Environmental Sensing Peripherals
I've been on the lookout for peripherals to use for the sensing. Wind direction and wind speed are probably the most complicated since the least expensive transducer is an anemometer, which has mechanical components. I have found one at Sparkfun that looks to be usable.
There are a number of other things at at Sparkfun that are also of interest. They have an environmental sensing board that combines a number of transducers and is intended to work with weather meter hardware.
I've been doing experiments with the Silicon Labs Thundersense 2 board, which has environmental sensors also, but I don't think that board can be used in a deployment. The bluetooth radio doesn't have enough range and there are FCC compliance issues to be considered. Alternatively, I've been looking at the Silicon Labs radio modules. These modules are FCC certified and act as daughter cards for a starter kit. These might be okay for an initial deployment.
The Thundersense boards have proven their use since it was necessary to insure that we could use the micca run-time in conjunction with the Silicon Labs Bluetooth software stack. You are dependent upon the SOC supplier for the complex software to run a radio in accordance with a standardized protocol. That's not something a project like Farfalle would be able to undertake.
Of course, the radio modules don't come with any sensing transducers, hence the need to look at other boards. We should be able to wire up any external board through the I/O pins that are brought out of the starter kits. There is an expansion header on the starter kit and with luck the connections should be able to be make through the header, just to keep things tidy.
Andrew
23 Feb 2019
I have ordered and received new radio boards and environmental sensors. These are the same parts discussed above. The Silicon Labs Bluegecko starter kit (SLWSTK6101C) comes with two radio daughter card modules. I have also gotten a BG13P module daughter cart (SLWRB4305A) and that will be the target for some additional experiments. I have also received the Sparkfun environment sensor shield and the weather meter. The assembled anemometer, wind direction and rain gauge that form the weather meter are quite a contraption. But it looks reasonably well built and is the right price for a project such as this.
I've begun to assemble all the glue parts to put this together. The radio mother board has an extension connector which connects a number of I/O pins from the radio SOC to the outside world. The environmental sensor transducers run on I2C. The weather stuff will require an I/O pin connected to a pulse counter, ADC,and an I/O pin to generate an interrupt for the rain gauge. The idea is to bring a connector off the board extender onto a prototype area, wire in the environmental sensor and weather stuff and use that as a base for experimenting with the low level data acquisition code that will be necessary as to base onto which domain dependencies can be bridged. To say the least, all this hardware wiring stuff is right at the limit of what I know how to do. All this electricity stuff is such a bother. Circuits, connectors, voltages -- Oh my!
Of course, I will have to repeat the experiments on integrating micca and the Silicon Labs bluetooth software, but there's nothing new there, just some work to pull it together. More later when I get some of this together.
Andrew
1 Mar 2019
Today I soldiered on some headers to the Sparkfun Environmental sense board and decided to take look at the signals coming out of the weather meter. There are three peripherals on the weather meter.
- Anemometer which pulses on each revolution
- Wind speed indicator which outputs a different voltage depending upon the direction it is pointing
- Rain gauge which pulses when its bucket is full and dumps.
I applied 3.3 Volts to the board and hooked up the weather meter cables to the appropriate connector. The board has some RJ11 telephone like four conductor cables.
Looking on a scope all the signals looked as expected.
- The anemometer pulses high on each revolution.
- The wind gauge output varies according to the direction. Giving it a good spin generates an interesting scope trace.
- The rain gauge pulses low each time the bucket dumps.
One serious issue we will have to deal with later is power consumption. The anemometer / wind direction was consuming about 3 mA when connected. This is way too much. I suspect the wind direction since it works by simply inserting a different resistor path (using switches) into the circuit and so you get a different output voltage. I think such a circuit would dissipate some power all the time it has a voltage applied to it. It may not look like much to some folks, but 3 mA is a lot for us. We will probably have to look at some way to gate the power to some of these things. That could impact how frequently we can measure, but that is an issue we can solve later.
Right now the good news is that the electro-mechanical parts work and produce signals that should be easy to handle. Next step is to get the I2C sensors hooked up to the radio board and see if we can read them. That would put us in a good place to start finalizing some decisions about hardware capabilities.
Andrew
March 16, 2019
This is what the prototype board is looking like. I made the wiring harness which now hooks up all the peripherals on the Sparkfun board.
March 19, 2019
I have finished up the work to talk to all the peripherals on the Sparkfun board. I now have all of them working as expected, including the rain gauge. We still have the problem of current draw on the wind direction. I suspect what we'll have to do is a little surgery on the connectors so that we can directly connect to the anemometer and wind vane separately. Then we should be able to prototype a small analog switch to the power line of the wind vane. The analog switch would be controlled by a GPIO line or maybe we can get a I2C controlled switch. The measurement of wind direction would then have to apply power to the wind vane before taking the measurement. A little messy, but not bad enough that I want to abandon the Sparkfun board. I knew the wind measurement transducers were going to be the most problematic. I'm not particularly happy with the Sparkfun anemometer, etc. but it is relatively inexpensive and getting good ones is a much more expensive undertaking.
Anyway, I'm finished with these experiments. I am convinced that the Silicon Labs radio modules works on the bluetooth side of things and that it has enough peripherals to control the environmental sensing side of things. Given that I haven't found any better radios or sensing transducers, I think we should plan on using these boards for the first phases of the project.
Andrew