This post details my external (ie non EnOcean) sensor that feeds the Hot Water temperature into openHAB.
electronichamsters (Eric Tsai) posted some great information regarding the use of RFM69x transceivers.
You can follow his posts here.
The RFM range are very well designed, but unlike many other Tx or RX you can't just stand them up and plug them in.
While we were generously supplied with a licence for Cadsoft Eagle, my other commitments meant I wasn't going to have time to design and send away for boards.
While I was in the www ( is that Wonderful World of Why not buy it), I ran across the Moteino range of Arduino which allows the RFM range of transceivers to be soldered on the rear.
image from LowPowerLab showing the raw unpopulated boards.
The price of the competed item was below what I could get the raw parts for.
The documentation and support also seemed to be there (unlike many other sites just selling the product), so it made sense to support someone else.
I decided to use these as my external temperature source.
(I may need to order some USB versions for future projects )
In my application, I wanted to ensure that 'Granny' was following regular hygiene, since it provides another method of checking for 'normality'.
By measuring the Hot Water temperature as it leaves the cylinder, you can see the change and build a picture of use.
In my case the hot water is by LPG burner, and while you could access the 'in use' indicator by hacking the protocol, this method is simple and universally applicable.
I couldn't use a solar powered sensor since there is little light inside the heater (or a cupboard/ceiling space if you have a conventional cylinder).
The burner needs power, and for some time now I've been checking the bargain bin at one of our large retail store chains each time I go past.
Generally they have older version phone chargers, in both Mains and 12v version, and they can be easily re-purposed to power an Arduino or similar low current device
The price is right at $1 or $2 (or sometimes less), and I had one that suited this purpose.
The transmitter was programmed and tested on the bench, and it only sends data if the temperature changes, which reduces the 'RF clutter' or noise.
( I may need to adjust this to only send after 1 or 2 deg change )
Fitting it turned out to be the typical ... yet another minor stumbling block, where the 2nd outlet is covered by the pipework.
photo of burner, Hot water is on the left, and I've removed the lagging (which should extend down the pipe as well )!.
Rotating the outlet is allowed under NZ law, since I'm the homeowner, but regardless it needs to be done safely.
Ideally you need a hand to identify the correct circuit breaker.
The practice is Prove, Test, Prove, and this means Prove your measuring item works, Test the appliance is not live, then Prove your measuring item again.
I also encourage you to read my blog about electrical safety here.
Mains voltages ARE LETHAL, so if you don't know what you're doing, its time to ring someone who does.
Since we know the socket is live (otherwise the water would be cold)
For this you can use an indicator (desk lamp or similar).
- Prove it works by powering it from the socket.
- Test by having the other person switch off the breaker, and verify the indicator/lamp goes off when they switch the breaker off.
- Prove by having them switch the breaker back ON.
Then have them switch it OFF and ensure the indicator goes off.
I have temperature control units that are powered from the LPG burner, so they were my indicator.
I juggled the outlet to allow the power pack and mains lead to be accessible, along with the switch.
It's tight but it works.
I needed to add the transmitter into this metal box that the burner lives in.
- First issue is the RF will be limited.
- Second is that having a device rattling around on a metal surface relying on paint as the insulator is not a good practice.
- Third is the creepy crawlies and moisture, since it's on the south side of the house where all our bad weather comes from.
The RF escaping is solved by poking the antenna down beside the plastic Hot Water pipe where it exits the metal box.
The lazy plumber didn't provide any sort of protection or sealing, so I can easily slip it past.
(This is on the south side of the house and moisture can enter and get directly into the cavity between the bricks and the internal walls ...not good building practice, and no wonder we have weathertight issues in NZ.)
Not an ideal installation, given that pipes move with temperature changes and no sealing against moisture.
I found a potting box that allows the transmitter to be held and covered from the moisture, while not shorting out on the metal surface.
The creepy crawlies will have a nice home, so they are happy as well.
Before I finished the installation, I figured it was a good idea to check it works.
fvan (Frederick Vandenbosch) has provided a very good post about getting serial data into the RPi and OpenHAB here
The one minor item to add is that you need to install minicom
$ sudo apt-get install minicom
I followed this guide, which covers the settings and what all those letters mean.
I'm using an Arduino Nano as the receiver and it plugs into one of the USB sockets.
I found I needed to change the line to point to the USB
edit: I wondered why my normal sensors didn't seem to be working and I had incorrectly pointed openHAB at the USB only ...this is the correct line
The temperatures were coming through just fine, so I left it while I did some of the other tasks around the place.
While I was doing these tasks, the outside temperature dropped from 20 to 12 degs C, and this showed very well while it was exposed.
Since it worked, I attached the probe to the Hot Water outlet pipe fitting.
However this photos shows the poor contact, and it had too much lag, so I have since moved it down onto the plastic pipe.
With the tests completed, it was time to stick it back together and tidy the wiring before the rain arrives.
The final install. The transmitter is behind the Hot Water pipe on the left side.
This post is already long enough, but a quick look at the graph from a sink fill of 50 deg water for washing some dishes shows the lag.
The graph for showers will tend to have a much longer peak period, and will tend to stand out over a 24hour period.
So another part ticked off the list.