Computing stuff tied to the physical world

RFM69s, OOK, and antennas

In Musings on Jul 15, 2015 at 00:01

Recently, Frank @ SevenWatt has been doing a lot of very interesting work on getting the most out of the RFM69 wireless radio modules.

His main interest is in figuring out how to receive weak OOK signals from a variety of sensors in and around the house. So first, you’ll need to extract the OOK information – it turns out that there are several ways to do this, and when you get it right, the bit patterns that come out snap into very clear-cut 0/1 groups – which can then be decoded:

FS20 histo 32768bps

Another interesting bit of research went into comparing different boards and builds to see how the setups affect reception. The good news is that the RFM69 is fairly consistent (no extreme variations between different modules).

Then, with plenty of data collection skills and tools at hand, Frank has been investigating the effect of different antennas on reception quality – which is a combination of getting the strongest signal and the lowest “noise floor”, i.e. the level of background noise that every receiver has to deal with. Here are the different antenna setups being evaluated:

RFM69 three antennas 750x410

Last but not least, is an article about decoding packets from the ELV Cost Control with an RFM69 and some clever tricks. These units report power consumption every 5 seconds:


Each of these articles is worth a good read, and yes… the choice of antenna geometry, its build accuracy, the quality of cabling, and the distance to the ┬ÁC … they all do matter!