23-8-14
Water Sensor WS1 - Progress
One of our sensors was hooked up to a humidifier to test it, it worked there day in and day out so we were able to collect quite a bit of data, we were happy with the overall results but there were still a few unanticipated situations:
- There were several repeated fluctuations in the water level as it left the critical point, and after a few occurrences we found this to be understandable as the water level was slowly decreasing in a wobbly process - the humidifier disturbs the flow of water as it works, perhaps the fish tanks do this as well? We didn't have a fish tank on hand to test.
- The condition would open and close repeatedly over a short period of time, maybe 10 minutes, where the open state is short-lived, it doesn't last, it's like water splashing up. And then it falls again.
From this pattern, it seems that a filter could be built into the firmware where only openings that last up to a few seconds are valid openings.
But in addition, it's an overall performance we like, as the connection is stable and the data is reported in a timely manner. Like the feeling of a faithful machine, it gives ha indeed a connection to this water container.
For that, it's great.
We are still conceptualizing to try some different possibilities when it comes to the housing design. But in terms of the hardware, we've pretty much settled on it and we've got the support of the manufacturer, which is a particularly enthusiastic manufacturer that seems to specialize in making these water level sensors and make a living out of it, and they have a lot of different kinds of water level sensors, and we chose this one because it's the smallest one.
To help you understand the progress of this sensor, we've given it its own page: https://docs.screek.io/ws1 The birth of a new idea may not end up being at all what was originally envisioned, but it's still well worth exploring them.
23-8-11
Water Level Sensor WS1 - Test Progress
During two consecutive days of testing, we changed two ESP32 motherboards, and in the end we decided to go with the C3, which we were most satisfied with for its relative stability.
We tested the sensor by sticking it all the way to the glass kettle and we were pleased with the results, the data stability was very high and there were no false positives.
It's worth mentioning the adjustment button on top of the sensor, it was interesting to see how it would repeatedly turn on and off for a few seconds after pressing it where there was no water level but was triggered by a false alarm, and then the optimization was complete and it would re-recognize the detected object.
Another discovery is that if the back of the sensor that the finger is on is not doing the sensitivity adjustment, it may also be seen as a water level.
Metal is always seen as having liquid and metal clad containers cannot be used. We spoke to the supplier about this and learned that their product only has ultrasonic waves that can pass through metal.
That's some new test situations, we are stable with the performance of the sensor so far, we have unfolded the design of the housing and are optimizing it, we believe that we will be able to share it soon.
23-8-9
Water Sensor WS1 Test2 In addition to the long-term stability test, we tested a thickened case today, and with this thicker surface of the bottle, the test still looks pretty accurate.
