, one on a buoy at St Heliers Beach and one at neighbouring Kohimarama Beach. The two units are slightly different, one has the same external antenna used in the first trial, the second has an internal antenna. The intention of this difference was to confirm if the signal strength was going to be sufficient without having to put another hole through the waterproof case.
See the results:
The battery life issue is solved!
The changes between version one and two to extend the battery life have been more successful than I expected. Trial one lasted 12 days (sending 6 updates an hour) before the battery voltage dropped below 3.4v and it stopped transmitting data. Trial two both units have been running for 23.5 days (sending 4 updates an hour) and the battery levels are at 65% and 56%! So it looks like I'm going to get nearly four times the battery life this round!
Here's a graph of the battery voltage vs the number of readings the unit has sent back:
I'm guessing the difference in voltage between the two units is to do with the life history of the batteries. They were all out of my parts drawers and have had different uses and abuses in the past.
When I retrieved the unit after the first sea trial it wasn't 100% dry when I opened it. I wasn't sure if this was because it was damp when sealed or if water got in while deployed.
Yesterday as I was out swimming I checked on the unit installed at Kohimarama beach. It has was appears to be 2ml of water inside the enclosure. I've left the unit out, the way it's setup there's a margin inside the bottom of the enclosure without any electronics so the water is not directly impacting anything yet.
I'm going to leave both units out until the batteries go flat or the water ingress gets too extreme. In the mean time Version 3 is underway, so far the only change is to add a light intensity sensor and start logging this too. When I close it up I'll add some O-ring grease (same as I use on my camera housings) and see if this helps the seal.
Got any ideas on other sensors to add? I2C or SPI.