RTL_433 and Thermopro tp 27

[Freymat]

Hi !
just a short question: is the thermopro TP-27 compatible with RTL_433. I would like to find a 4 probe thermometer and use it with RTL_433.

Thank you !
Mat

[zuckschwerdt]

We have never seen a 4 probe thermometer. But the ThermoPro codes are not complicated. If you get a TP-27 then just send some samples and we'll figure it out. https://triq.org/rtl_433/ANALYZE.html

[Freymat]

Thank you. I ordered one today. I hope the radio transmission is still on 433MHz.

[ngramste]

I was able to record a sample
g021_915M_1000k.zip

[zuckschwerdt]

Thanks. That looks to be related to the known TP28b. This will get the raw data:
rtl_433 -X 'n=tp27,m=FSK_PCM,s=105,l=105,r=5500'

(also drop the cu8 on https://triq.org/pdv/ and select PCM 100 100 to see the data)

There is a preamble of aaaaaaaaaaaaaa but the sync word is not the expected d2aa 2dd4, this sample starts with d236c449… (depending on alignment)

Can you record a few data codes, maybe ramp the temperature up or down slowly so we can spot one of the temperature fields. Just paste a list of codes as text here.

[ngramste]

I have been playing around with it for a while now, this is what I have found without knowing anything haha.

I was using the following settings for RTL_433

rtl_433 -f 915M -X "n=name,m=FSK_PCM,s=100,l=100,match={40}0xb48db11276,r=8172"

I looked at another issue asking about the TP28b and copied their "homework", I have been working in https://triq.net/bitbench to figure out the messages and this is what I found so far while monitoring temperatures on probe 1.

PREAMBLE:hhhhhhhhhhhhhh SYNC:hhhhhhhhhh TEMP_A:hhhhh ?:hhh TEMP_B:hhhhh ?:hhh TEMP_C:hhhhh ?:hhh TEMP_D:hhhhh ?:hhh ?:bb IS_MEAT:bbbb ?bb ?hhhhhhhhhhhhhhhhhhhhhhhhhhh

aaaaaaaaaaaaaab48db11276003d3f803f803f803f803f803f803f80000000000 ---- 24.4C aaaaaaaaaaaaaab48db11276003d7f803f803f803f803f803f803f80000000000 ---- 24.5C aaaaaaaaaaaaaab48db11276003dbf803f803f803f803f803f803f80000000000 ---- 24.6C aaaaaaaaaaaaaab48db11276003dff803f803f803f803f803f803f80000000000 ---- 24.7C aaaaaaaaaaaaaab48db11276003e3f803f803f803f803f803f803f80000000000 ---- 24.8C aaaaaaaaaaaaaab48db11276003e7f803f803f803f803f803f803f80000000000 ---- 24.9C aaaaaaaaaaaaaab48db11276003eff803f803f803f803f803f803f80000000000 ---- 25.1C aaaaaaaaaaaaaab48db11276003f3f803f803f803f803f803f803f80000000000 ---- 25.2C aaaaaaaaaaaaaab48db11276003f7f803f803f803f803f803f803f80000000000 ---- 25.3C aaaaaaaaaaaaaab48db11276003fbf803f803f803f803f803f803f80000000000 ---- 25.4C aaaaaaaaaaaaaab48db11276003fff803f803f803f803f803f803f80000000000 ---- 25.5C

The temp is a hex encoded value that will give you the degrees C when divided by 4. Example for the first entry I captured which reported 24.4 C...
0x003d3 == 979.
979 / 4 = 244

This just seems odd to me as it appears that everything is off by 2 bits which would explain the divide by 4

[zuckschwerdt]

Very good results. That you need to divide by 4 (and that those bits are fixed) indicates that you are off by two bits. Either the field is 2 bits shorter or the whole message needs to be aligned two bits left. See this BitBench with proper temp readings and this BitBench with the shifted message -- with more plausible field lengths and it looks like 8 temp fields, maybe there are setpoints or max values or such?

[ngramste]

I had tried playing around with set points and other settings and I couldn't find anything in the data change. I even tried triggering the different alarm types, low alarm and high alarm, but they all just set the same alarm bit. I will post some more data of the other probes reporting temps so you have that data. I only just learned of bitbench yesterday so I appreciate the help with dialing it in!

[ngramste]

Here are my tests and my bitbench parse string (Edited to include my latest bitbench settings as my link was broken). One additional note, I saw that the CHANGE_ALARM bit was set for the corresponding alarm that was going off for 1 frame. The bits are in reverse order like they are for IS_MEAT. I also noticed multiple bits getting set when the meat was alarming so I will play with that in different alarm modes to see if I can figure out if they yield different results.

Align: b48db11276
Preamble: On

OFFSET?:bb TEMP_A:16d ?:hhhh TEMP_B:16d ?:hhhh TEMP_C:16d ?:hhhh TEMP_D:16d ?:hhhh IS_MEAT:bbbb MEAT_ALARMING:bbbbbbbb ALARM:b SNOOZE:b ?:b IS_FAHRENHEIT:b CHANGE_ALARM:bbbb ?:hhh CHECKSUM?:hh 512x

[ ----------- probe 1 tests --------------- ]
aaaaaaaaaaaaaab48db1127600b2bf803f803f803f803f803f803f800c0242400d802aaaaa80000000000000[71.4f]
aaaaaaaaaaaaaab48db1127600b33f803f803f803f803f803f803f800c02424033002aaaaa80000000000000[71.6f]
aaaaaaaaaaaaaab48db11276012fff803f803f803f803f803f803f800c02424025c02aaaaa80000000000000[119.5f]
aaaaaaaaaaaaaab48db112760135bf803f803f803f803f803f803f800c0242401b802aaaaa80000000000000[123.8f]
aaaaaaaaaaaaaab48db11276004cff803f803f803f803f803f803f800c0242403b802aaaaa80000000000000[30.7f]
aaaaaaaaaaaaaab48db112763f027f803fb6bf803f803f803f803f800c02024022402aaaaa80000000000000[-0.9C]
aaaaaaaaaaaaaab48db112763f04bf803fb6bf803f803f803f803f800c02024005402aaaaa80000000000000[-1.8C]
aaaaaaaaaaaaaab48db11276007a7f803fb6bf803f803f803f803f800c0202403bc02aaaaa80000000000000[48.9C]

[ ----------- probe 2 tests --------------- ]
aaaaaaaaaaaaaab48db112763f803f8000b33f803f803f803f803f800c02424039c02aaaaa80000000000000[71.6f]
aaaaaaaaaaaaaab48db112763f803f80004d7f803f803f803f803f800c0240001ec02aaaaa80000000000000[30.9f]
aaaaaaaaaaaaaab48db112763f803f80006dff803f803f803f803f800c02400005c02aaaaa80000000000000[43.9f]
aaaaaaaaaaaaaab48db112763f803f80007aff803f803f803f803f800c02400027002aaaaa80000000000000[49.1f]

[ ----------- probe 3 tests --------------- ]
aaaaaaaaaaaaaab48db112763f803f803f803f8000a77f803f803f800c0240000bc02aaaaa80000000000000[66.9f]
aaaaaaaaaaaaaab48db112763f803f803f803f80007a7f803f803f800c0240001fc02aaaaa80000000000000[48.9f]
aaaaaaaaaaaaaab48db112763f803f803f803f80005d3f803f803f800c0240002f802aaaaa80000000000000[37.2f]
aaaaaaaaaaaaaab48db112763f803f803f803f80010cbf803f803f800c02400001c02aaaaa80000000000000[107.4f]

[ ----------- probe 4 tests --------------- ]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f8001113f800c0240003e802aaaaa80000000000000[109.2f]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f8001113f800c0240003e802aaaaa80000000000000[108.9f]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f800057ff800c0240003e802aaaaa80000000000000[35.1f]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f8000967f800c02400000c02aaaaa80000000000000[60.1f]

[ ----------- is meat tests, no probes --------------- ]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f800002400039c02aaaaa80000000000000[all hi/low]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f800402400020002aaaaa80000000000000[probe 1 only set to meat]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f80080240000a402aaaaa80000000000000[probe 2 only set to meat]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f80100240001f002aaaaa80000000000000[probe 3 only set to meat]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f802002400035802aaaaa80000000000000[probe 4 only set to meat]

[ ----------- alarm tests, alarm for all is set to meat at 40f --------------- ]
aaaaaaaaaaaaaab48db11276005a3f803f803f803f803f803f803f803c0240001b802aaaaa80000000000000[probe 1 no alarm]
aaaaaaaaaaaaaab48db11276007aff803f803f803f803f803f803f803c47c00033002aaaaa80000000000000[probe 1 alarm]
aaaaaaaaaaaaaab48db112760043ff803f803f803f803f803f803f803c02400033c02aaaaa80000000000000[probe 1 no alarm]

aaaaaaaaaaaaaab48db112763f803f800049ff803f803f803f803f803c02400037402aaaaa80000000000000[probe 2 no alarm]
aaaaaaaaaaaaaab48db112763f803f80006b3f803f803f803f803f803c8bc0402c002aaaaa80000000000000[probe 2 alarm]
aaaaaaaaaaaaaab48db112763f803f8000553f803f803f803f803f803c02400009402aaaaa80000000000000[probe 2 no alarm]

aaaaaaaaaaaaaab48db112763f803f803f803f800052ff803f803f803c0240002bc02aaaaa80000000000000[probe 3 no alarm]
aaaaaaaaaaaaaab48db112763f803f803f803f80006d7f803f803f803d13c0803e802aaaaa80000000000000[probe 3 alarm]
aaaaaaaaaaaaaab48db112763f803f803f803f800055ff803f803f803c02400037002aaaaa80000000000000[probe 3 no alarm]

aaaaaaaaaaaaaab48db112763f803f803f803f803f803f8000583f803c02400036c02aaaaa80000000000000[probe 4 no alarm]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f800095ff803e23c0c01d802aaaaa80000000000000[probe 4 alarm]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f80005a7f803c0240003b802aaaaa80000000000000[probe 4 no alarm]

[ ----------- enable beep alarm on send unit --------------- ]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f803c02400039002aaaaa80000000000000[alarm enabled]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f803c007c0028402aaaaa80000000000000[disable alarm]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f803c0040002d402aaaaa80000000000000[alarm disabled]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f803c027c003c002aaaaa80000000000000[enable alarm]
aaaaaaaaaaaaaab48db112763f803f803f803f803f803f803f803f803c02400039002aaaaa80000000000000[alarm enabled]

[ngramste]

Here is some more data to play with after letting it sit for about 10 minutes and all 4 probes connected. probe 1 was in some melty ice water
tp27.log

[ngramste]

I discovered one more bit, the "snooze" button on the base station clears an alarm bit of sorts

[ ----------- alarm tests, probe 1 hi=250f low=50f --------------- ]
aaaaaaaaaaaaaab48db11276006fbf803f803f803f803f803f803f80380240c024802aaaaa80000000000000[probe 1 no alarm below low]
aaaaaaaaaaaaaab48db1127600927f803f803f803f803f803f803f80380240c004802aaaaa80000000000000[probe 1 no alarm above low]
aaaaaaaaaaaaaab48db1127600717f803f803f803f803f803f803f803847c40008802aaaaa80000000000000[probe 1 alarm 1st frame]
aaaaaaaaaaaaaab48db11276006c3f803f803f803f803f803f803f803847c0000f002aaaaa80000000000000[probe 1 alarm 2nd+ frame]
aaaaaaaaaaaaaab48db1127600727f803f803f803f803f803f803f803846c0000b402aaaaa80000000000000["snooze" alarm on base station]

[ ----------- alarm tests, probe is set to meat at 50f --------------- ]
aaaaaaaaaaaaaab48db112760072ff803f803f803f803f803f803f803c02400020c02aaaaa80000000000000[probe 1 no alarm]
aaaaaaaaaaaaaab48db112760095ff803f803f803f803f803f803f803c47c40033c02aaaaa80000000000000[probe 1 alarm 1st frame]
aaaaaaaaaaaaaab48db112760095ff803f803f803f803f803f803f803c47c00026002aaaaa80000000000000[probe 1 alarm 2nd+ frame]
aaaaaaaaaaaaaab48db1127600973f803f803f803f803f803f803f803c47c00011c02aaaaa80000000000000[probe 1 alarm 2nd+ frame]
aaaaaaaaaaaaaab48db11276009aff803f803f803f803f803f803f803c46c00001402aaaaa80000000000000[probe 1 "snooze"]

[ngramste]

I have been trying to figure out how the checksum of this message works and I have hit a wall. This is the latest I have in bitbench

[zuckschwerdt]

Try e.g. 2x 56x h SYNC:32h ?:h TEMP_A:16d ?:16h TEMP_B:16d ?:16h TEMP_C:16d ?:16h TEMP_D:16d ?:16h IS_MEAT:4b MEAT_ALARMING:8b ALARM:b SNOOZE:b ?:b IS_FAHRENHEIT:b CHANGE_ALARM:4b ?:12b CHECKSUM?:8h 512x
where 56x is the preamble. Skipping the preamble the next 25 bytes are a CRC-8, poly 0x07, init 0xe4.
But you could also skip more bytes, e.g. without the first byte (which may be sync) it's init 0x82.