Salus RT505TX

Adrian Stevenson

unread,

Oct 25, 2017, 5:12:54 PM10/25/17

to rtl_433

I am attempting a new device decoder for the Salus RT505TX room thermostat but I need some advice and guidance.

At the moment I have the following limits set:

r_device salus = {

.name = "Salus RT505TX",

.modulation = FSK_PULSE_PCM,

.short_limit = 58,

.long_limit = 58,

.reset_limit = 9937, //no idea what this should be?

.json_callback = &salus_callback,

.disabled = 0,

.demod_arg = 1,

.fields = output_fields

};

When I look at the signal analysis I can see that the modulation seems to change part way through. 37 Pulses with a width of 14511 (short distance 58 ms) then 22 pulses with a width of 6985 with a short distance of 28 ms).

After 0.262144s the signal is then repeated another 5 times.

Are there any other devices that exhibit this behaviour, I can then hopefully look at the code and learn from?

What should the reset limit be set at?

Detected FSK package @ @0.000000s

Analyzing pulses...

Total count: 37, width: 14511 (58.0 ms)

Pulse width distribution:

[ 0] count: 1, width: 12 [12;12] ( 48 us)

[ 1] count: 25, width: 104 [104;105] ( 416 us)

[ 2] count: 6, width: 208 [208;209] ( 832 us)

[ 3] count: 2, width: 313 [313;313] (1252 us)

[ 4] count: 2, width: 418 [418;418] (1672 us)

[ 5] count: 1, width: 147 [147;147] ( 588 us)

Gap width distribution:

[ 0] count: 3, width: 295 [257;314] (1180 us)

[ 1] count: 21, width: 104 [104;105] ( 416 us)

[ 2] count: 6, width: 209 [209;209] ( 836 us)

[ 3] count: 1, width: 1462 [1462;1462] (5848 us)

[ 4] count: 4, width: 600 [522;627] (2400 us)

[ 5] count: 1, width: 835 [835;835] (3340 us)

Pulse period distribution:

[ 0] count: 5, width: 304 [269;314] (1216 us)

[ 1] count: 18, width: 208 [208;209] ( 832 us)

[ 2] count: 6, width: 417 [417;418] (1668 us)

[ 3] count: 2, width: 1409 [1253;1566] (5636 us)

[ 4] count: 2, width: 574 [522;626] (2296 us)

[ 5] count: 2, width: 783 [731;835] (3132 us)

[ 6] count: 1, width: 1044 [1044;1044] (4176 us)

Level estimates [high, low]: 15483, 224

Frequency offsets [F1, F2]: 17418, -15651 (+66.4 kHz, -59.7 kHz)

Guessing modulation: No clue...

Total count: 22, width: 6985 (27.9 ms)

Pulse width distribution:

[ 0] count: 1, width: 0 [ 0; 0] ( 0 us)

[ 1] count: 2, width: 331 [314;348] (1324 us)

[ 2] count: 17, width: 104 [104;105] ( 416 us)

[ 3] count: 1, width: 209 [209;209] ( 836 us)

[ 4] count: 1, width: 154 [154;154] ( 616 us)

Gap width distribution:

[ 0] count: 1, width: 10 [10;10] ( 40 us)

[ 1] count: 15, width: 104 [104;105] ( 416 us)

[ 2] count: 3, width: 313 [313;313] (1252 us)

[ 3] count: 1, width: 522 [522;522] (2088 us)

[ 4] count: 1, width: 1148 [1148;1148] (4592 us)

Pulse period distribution:

[ 0] count: 1, width: 10 [10;10] ( 40 us)

[ 1] count: 5, width: 424 [417;453] (1696 us)

[ 2] count: 12, width: 208 [208;209] ( 832 us)

[ 3] count: 1, width: 313 [313;313] (1252 us)

[ 4] count: 1, width: 626 [626;626] (2504 us)

[ 5] count: 1, width: 1253 [1253;1253] (5012 us)

Level estimates [high, low]: 15522, 266

Frequency offsets [F1, F2]: 17593, -13617 (+67.1 kHz, -51.9 kHz)

Guessing modulation: No clue...

Detected FSK package @ @0.262144s

Analyzing pulses...

Total count: 37, width: 14503 (58.0 ms)

Pulse width distribution:

[ 0] count: 1, width: 0 [ 0; 0] ( 0 us)

[ 1] count: 3, width: 323 [313;342] (1292 us)

[ 2] count: 24, width: 104 [104;105] ( 416 us)

[ 3] count: 6, width: 209 [209;209] ( 836 us)

[ 4] count: 2, width: 417 [417;418] (1668 us)

[ 5] count: 1, width: 154 [154;154] ( 616 us)

Gap width distribution:

[ 0] count: 1, width: 16 [16;16] ( 64 us)

[ 1] count: 21, width: 104 [104;106] ( 416 us)

[ 2] count: 2, width: 313 [313;314] (1252 us)

[ 3] count: 6, width: 209 [209;209] ( 836 us)

[ 4] count: 1, width: 1462 [1462;1462] (5848 us)

[ 5] count: 4, width: 600 [522;627] (2400 us)

[ 6] count: 1, width: 836 [836;836] (3344 us)

Pulse period distribution:

[ 0] count: 1, width: 16 [16;16] ( 64 us)

[ 1] count: 8, width: 434 [417;522] (1736 us)

[ 2] count: 17, width: 208 [208;209] ( 832 us)

[ 3] count: 4, width: 313 [313;313] (1252 us)

[ 4] count: 2, width: 1409 [1253;1566] (5636 us)

[ 5] count: 3, width: 730 [626;835] (2920 us)

[ 6] count: 1, width: 1044 [1044;1044] (4176 us)

Level estimates [high, low]: 15503, 330

Frequency offsets [F1, F2]: 15144, -11583 (+57.8 kHz, -44.2 kHz)

Guessing modulation: No clue...

Detected FSK package @ @0.262144s

@0.262144s

Analyzing pulses...

Total count: 22, width: 6986 (27.9 ms)

Pulse width distribution:

[ 0] count: 1, width: 0 [ 0; 0] ( 0 us)

[ 1] count: 2, width: 327 [313;341] (1308 us)

[ 2] count: 17, width: 104 [104;105] ( 416 us)

[ 3] count: 1, width: 208 [208;208] ( 832 us)

[ 4] count: 1, width: 154 [154;154] ( 616 us)

Gap width distribution:

[ 0] count: 1, width: 17 [17;17] ( 68 us)

[ 1] count: 15, width: 104 [104;105] ( 416 us)

[ 2] count: 3, width: 313 [313;314] (1252 us)

[ 3] count: 1, width: 522 [522;522] (2088 us)

[ 4] count: 1, width: 1148 [1148;1148] (4592 us)

Pulse period distribution:

[ 0] count: 1, width: 17 [17;17] ( 68 us)

[ 1] count: 5, width: 423 [417;446] (1692 us)

[ 2] count: 12, width: 208 [208;209] ( 832 us)

[ 3] count: 1, width: 313 [313;313] (1252 us)

[ 4] count: 1, width: 627 [627;627] (2508 us)

[ 5] count: 1, width: 1252 [1252;1252] (5008 us)

Level estimates [high, low]: 15447, 314

Frequency offsets [F1, F2]: 15352, -12955 (+58.6 kHz, -49.4 kHz)

Guessing modulation: No clue...

At the moment the output from the decoder looks like this:

Thermostat On Signal

[00] {478} 7f ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f fc

[00] {478} 7f ff ff 00 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f f8

[00] {478} 7f ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f fc

[00] {478} 7f ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f f8

[00] {478} 7f ff ff 00 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f f0

[00] {478} 7f ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f f0

[00] {478} 7f ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f f8

[00] {478} bf ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f fc

[00] {478} 7f ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 00 7f 00 00 03 f8 00 00 1f f8

Thermostat Off Signal

[00] {478} 7f ff ff 00 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 3f 80 00 00 03 f8 00 0f e0 00

[00] {478} 7f ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 3f 80 00 00 03 f8 00 0f e0 00

[00] {478} 7f ff ff 00 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 3f 80 00 00 03 f8 00 0f e0 00

[00] {478} 7f ff ff 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 3f 80 00 00 03 f8 00 0f e0 00

[00] {478} 7f ff ff 00 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 3f 80 fe 03 f8 0f e0 00 00 7f 01 ff f8 07 ff ff e0 3f 80 fe 00 00 00 00 1f c0 00 00 00 00 00 00 00 00 3f 80 00 00 03 f8 00 0f e0 00

Bytes 51 & 52

0000 0000 0111 1111=007f= on

0011 1111 1000 0000=3f80= off

Christian Zuckschwerdt

unread,

Oct 26, 2017, 3:22:34 AM10/26/17

to rtl_433

If this is really FSK then then width distribution and talking about pulses is not very meaningful. Look at the spectrogram, and analyze the signal in Audacity. Post a good sample here for others to take a look maybe.

Adrian Stevenson

unread,

Nov 4, 2017, 10:03:27 AM11/4/17

to rtl_433

Christian,

Thanks for the response to my post.

I have uploaded the Salus samples to rtl_433_tests/ and have attached some files here.

Viewing the waveform signal with Audacity,I can see that the waveform frequency is varying and as I am getting some consistent data using FSK_PULSE_PCM.

I think, I just need some confirmation/guidance about the limit levels and how to get all 20 signals into the same bit_buffer array.

Thanks

Adrian

Salus RT505TX.png

gfile003.data

Salus RT505TX-Zoomed-out.png

Salus RT505TX-spectrogram.png

Christian Zuckschwerdt

unread,

Dec 4, 2017, 11:52:14 AM12/4/17

to rtl_433

The Salus RT505TX transmission has 20 packets with alternating long and short packets.
There are 8162 samples between packets, long packet 14560 samples, short packet 7026 samples.
A bit is 105 samples (about 8x the width I expect to see @250kHz -- this is sampled with 2.048MHz, right?)
The packet starts with 3-4 zeros and ~12 one/zero sync (one/zero can be interchanged, it's just f1/f2)

A quick trial gives (which looks reasonable):

{139} 15 55 55 45 ba 94 a2 00 04 1a c0 cf 00 a1 00 e6 4d e0
{66} ea aa aa 8b 75 04 00 44 40
{138} ea aa aa 8b 75 29 44 00 08 35 81 9e 01 42 01 cc 9b c0
{66} ea aa aa 8b 75 04 00 44 40
{138} ea aa aa 8b 75 29 44 00 08 35 81 9e 01 42 01 cc 9b c0
{66} ea aa aa 8b 75 04 00 44 40
{138} ea aa aa 8b 75 29 44 00 08 35 81 9e 01 42 01 cc 9b c0
{66} ea aa aa 8b 75 04 00 44 40
{138} ea aa aa 8b 75 29 44 00 08 35 81 9e 01 42 01 cc 9b c0
{66} ea aa aa 8b 75 04 00 44 40
{138} ea aa aa 8b 75 29 44 00 08 35 81 9e 01 42 01 cc 9b c0
{66} ea aa aa 8b 75 04 00 44 40
{138} ea aa aa 8b 75 29 44 00 08 35 81 9e 01 42 01 cc 9b c0
{66} ea aa aa 8b 75 04 00 44 40
{138} ea aa aa 8b 75 29 44 00 08 35 81 9e 01 42 01 cc 9b c0
{66} ea aa aa 8b 75 04 00 44 40
{139} 75 55 55 45 ba 94 a2 00 04 1a c0 cf 00 a1 00 e6 4d e0
{66} ea aa aa 8b 75 04 00 44 40
{138} ea aa aa 8b 75 29 44 00 08 35 81 9e 01 42 01 cc 9b c0
{67} 75 55 55 45 ba 82 00 22 20

The reset_limit should be 8162/2048000*1000000 => 4000 but that is too long, so use 800 µs (wider than your longest stretch of 0 or 1).

Adrian Stevenson

unread,

Dec 19, 2017, 9:43:00 AM12/19/17

to rtl_433

Hi Christian,

Your example output looks a lot better than what I have. What modulation and reset limits are you using?

see http://www.datasheetlib.com/datasheet/1098623/rf02_hope-microelectronics.html?page=18#datasheet

The sampling rate would of been the default of 250000.

With regards "The reset_limit should be 8162/2048000*1000000 => 4000"

Should that say 8162 / 2048 000 000 Hz * 1000000 = 3.9 us or put another way 8162 / 2048 000 Khz * 1000 = 3.9 us

but the sampling is 250 000 so 8162 /250 = 32 us which would be too short..

There is some related information here https://the.cyclingengineer.co.uk/2013/11/23/home-automation-integrating-salus-rt500-rf-in-openhab-using-a-jeelink/

Thanks

Christian Zuckschwerdt

unread,

Dec 19, 2017, 10:20:42 AM12/19/17

to rtl_433

Re: reset_limit "8162/2048000*1000000 => 4000" breaks down to: 8162 samples by 2048k sample rate = seconds, then times 10e6 to get microseconds.
I.e. I assumed 4000 us. With the default 250k it's 32648. 32 ms! Thats really slow/long.

Use http://triq.net/iqs -- drop some sample in, set to "x8" and you can roughly read the timing.

It's basic FSK, PCM coding with a pulse width of 105 samples @250k = 420 us. Try the feat-gedecode branch with:

rtl_433 -q -F json -X salus:FSK_PCM:420:420:33000 -r rtl_433_tests/tests/salus/01/gfile001.cu8

Reply all

Reply to author

Forward