Weird modulation?

[Niels Ole Kirkeby]

Hi all

Please see this link:

https://triq.org/pdv/#AAB1050A7803D01CB00808FFFF818091818180909090918180909090909181818181818181809090909180909180909090909090909091809091818090918181818180909180918180918091818180A18091818180909090918180909090909181818181818181809090909180909180909090909090909091809091818090918181818180909180918180918091818180A180918181809090909181809090909091818181818181818090909091809091809090909090909090918090918180909181818181809091809181809180918181818181818181B455

Does anyone recognize this protocol/modulation? It is a cheap rain gauge with temperature readings from an danish company. I've tried to find the chip but it's just a "black dot".

The modulation ressembles mostly a PWN or PPM modulation but not quite. There are both long and short pulses and long and short gaps. Nothing I've seen before. Could anyone come up with a guess?

Niels Ole

[Benjamin Larsson]

Hi, it looks like differential PWM can you post a signal recording? That
might help analyzing this

MvH
Benjamin Larsson

[Christian Z.]

Yes, PWM with a sync pulse of 7300 µs. The weirdness is that the gaps preceed the pulses, thus look as they contain data if reading any other way ;)

[Benjamin Larsson]

It looks like it is a PWM with a fixed duration for every pulse +
silence. But it doesn't add up in all places. So it could be something
strange but most like ly just PWM.

MvH
Benjamin Larsson

[Niels Ole Kirkeby]

Here is a recording

[Niels Ole Kirkeby]

The only pattern I can see is that every sequence of long pulse/short gap is preceeded by a short pulse/short gap. Otherwise, besides something which looks like sync-pulses, everything is short pulse/long gap or long pulse, short gap.

[Christian Z.]

Yes, the gaps *preceed* the pulses. A symbol consist of gap+pulse.

The symbols are:

- 960+2690

- 2700+980

And then before the sync symbol of

- 2700+7340

there is some truncation with

- 960+2464

But that's it, just PWM with sync.

rtl_433 -R 0 -X 'n=name,m=OOK_PWM,s=980,l=2700,y=7340,r=3000' sample.cu8

codes     : {66}245c207b7fcc836a0, {65}48b840f6ff9906d40, {71}48b840f6ff9906d400

Three repeats of the same packet "48b840f6ff9906d40" (the first one has a leading bit, that last trailing bits).

[Niels Ole Kirkeby]

@Christian Z. Brilliant! I didn't see that at first. Now I can go on, thanks for the input.