Energy metering and how to listen in on the 868MHz band?

[casper.bang]

I'm almost done hacking an industrial energy meter which I will install as a secondary meter, allowing me to continuously monitor energy consumption in my house across 3 phases. However, this is not quite fine-grained enough for my liking, ideally I'd like to drill down to specific devices/outlets.

Now I can't very well construct my own meters and connect to mains; mostly due to safety and cost issues. However, I came across this product, which looks nice and cheap. The downside is that it appears only a maximum of 5 meters can be used (though I hope this is strictly a limitation of the receiver) and also I doubt if there are rs-232/USB interfaces of the receiver I can utilize.

How hard do you guys think it would be to construct hardware to listen in on the popular 868MHz band for me to start reverse engineering the data protocol itself? The 868MHz is just a carrier, I assume there are countless ways to modulate the actual digital signals on top of this, but are there cheap de-facto standards I may cut my teeth on first? I realize encryption might also be an issue, but for now I am mostly interested in the hardware aspect and whether there are easy and popular 868MHz kits/builds for starting hacking.

Naturally, I'd also like to hear if anyone are aware of alternatives to the product I mentioned.

[madires]

An inexpensive solution would be a DVB-T dongle based SDR like http://sdr.osmocom.org/trac/wiki/rtl-sdr.

[jeroen74]

Likely IEEE 802.15.4. You could use a AT86RF212. Though energy meters very likely use encryption. Good luck with that

[fake-name]

I'm actually working on a project that does exactly this.

It's not quite there yet (it's at the prototype stage), but I think it's exactly what you're looking for:

• Completely open-source(hardware and firmware)
• Scales to large numbers of nodes (hundreds, hopefully)
• Target cost per-plug of $20-30 each, plus a single <$50 base-station
• 120V and 240V versions
• (Hopefully) mesh-based, using 2.4 Ghz comms (nRF24L01+ based)

Attached are the first rev prototype (energy metering is working, the wireless network part isn't done), and the potential next-revision.

It's all open source, and the project files and some discussion can be found on my website:
 http://www.imaginaryindustries.com/blog/.

Look for entries labeled "Power Meter".

[fake-name]

An inexpensive solution would be a DVB-T dongle based SDR like http://sdr.osmocom.org/trac/wiki/rtl-sdr.

While the data is probably encrypted (at least for smart meters), a RTL-SDR based radio is an EXCELLENT $20 way to get started with analysing radio transmissions.

Here are what I'm pretty sure are the local smart-meters talking:

It looks like pretty standard FSK. If you piped it into GNURadio, you could probably even decode it pretty easily, though decrypting the decoded data would be the real challenge.

[casper.bang]

Thanks guys for making me aware of rtl-sdr, will have to order one and find some time to play with it!

@fake-name that's a cool project, open source hardware even... good luck, I shall be tracking your progress.

[amyk]

Take one apart and look inside... if you're lucky you might be able to extract the controller firmware, or at least determine the transmitter it uses so you can find more info about it (maybe sample code/reference designs.)

[ptricks]

I'm guessing the 5 device limit is because of the monitoring rate. 5 devices x 5 second intervals is 25 seconds and then you need time for the devices to send data and not overlap.  They probably figured that 5 was enough and adding more radio channels wasn't cost justified for the people that want more than 5 monitor points.