Visonic bitstream protocol

[jcw]

Hello - I'm poking around with a Visonic door sensor, trying to figure out the protocol it sends out over 868 MHz.

What I can see is a very clear pattern of 001/011/101/111 transmit on/off transitions with all pulses either 80 or 160 µS (approx).

When you take out the fixed 1's, you end up with 36x2 = 72 bits:
_-__-_--__--_-_-__-_--_-__--_-_-__--_-__--__-_--__--__-_--_-__-_--_-__--

Probably a header / preamble, the payload, and then a checksum/trailer.

But what is what? Does anyone have more information about the bit allocations used for these packets? I haven't found anything on the web.

(I'm not interested in the secure PowerMax "rolling-code" encoding, just simple home stuff such as this door sensor)

[jcw]

Some corrections and progress: the pulses are 400 and 800 µs, using Manchester encoding, with a packet size of 44 .. 46 bits. Packets are sent 3 times within 2..3 seconds.

I've got a JeeNode (ATmega328 with RF) reliably decoding these packets, but preamble, byte order, bit allocation, and checksum are still unknown.

[jcw]

No responses and replies yet - how surprising!

For future viewers of this thread: please email me at jcw@equi4.com if you have any tips re. Visonic. I'd be interested to add it to the protocols already supported at www.jeelabs.org - so far, that includes the whole range of FS20 devices, KAKU, and (K)S300 weather station sensors.

[j.hoekstra]

I guess it has something to do with a lot of people already using RFXCom or visonic receivers, which at least in the first case already do the decoding. So there's not really a need to re-invent the wheel.
I encourage you to keep at it however, Jeenodes seem to promising parts and I'm thinking about using them for a irrigation thingie.
Maybe B_weijenberg can give you some tips or pointers where to get the info, he's active on this forum and others.

[jcw]

Thanks, good point. Main reason for doing this, is because I've already got the entire infra-structure in place for receiving and sending 433 & 868 MHz signals. Many Visonic sensors have FS20 equivalents, so it's no big deal for me, but it'd be fun to add - and since what I do is open source, it might be of use for a slightly different audience.