Status:
- all supporting custom libraries are now written and include tests
- main application functional
- schematic for board produced (one hole need fix)
- 3D enclosure design with wooden plate completed
Picture of the final barograph
There is also a complete user manual with a lot of cool build pictures. (The brochure has been written in French because I offered a Digibaro to my father for his 80th birthday, but it should not prevent you to have a look at it!)
Develop an open source digital barograph that does not require an external power source, has a very low power consumption, and works on a boat independantly of any wifi/cell connection.
I still believe that the barometric pressure is an essential decision making tool when sailing. Displaying a graph rather than writing down single values provides an instant evaluation of the trend.
The main idea is that the beautifull conventional mechanical barograph are not suited for marine use (not resistant to vibration) and require a weekly paper change.
The paper-electronic barographs aleviate this problem, but I remember that ours was constantly jamming.
Several electronic barograph already exist on the market, but they all require an external source of power (and are pricey).
Today you can have an app that will siphon the data from a nearby weather station and display the barometric pressure in any format using your brillant smartphone screen. And this is even not really necessary since having data connectivity would allow to also get a full detailed weather forecast, much more accurate, and covering a wider area, than what you could analyze yourself.
Despite our ever more connected world, I am old fashioned and like to believe that having systems not relying on data flowing wirelessly to your boat can be useful. It turns out that the fundamental building block for such a project is readily available: several extremely accurate digital pressure sensors can be sourced at low price (you may already have one in your cellphone).
I also wanted not to rely on an external source of power. This is not so much to avoid battery drain (the most basic solar panel would provide more than enough), but to avoid having extra wiring.
The final concept is that you would be able to leave the barograph on your boat, and when coming back after six months, you would be able to read the barometric pressure of the last few days because the barograph would keep collecting data. Of course, this is probably superfluous since after leaving your boat six months, you probably have more than 5 days of work before taking the sea again (hence you could turn on the barograph).
At any rate, being a style exercise or a real need, the goal is to develop an extremly low power barograph that could sustain at least six months on a set of typical AA alkaline batteries!
There are few major difference between this project an my previous (never completed) digibaro v1:
| Digibaro v1 | Digibaro v2 |
|---|---|
| Target low cost | Cost is not the main priority |
| Classic UI with 5 buttons | Mimimalist UI (analog feel) |
| LCD screen | e-Paper display |
| Only stores 5 days of data | Long term recording (>20years) |
| Typical "maker" approach | High reliability desired |
The first version of my digibaro was based one BMP085 sensor, a momochrome graphic LCD and an Arduino Pro-Mini (3.3V). This version use modernized components.
The micro-controller is a ARM Cortex M0+ (SAMD21) packaged by RocketScream. This board was selected because it was specifically designed for low power applications (no extra LEDs, efficient regulator, etc.), and it provides 2MB of flash memory on board.
The pressure sensor is a BOSH BME280. It provides excellent accuracy plus a humidity and temperature sensor on the chip. The BMP680 would provide an extra Volatile Organic gas detector, but a the cost of an extra complexity in managing the sensor.
The display is based on a 400x300 black and white e-Paper system packaged by Waveshare. E-paper display would seem ideal for this type of low power application. However, since the refresh rate is so slow (~4s for this particular display), updating the display takes a significant amount energy. There is a cutting-point between a LCD (that refresh very quickly and consume very low power in standby) and an e-Paper display that depends on the number of refresh per day. Still, for this application, the e-Paper provides a nice, more natural, look.
- Require external power source
- 10mA without backlight
- Retired
- ~300 Euro
- Require external power source
- 20mA @ 5V DC (160mA with LED backlight)
- Storm alarm
- Factory calibrated
- DBX2 User Manual
- ~$500
- Consumes 100mA (without backlight), require external source of power
- Pressure change graph
- Meteoman screenshots and discussion
- ~$200
- Paper recording :-)
- Swiss made ;-)
- One single C type battery for a year!
- Very cool, however we had a similar model that jammed quite often.
- Meteograf User Manual
- ~$900 without case
In May 2019, I just discovered this: Arduino + BMP085 Barograph from PauloDevelo on Arbatus.
Currently digibaro2 is built using platformIO. This mainly constrains the directory structure. Each test program show in the comments the command to build it. The main application uses the platormio.ini definitions and calls an extra script (version.py) to inject the git hash. This last part is not directly portable to another build system.