variant of the somfy-rts protocol?

[xjazzor]

Hi there,
I have some Roller Shades at home, which seem to be using the Somfy-RTS protocoll. I sniffed some packets with the rtl_433-plugin for SDR# as seen here:

You can see, everything is recognised except Lines 8 and up. These are the lines, where I want to program a new remote onto the shade.
If I press the same "buttons" on the virtual remote from the great EspSomfy (https://github.com/rstrouse/ESPSomfy-RTS) I get the following:

Are my shades on a different protocol, or do I need to edit something? I can provide more info if needed and want to solve this riddle :D

[zuckschwerdt]

If you can try listening with current rtl_433. Try -R 167:vv to get the raw packets also.
There are some senders with extra data or slightly different meaning, see https://github.com/merbanan/rtl_433/blob/master/src/devices/somfy_rts.c

[xjazzor]

Thank you for your help!
I pulled the data from rtl_433 and attached the readout as a file. As an excerpt, here are some readouts from down, stop, up and then program: (Attention: the examples below are not directly one after the other. The counter increments in single steps)

`_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
time : 2024-08-21 20:00:55
model : Somfy-RTS id : C5154A
Control : Down (8) Counter : 1086 Retransmission: 1 Integrity : CHECKSUM
[pulse_slicer_pcm] Somfy RTS
codes : {179}f0f0f0f0f0f0f0fe696a965a966a9996a95aaa695a5a0
[pulse_slicer_pcm] Somfy RTS
codes : {139}f0f0fe696a965996699996a95aaa695a5a0
[somfy_rts_decode] seed=0x98, chksum=0x5

---

time : 2024-08-21 20:00:58
model : Somfy-RTS id : C5154A
Control : Stop (5) Counter : 1088 Retransmission: 1 Integrity : CHECKSUM
[pulse_slicer_pcm] Somfy RTS
codes : {180}f0f0f0f0f0f0f0fe6999969996a9a6a99665955665650
[somfy_rts_decode] seed=0x95, chksum=0x0

---

time : 2024-08-21 20:01:00
model : Somfy-RTS id : C5154A
Control : Up (6) Counter : 1089 Retransmission: 1 Integrity : CHECKSUM
[pulse_slicer_pcm] Somfy RTS
codes : {180}f0f0f0f0f0f0f0fe6996969a96aaa6a99665955665650
[somfy_rts_decode] seed=0x96, chksum=0x2

---

time : 2024-08-21 20:01:16
model : Somfy-RTS id : C5154A
Control : ? (12) Counter : 1091 Retransmission: 1 Integrity : CHECKSUM
[pulse_slicer_pcm] Somfy RTS
codes : {180}f0f0f0f0f0f0f0fe695a969696a6a6a99665955665650
[somfy_rts_decode] seed=0x9c, chksum=0xa

---

time : 2024-08-21 20:01:16
model : Somfy-RTS id : C5154A
Control : ? (12) Counter : 1091 Retransmission: 1 Integrity : CHECKSUM
[pulse_slicer_pcm] Somfy RTS
codes : {180}f0f0f0f0f0f0f0fe695a969696a6a6a99665955665650
[somfy_rts_decode] seed=0x9c, chksum=0xa
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _`

From my - somewhat limited perspective - it looks, like the remote just spams its id with another seed code - Before it was 0x95 now its 0x9c. I'm just gessing here. ;)

somfy-like-packets.txt

[zuckschwerdt]

Those codes are great to analyze this. Maybe @matthiasseemoo (original author in #1496) and @Velocet (good info in #2356) can help?

[xjazzor]

Should I provide more packets/ recordings? Can I help in any way?
My second remote died and I can't find any shop to buy a new one. Therefore I want this to work. ;)

[Velocet]

There are many RTS variants (at least 5 .. RTS, RTD, RTDS, RTW, RTN..). In case of RTS there are two different protocol modes which differ in length but are compatible to one and another.

You can find my notes i did for somfy-rts here:
RTx: Somfy RTS and its different versions

[xjazzor]

There are many RTS variants (at least 5 .. RTS, RTD, RTDS, RTW, RTN..). In case of RTS there are two different protocol modes which differ in length but are compatible to one and another.

I am with you that they are nearly compatible. Sadly I can't get espsomfy to be paired with my blinds because that parts seems to differ. Maybe in this protocol the pairing sequence is only the remote-id repated 7times?

[gdt]

Where are we on this? Is there a path to a PR to improve rtl_433, or should we just close it?

[xjazzor]

I would love to see a PR to improve rtl_433, as I believe it could benefit not only me but others as well.

[Crsarmv7l]

I noticed in lines 8 and up the counter is the same, which means a long (held) press. With somfy long presses the beginning preamble will be different from the repeats, repeats are shorter preambles. Likely rtl_433 is having a hard time decoding the packets with the short preambles on a button hold.

Also as you noticed, the seed does change with each button press (something I noticed when building a decoder/encoder for the Yardstick One), but only the second nibble and the change isn't really checked by the receiver. I was able to use a constant value for the second nibble.

Something that was a big help to me was using both rtl_433 decoders and protocol info from here (around line 548) plus seeing how they are doing it: https://github.com/DarkFlippers/unleashed-firmware/blob/306b7285879abe05b1e3e7bfdca6442c86161b94/lib/subghz/protocols/somfy_telis.c

Slightly different approaches between the two, but I was able to get enough information with both sources to get rx and tx working.

I would need to go back and look at my notes for more specifics, its been a while since I touched somfy